Apoptotic mesenchymal stem cells and their secreted apoptotic extracellular vesicles: therapeutic applications and mechanisms

Abstract

Mesenchymal stem cells (MSCs), an accessible and less ethically controversial class of adult stem cells, have demonstrated significant efficacy in treating a wide range of diseases in both the preclinical and clinical phases. However, we do not yet have a clear understanding of the mechanisms by which MSCs exert their therapeutic effects in vivo. We found that the transplanted MSCs go an apoptotic fate within 24 h in vivo irrespective of the route of administration. Still, the short-term survival of MSCs do not affect their long-term therapeutic efficacy. An increasing number of studies have demonstrated that transplantation of apoptotic MSCs (ApoMSCs) show similar or even better efficacy than viable MSCs, including a variety of preclinical disease models such as inflammatory diseases, skin damage, bone damage, organ damage, etc. Although the exact mechanism has yet to be explored, recent studies have shown that transplanted MSCs undergo apoptosis in vivo and are phagocytosed by phagocytes, thereby exerting immunomodulatory effects. The apoptotic extracellular vesicles secreted by ApoMSCs (MSC-ApoEVs) play a significant role in promoting immunomodulation, endogenous stem cell regeneration, and angiogenesis due to their apoptotic properties and inheritance of molecular characteristics from their parental MSCs. On this basis, this review aims to deeply explore the therapeutic applications and mechanisms of ApoMSCs and their secretion of MSC-ApoEVs, as well as the challenges they face.

Background

The efficacy of mesenchymal stem cells (MSCs) in a wide range of preclinical disease models has been well documented over the past decades [1,2,3,4,5]. There is even an increasing number of clinical trial cases using MSCs as a therapeutic tool [6,7,8,9]. For a long time, it has been believed that surviving MSCs were essential for therapeutic efficacy and their therapeutic effects depended mainly on their differentiation properties, paracrine mechanisms, and intercellular surface molecular interactions [10,11,12,13,14,15]. However, recent studies have shown that transplanted MSCs may go into apoptosis within a short period of time because of the disease microenvironment such as hypoxia, ischemia, and inflammation, but apoptosis does not affect the efficacy of MSCs [16]. Thus, recent studies have proposed a new concept that apoptosis of MSCs in vivo plays a key role in the treatment of disease.

As early as 2005, thum et al. proposed a new immunomodulatory mechanism: the ‘dying stem cell’ hypothesis [17]. Specifically, the authors found that the viability of peripheral blood or whole bone marrow MSCs (BMMSCs) isolated from patients with acute myocardial infarction for transplantation ranged from 75 to 95%, implying that some of the cells had undergone apoptosis and necrosis prior to transplantation [17]. The influence of the myocardial infarction microenvironment on the viability of transplanted MSCs was implicated [17]. They proposed a new insight into the effect of transplanted MSCs on improving cardiac function: transplanted MSCs that underwent apoptosis in vivo would recruit immune cells and suppress inflammatory macrophages and dendritic cells [17]. Although this was hypothesized, a growing number of recent studies have confirmed this hypothesis. They found that MSCs underwent extensive apoptosis within hours of infusion and were subsequently engulfed by local myeloid phagocytes, which were the ultimate players in the anti-inflammatory response [18].

With more and more research, direct transplantation of apoptotic MSCs (ApoMSCs) induced in vitro were preferred by more and more researchers for more efficient treatment. It has been found that transplantation of ApoMSCs showed similar or even better efficacy than viable MSCs, which has been demonstrated in various preclinical diseases such as autoimmune diseases and inflammatory diseases [16, 19]. Studies applying anti-ApoMSCs on animal disease models revealed limited immunomodulatory effects and diminished efficacy.

MSCs may be exposed to the risk of pulmonary embolism due to their excessive size [20]. In response to the popularity of cell-free therapies, the apoptotic extracellular vesicles secreted by ApoMSCs (MSC-ApoEVs) have been used by an increasing number of researchers as a result of ApoMSCs [21]. Due to their phagocytic targeting like ApoMSCs and carrying abundant functional proteins and RNAs from their parental MSCs, they have the potential to be more direct therapeutic agents than ApoMSCs themselves. It has been demonstrated in preclinical disease models such as bone damage, skin damage, dental damage, diabetes, cancer, and inflammatory diseases [22]. MSC-ApoEVs have been explored as drug carriers due to their easy availability, high yield, and high drug loading rate [10, 23].

In this review, we first summarized the spontaneous apoptotic fate of transplanted MSCs in vivo provided evidence for the therapeutic efficacy of MSCs after apoptosis, and explored the causes of apoptosis occurs in MSCs. In addition, we summarized in detail the specific applications and mechanisms of action of ApoMSCs and their MSC-ApoEVs for therapy. Finally, we analyzed and concluded the challenges and prospects arising from the therapeutic application of ApoMSCs and MSC-ApoEVs.

Spontaneous apoptotic fate of transplanted MSCs

The fate of MSCs with different routes of administration for transplantation to undergo apoptosis in vivo has been confirmed by an increasing number of studies [24]. Intravenous injection was the most common, minimally invasive, and highly reproducible route of administration, with data suggesting that it accounted for 43% of clinical trials [25]. Various MSCs detection techniques such as MSCs constitutively expressing fluorescent proteins or luciferase, or MSCs labeled with fluorescent dyes or radiotracers were used to study the biodistribution of MSCs, which showed that intravenously infused MSCs rarely reached the transplantation damage site, and most MSCs were trapped in the lungs due to their inability to pass through the capillary system of the lungs and were later redistributed to liver and spleen, and almost none at other sites(< 0.1%) [20, 26, 27]. Different tissue culture or polymerase chain reaction (PCR) of MSCs-specific marker genes results showed that surviving MSCs rapidly decreased and disappeared within only 24 h [26, 28]. Interestingly, more than half of the fluorescence-positive MSCs in the lungs showed co-expression of calreticulin within 30 min after MSCs injection and exhibited a high level of expression of the apoptotic marker caspase 3 within 1 h [26, 29]. These suggested that intravenously injected MSCs underwent rapid apoptosis during their stay in the lungs [26, 29]. Among them, calreticulin is a marker that serves as a signal for the host immune system to consecutively induce phagocytosis [26]. After 24 h of intravenous infusion of MSCs, it was found that MSCs were contained in monocytes of predominantly non-classical Ly6C^low phenotype, and MSCs-containing monocytes were also detected systemically, further demonstrating that intravenously transplanted MSCs interacted with phagocytosis after apoptosis occurred [30].

In addition to the survival status of MSCs, it was found that most MSCs were no longer complete soon after transplantation [26]. Using microscopy to analyze the size changes of green fluorescent MSCs, it was found that fluorescent MSCs in transplanted mouse lungs appeared to be smaller, and Hoechst 33,342 nuclear staining showed that most of the nuclear signals were absent in fluorescence-positive MSCs after transplantation, indicating that MSCs continued to decrease in size and undergo fragmentation after transplantation [26]. Taken together, these suggested that MSCs underwent apoptosis, fragmentation, and phagocytosis by immune cells after intravenous transplantation.

Direct injection of MSCs into damaged tissues showed similar results to intravenous injection, which reported apoptosis of locally transplanted MSCs after 3–5 days and the appearance of almost all locally transplanted MSCs in tissue-specific phagocytes within a week [27, 31]. Preda et al. explored the fate of MSCs under different routes of administration by using a dual tracking system for MSCs, i.e., luciferase expression and VivoTrack680 labeling and in vivo optical imaging. It was found that MSCs infused into intraventricular, intrapancreatic, intrasplenic, and subcutaneous areas of diabetic mice activate hypoxic signaling pathways within one day after transplantation, followed by caspase 3-mediated apoptosis [16]. Immune cells were locally recruited at the transplantation site, resulting in macrophage phagocytosis of apoptotic MSCs [16].

Impact of spontaneous apoptotic MSCs on diseased animals

The short-term survival of MSCs did not seem to interfere with their effectiveness, as the beneficial effects of MSCs remained potent after apoptosis [30]. In graft-versus-host disease (GVHD) mice, 1 h after MSCs infusion, extensive apoptosis and perforin-dependent caspase activation were observed due to the presence of cytotoxic immune cells [32]. This was necessary for the MSCs to function as an immunosuppressor [32]. MSCs infusion effectively reduced GVHD effector cells infiltrating the spleen and lungs by up to 10% compared to untreated GVHD mice [32]. When MSCs were intravenously infused in acute liver-injured mice, MSCs underwent apoptosis within 2–4 h after [33]. With the elevation of phosphatidylserine (PS) levels on apoptotic MSCs, monocyte-derived macrophages were recruited for phagocytosis and shifted to the M2 phenotype [33]. The transplantation of spontaneously apoptotic MSCs was effective in ameliorating acute liver injury in mice after 12 h of treatment [33]. The area of hepatic necrosis was reduced, the serum levels of alanine aminotransferase and aspartate aminotransferase were restored to normal, and the number of apoptotic hepatocytes in mice was also significantly elevated [33]. In myocardial infarction mice, direct intracardiac injection of MSCs underwent apoptosis within 24 h [34]. However, the left ventricular function and the fibrotic area of cardiac infarcts were effectively improved in mice after 4 weeks of MSCs transplantation compared with phosphate buffered saline-treated controls [34].

Studies applying anti-ApoMSCs on animal disease models revealed limited immunomodulatory effects and diminished efficacy [29, 35]. Liu et al. established anti-ApoMSCs by pretreatment with the caspase 3 inhibitor Z-DEVD-FMK [35]. Although there was some therapeutic effect, the Z-DEVD-FMK-treated MSCs secreted less anti-inflammatory factor tumor necrosis factor-alpha- -stimulated gene 6 (TSG-6), and their ability to regulate inflammation and inhibit scar formation was reduced compared with viable MSCs and H₂O₂-treated ApoMSCs [35]. Among them, H₂O₂-treated ApoMSCs had the best therapeutic effects [35]. Pang et al. established anti-ApoMSCs to treat ovalbumin (OVA) -sensitized asthmatic mice by deleting BAK/BAX (key proteins for apoptosis) in MSCs through the caspase 9 ribonucleoprotein complexes introduced via electroporation [29]. Although they were efficacious in improving lung histology, the extent of treatment was reduced compared to the viable MSCs-treated group [29]. In terms of immunosuppressive capacity, anti-ApoMSCs were shown not to downregulate the proliferation of activated T cells in diseased mice [29]. Serum specific pro-inflammatory factors interleukin (IL)-5 and IL-13 were also less down-regulated compared to the viable MSCs-treated group [29]. These therapeutic results with anti-ApoMSCs further side-step the fact that apoptosis of MSCs was necessary for immunosuppression and therapeutic efficacy in vivo.

The causes of spontaneous apoptosis of transplanted MSCs

Currently, the specific cause of apoptosis of MSCs after transplantation is still unclear. It is considered that the microenvironment at disease conditions might influence the fate of transplanted MSCs by factors like cytotoxic immune cells, hypoxia, inflammatory factors, nutritional deficiencies, microorganisms, etc [36].

Galleu et al. speculated that apoptosis of transplanted MSCs involves the presence of cytotoxic immune cells such as CD8⁺ T cells and CD56⁺ natural killer (NK) cells [32]. In GVHD mice, it was found that the apoptosis of MSCs was induced only in diseased mice infused with CD8⁺ T cells, but not in those not infused with CD8⁺ T cells [32]. The same result was also obtained by using Mh/ perforin knockout mice (defective cytotoxic GVHD effector cells) as the recipients [32]. Similar cytotoxicity on MSCs observation was made in an in vitro co-culture study [32]. Peripheral blood mononuclear cells (PBMCs) were isolated from GVHD patients who had favorable therapeutic outcomes with MSCs treatment (noted as PBMC⁺) and co-cultured with human BMMSCs for 4 h [32]. In contrast to the co-culture of PBMCs collected from patients who had poor response to MSCs treatment (noted as PBMCs⁻), apoptosis of MSCs in PBMCs⁺ was observed to be upregulated 4-fold by flow cytometry [32]. Among them, CD56⁺ NK cells and CD8⁺ T cells populations in the PBMCs⁺ were the cells responsible for initiating MSCs apoptosis [32]. Apoptosis of MSCs may be caused by these toxic immune cells via granzyme B release and perforin-dependent and Fas/FasL ((Fas ligand) pathways [32]. This result was similarly validated in PBMCs extracted from patients with Crohn’s disease [37]. Compared to PBMCs⁻, the PBMC⁺ group induced more significant human adipose-derived MSCs (ADMSCs) to apoptosis and higher levels of prostaglandin E2 (PGE2) secreted from human ADMSCs [37]. Researchers believed that these cytotoxic immune cells may serve as a clinical biomarker to assist practitioners in predicting the treatment effect of MSCs therapy as precision medicine [32, 37].

It has also been claimed that apoptosis of transplanted MSCs in vivo occurred due to hypoxia at the damaged site [16]. In diabetes mice, Preda et al. found that MSCs transplanted in the pancreas transiently triggered hypoxia-induced signaling pathways and activated caspase 3/7-mediated apoptosis in the presence of pro-inflammatory factors such as tumor necrosis factor-alpha (TNF-α) and interferon-gamma (IFN-γ) [16]. Interestingly, the authors found that both hypoxia and pro-inflammatory factors are essential for apoptosis [16]. In periodontitis mice, the expression level of the hypoxia-related factor hypoxia-inducible factor 1 alpha (HIF-1α) was significantly elevated in periodontitis tissues, demonstrating the presence of a hypoxic microenvironment, which induced apoptosis in the transplanted MSCs [38].

MSCs have been shown to treat a variety of inflammatory diseases, but few have explored the impact of the inflammatory environment on the fate of transplanted MSCs. Li et al. demonstrated that the presence of IFN-γ and TNF-α stimulated the high expression of inducible nitric oxide synthase in MSCs, which contributed to the production of intracellular nitric oxide (NO), thus up-regulating the expression of Fas on the surface of MSCs and aggravating endoplasmic reticulum stress to induce apoptosis [39].

In the ischemic heart, exogenous MSCs were exposed to a variety of pro-apoptotic or cytotoxic factors such as inadequate blood supply, hypoxic pressure, and inflammation triggered by tissue damage that may produce oxidative stress, cytotoxic free radicals, and proteins, all of which may trigger apoptosis [40]. After infusion of MSCs, any sites, and environments they pass through may induce any form of cell death, so the mechanism of apoptosis in transplanted MSCs in vivo remains a major challenge that is complex and still needs to be explored in depth.

Therapeutic applications of ApoMSCs and MSC-ApoEVs

Therapeutic applications of ApoMSCs

A growing number of studies have demonstrated that apoptosis of MSCs in vivo was necessary for their therapeutic effects. Therefore, infusion of ApoMSCs in vivo may be a more direct and efficient therapeutic option. Currently, the efficacy of ApoMSCs in various preclinical disease models such as sepsis, acute tissue injury, acute inflammation, autoimmune disease, and proliferative scarring of the skin has shown exciting results, as shown in Table 1.

Table 1 Typical therapeutic applications of ApoMSCs from different sources

Sepsis is commonly associated with the persistence of infection-induced immune dysregulation, with a strong attack of the hyperinflammatory response leading to functional organ dysfunction [41, 42]. Chang et al. effectively upregulated the survival rate of septic rats by using apoptotic ADMSCs(Apo-ADMSCs) [43, 44]. They effectively ameliorated disease-related cardiac, hepatic, pulmonary, and renal injuries through anti-inflammatory, anti-apoptotic, and antioxidant effects [43, 44]. To explore the acute immune response, the study observed peripheral blood and spleen-derived serum [43, 44]. It was found that circulating TNF-α levels in peripheral blood decreased significantly after treatment with Apo-ADMSCs, as well as the levels of toxic T-lymphocytes [43, 44].

In acute lung injury, apoptotic human umbilical cord MSCs ( Apo-hUCMSCs) have been shown to have better therapeutic effects than viable hUCMSCs [45]. ApoMSCs effectively reduced neutrophil infiltration and total protein infiltration in rat lungs, and improved alveolar-capillary membrane integrity [45]. The expression levels of relevant inflammatory factors such as the pro-inflammatory factor INF-γ decreased and the anti-inflammatory factor IL-4 increased in lung tissues [45].

Many studies have shown that melatonin could be effective in maintaining body function and homeostasis by down-regulating inflammatory and oxidative responses [46]. Interestingly, Yip et al. combined melatonin with Apo-ADMSCs for the treatment of acute pulmonary ischemia-reperfusion rats, and found that the therapeutic efficacy was superior to that of the melatonin group, and even superior to that of the melatonin-combined-viable MSCs group [47]. Combined treatment with melatonin and Apo-ADMSCs better-ameliorated hypoxemia and pulmonary arterial hypertension in the diseased mice [47]. Both inflammatory and oxidative protein markers were significantly decreased in the lung tissues, while antioxidant indicator levels increased [47]. The study also revealed anti-apoptotic effects in the tissues, such as a significant decrease in the levels of apoptotic markers phosphorylated H2AX (γ-H2AX), cleaved caspase 3, and poly (ADP-ribose) polymerase proteins (PARP), as well as an increase in the protein levels of anti-apoptotic protein markers such as Bcl-2 [47]. Recent studies have shown that ApoMSCs have more potent anti-inflammatory and oxidative stress-reducing properties than viable MSCs, which partly explains why melatonin works better in combination with Apo-ADMSCs [43, 44]. Overall, these studies offered better prospects for the treatment of ApoMSCs. See Fig. 1 for details.

Fig. 1

Therapeutic applications of ApoMSCs and MSC-ApoEVs

MSCs induced apoptosis in vitro in various ways such as STS, inflammatory factor, H₂O₂, hypoxia, UVC light, and serum free. ApoMSCs and MSC-ApoEVs have great potential for the treatment of a wide range of diseases. Among them, for the characterization of MSC-ApoEVs, they were found to express apoptosis-related markers and EVs-related markers.

Therapeutic applications of MSC-ApoEVs

Biological properties of MSC-ApoEVs

MSCs underwent apoptosis during which they released apoptotic extracellular vesicles, which resulted from apoptosis [21]. This predicted that MSC-ApoEVs may play a similarly important role as ApoMSCs in the treatment of disease [21]. Before summarizing the therapeutic applications of MSC-ApoEVs, it is necessary to introduce the biological properties of MSC-ApoEVs.

Extracellular vesicles (EVs) are membrane-bound particles released by cells into the extracellular space [48]. EVs have played an important role in the diagnosis and treatment of various diseases due to the carrying of a wide range of molecules derived from the parental cells, such as lipids, proteins, nucleic acids, and metabolites [49]. Even EVs are increasingly being used as drug-delivery vehicles due to their favorable safety and unique biodistribution capabilities [50].

Classically, EVs can be classified into three subtypes depending on their origin: exosomes (Exos) (produced by endosomes, ∅= 30–150 nm in diameter), microvesicles (MVs) (generated at the plasma membrane from its outward budding, ranging in diameter from less than 100 nm to a few micrometers) and apoptotic extracellular vesicles (ApoEVs) (produced by apoptotic cells, ∅= 50–5000 nm in diameter) [23, 51, 52]. EVs produced by viable cells, including Exos and MVs, have been extensively studied. However, most people still limit their knowledge of ApoEVs to the level of the apoptotic bodies (ApoBDs) with large diameters [53,54,55,56]. Currently, ApoEVs can be classified into three subtypes according to their biogenesis and size: ApoBDs (∅=1–5 μm in diameter), microvesicle-like ApoEVs (ApoMVs) (∅=100–1000 nm in diameter), and exosome-like ApoEVs (ApoExos) (∅< 150 nm in diameter) [52, 57].

The biogenesis of ApoBDs is mainly related to actomyosin reorganization in the late stages of apoptosis, where the plasma membrane blebbing, protrusions, and fragmentation produce membrane-bound ApoBDs of varying sizes, which prevents the loss of cellular contents, especially some pro-inflammatory substances [55, 58].

For the biogenesis of ApoMVs, they may be generated during the early stages of apoptosis and by a mechanism like MVs production from viable cells: plasma membrane budding, but the exact molecular regulatory mechanism is not yet clear [57, 59].

ApoExos are also like Exos production from viable cells: they are released upon fusion of the multivesicular body with the plasma membrane [57, 60]. ApoExos were analyzed using transmission electron microscopy, where multivesicular endosomes containing internal vesicles were visible [60]. The characterization revealed the expression of Exos-associated protein markers such as CD63 and CD9 [60]. The mechanism of the formation may involve the sphingosine-1-phosphate/sphingosine-1-phosphate receptors signaling pathway, the autophagy-lysosomal pathway, or the Gasdermins-mediated pathway [60,61,62].

Functionally, ApoEVs, as key messengers released by apoptotic cells, not only recruit peripheral phagocytes for effective removal of apoptotic substances but also contain a variety of biomolecules, including microRNAs and proteins, to facilitate intercellular communication [57]. ApoEVs derived from a variety of different cells, such as MSCs, immune cells, and endothelial cells, have been found to play important roles in efferocytosis, immunomodulation, cell proliferation and survival, and angiogenesis [52, 63]. Here, we focus on MSC-ApoEVs.

Unlike the method of collecting viable MSC-EVs for therapy by centrifugation at 120,000 g, the isolation of MSC-ApoEVs for therapy is currently collected mainly by centrifugation at 16,000 g. Transmission electron microscopy and nanoparticle tracking analysis showed a larger and wider range of diameters than viable MSC-EVs, generally ranging from 150 to 1000 nm. Notably, the method of isolation was unable to distinguish between EVs from healthy, stressed, etc. with the same density and particle size, which highlight the limitations of isolating and purifying MSC-ApoEVs. They were characterized to express apoptosis-related specific surface markers such as PS, calreticulin (CRT), and component 1q (C1q) and to express apoptosis-specific proteins including cleaved caspase 3. Even EVs-related markers such as transmembrane/lipid-bound protein (e.g., CD63, CD9, CD81, Flotillin1) and cytosolic protein (e.g., tumor susceptibility gene 101 (TSG101), Alix) were characterized in MSC-ApoEVs. It predicted that the biogenesis of MSC-ApoEVs for therapy may involve exosome-like biogenesis. Related markers of parental MSCs: CD29(+), CD44(+), CD90(+), CD34(-) and CD45(-) were also characterized, as shown in Table 2.

Table 2 Characterization of MSC-ApoEVs

It has been shown that MSC-ApoEVs contain a variety of contents such as proteins, DNA, RNA, lipids, and metabolites from the parental MSCs, suggesting their therapeutic potential in cell-to-cell communication [23]. Proteomic analysis of ApoEVs derived from different sources of MSCs, such as human BMMSCs and human ADMSCs, revealed that 80% of the total proteins of the parental MSCs were present in MSC-ApoEVs, which were higher than viable MSC-EVs [78, 85]. They were enriched with many functional proteins related to cell behavior, cell metabolism, cellular transport, and regulation of various diseases [78, 85]. Furthermore, previous studies have shown significant similarity in total RNA profiles between ApoEVs and corresponding parental osteoclasts [86]. Combined with the fact that MSC-ApoEVs carry apoptosis-related markers, it predicted an immunomodulatory capacity similar to the parental ApoMSCs that can induce phagocytosis by phagocytes.

In conclusion, MSC-ApoEVs have the potential to be a more direct therapeutic agent than ApoMSCs, which is more in line with the current trend of advocating cell-free therapy. More and more studies are now demonstrating the therapeutic potential of MSC-ApoEVs, as shown in Table 3.

Table 3 Typical therapeutic applications of MSC-ApoEVs from different sources

Direct therapeutic applications

In mice with dorsal wound injury, BMMSCs transplanted directly at the wound site promoted skin wound healing but also underwent apoptosis within a short period of time [67]. Direct treatment of wounds with BMMSC-ApoEVs dissolved in the hydrogel Pluronic F (PF)-127 resulted in a reduction of the scar area, a more complete skin structure, the production of newly formed epithelium and hair follicles, and even a reduction of the inflammatory cell infiltration was observed [67]. Type 2 diabetes is often associated with poor wound healing, which can be difficult to treat. However, research has shown that applying hydrogel containing hUCMSC-ApoEVs to wounds in diabetic mice effectively accelerated wound healing [64]. Interestingly, ADMSC-ApoEVs of different diameter sizes showed different efficacy, and it was found that small diameter ADMSC-ApoEVs (less than 200 nm) were significantly more efficacious than large diameter ADMSC-ApoEVs (greater than 500 nm) [65]. Meanwhile, the age of the ADMSCs donor showed differences in the treatment of skin injuries. Although both young and old ADMSC-ApoEVs induced skin healing and reduced scar formation, the younger ADMSC-ApoEVs were more effective [87]. In addition to direct treatment at the wound site, it was found that intravenous injection of BMMSC-ApoEVs in wound-injured mice resulted in the accumulation of BMMSC-ApoEVs in the liver, skin, spleen, and lungs, with peak levels of BMMSC-ApoEVs occurring on the third day and significantly decreasing on the seventh-day post-infusion [68]. The study focused on the regression of BMMSC-ApoEVs at the skin injury and found that fluorescently labeled BMMSC-ApoEVs showed accumulation at the skin injury in mice within 12 h but disappeared on day 7 [68]. It may be related to the metabolism of BMMSC-ApoEVs by the epidermis and hair follicles of the skin as shown by structure illumination microscopy analysis, and the results were further confirmed in the symbiotic model [68]. The accumulation of BMMSC-ApoEVs at the site of skin damage effectively activated skin and hair follicle MSCs, thus effectively promoting skin healing and hair growth [68, 88].

MSC-ApoEVs have been shown to be effective in the treatment of bone damage. Zhu et al. found that whole-body fluorescence imaging of intravenously injected BMMSC-ApoEVs showed that BMMSC-ApoEVs were predominantly enriched in the liver, whereas the enrichment of BMMSC-ApoEVs in the femur gradually increased over time [75]. Injections of BMMSC-ApoEVs significantly ameliorated bone loss induced by primary (aged mice) and secondary (ovariectomy (OVX) mice) osteoporotic mice, with increased bone mass and improved bone microarchitecture in the mice [75]. Even in situ treatment with scaffolds in combination with BMMSC-ApoEVs was effective in promoting new bone generation in rats with cranial bone defects [75]. Interestingly, BMMSC-ApoEVs at different apoptotic stages differed in promoting osteogenesis. It has been claimed that BMMSC-ApoEVs from mid-apoptotic stages had the strongest regenerative effect relative to early and late apoptotic stages, which may be related to the differences in the substances contained in BMMSC-ApoEVs produced at different apoptotic stages [89]. Many studies have shown that a hypoxic environment can enhance the biological functions of MSCs [90]. On this basis, Ding et al. utilized a modified gelatin matrix/3D printed extracellular mesenchyme scaffold complex as a carrier to inject ADMSC-ApoEVs induced to apoptosis after hypoxic incubation directly into the damaged joint cavity exhibited better cartilage repair and regeneration than ADMSC-ApoEVs generated under normoxic conditions [79]. It may be related to the fact that ADMSC-ApoEVs generated under hypoxia were superior to those generated under normoxic conditions in terms of production, microRNA (miRNA), and functional proteins [79]. This engineering of 3D printed cells in combination with hypoxic preconditioning treatment protocol offers greater advantages and possibilities for the treatment of MSC-ApoEVs.

Of course, MSC-ApoEVs can also be an excellent treatment for dental injury diseases. In periodontitis mice, gingival sulcus injection of BMMSC-ApoEVs significantly ameliorated their alveolar bone destruction [38]. In a pulp removal model, treatment with stem cells from human exfoliated deciduous teeth (SHED)-ApoEVs promoted pulp revascularization and tissue regeneration [81]. In addition, MSC-ApoEVs showed great potential in the treatment of diseases such as hemophilia A [78], diabetes mellitus [91], myocardial infarction [34], autoimmune disorders [92, 93], sepsis [74], noisy hearing loss [85], cancer [69, 77], uterine adhesions [94], and aging [72].

Drug delivery therapy

In addition to direct therapeutic effects, MSC-ApoEVs can be used as packaged drug carriers for therapy. The apoptotic property of MSC-ApoEVs induces efferocytosis, which can be utilized to enable nanomedicine-loaded monocytes/macrophages to naturally target the damaged site for therapy [10]. It has been demonstrated that systemically infused MSC-ApoEVs can be home to hepatic macrophages [95]. Xiang et al. exploited this homing property and combined the antibacterial function of copper-doped carbon dots (Cu-CDs) to effectively enhance the ability of hepatic macrophage endocytosis of bacteria in septic mice by loading Cu-CD onto the entire surface of BMMSC-ApoEVs via electrostatic action [95]. In this case, small GTPase Rab5, which was enriched in BMMSC-ApoEVs, played an important role in the internalization of bacteria by hepatic macrophages and promoted the maturation of early phagolysosomes [95].

In a study by Yang et al., it was found that high concentrations of alpha-mangostin (α-M), a drug with anti-inflammatory, antioxidant, neuroprotective, and cardiovascular protective efficacy, were able to induce apoptosis in hUCMSCs, producing hUCMSC-ApoEVs enriched with α-M (α-M/ApoEVs) [96]. Surface modification of α-M/ApoEVs by co-administration of matrix metalloproteinase-activated cell-penetrating peptide could target brain damage and be phagocytosed by brain macrophage microglia [96]. More effective treatment of ischaemic stroke compared to untreated hUCMSC-ApoEVs [96].

Bortezomib (BTZ) is often used as a therapeutic agent for multiple myeloma, but side effects such as peripheral neurotoxicity, nephrotoxicity, and leukopenia are often associated with its treatment [97,98,99]. Cao et al. found that co-culturing Apo-BMMSCs with bortezomib, the naturally occurring apoptotic encapsulation system of Apo-BMMSCs can encapsulate bortezomib in BMMSC-ApoEVs [77]. The drug loading rate was higher than conventional sonication loading [77]. Bortezomib-loaded BMMSC-ApoEVs not only reduced the toxicity of BTZ but also produced better anti-multiple myeloma effects compared to natural BMMSC-ApoEVs and BTZ administered alone [77]. This may be a result of the combined effect of the small Rab GTPase Rab7 (a key regulator of vesicle maturation and secretion) and BTZ in BMMSC-ApoEVs [77].

The potential of MSC-ApoEVs in the treatment of bone damage has been demonstrated [71, 79]. Gui et al. affixed a bone-targeting peptide (Asp-Ser-Ser)6 ((DSS)6) to the surface of BMMSC-ApoEVs, and co-loaded ubiquitin ligase RING finger protein (RNF) 146 onto BMMSCs by adenoviral transduction to obtain functional BMMSC-ApoEVs [100]. This effectively improved the bone-targeting ability of BMMSC-ApoEVs and induced the osteogenic capacity of endogenous MSCs [100].

Therapeutic mechanisms of ApoMSCs

ApoMSCs and efferocytosis

The efficient clearance of apoptotic cells by phagocytes is also known as “efferocytosis” [101]. In most tissues, efferocytosis is carried out by either highly phagocytic tissue-resident professional phagocytes (e.g., macrophages and dendritic cells) or less phagocytic non-professional phagocytes (which can be neighboring cells such as epithelial cells and fibroblasts) [101]. Phagocytes have positive anti-inflammatory and immunosuppressive effects on apoptotic cells after efferocytosis [101]. ApoMSCs can also interact with phagocytes to exert anti-inflammatory mechanisms. See Fig. 2 for details.

Fig. 2

Therapeutic mechanisms of ApoMSCs

MSCs transplanted in vivo undergo apoptosis for various reasons such as inflammatory factor, hypoxia, microorganisms, cytotoxic immune cells, and nutritional deficiencies. Mechanisms of therapeutic effects of ApoMSCs. (1) ApoMSCs and efferocytosis: ApoMSCs induce macrophages to exert immunomodulatory effects by binding to the macrophage surface receptor MerTK via surface PS, in which ApoMSCs attract more CCR2 chemokines to recruit more macrophages. Macrophages regulated by ApoMSCs promote their own M2 polarisation and inhibit M1 polarisation, inhibit neutrophil and T-cell activation. (2) Release of soluble factors by ApoMSCs: MSCs secrete a variety of soluble factors after apoptosis, e.g. TSG-6, PGE2, PAI-1, M-CSF, IL-33, paracrine factors inhibit tissue inflammation, promote macrophage M2 polarisation and inhibit T-cell activation. (3) Release of MSC-ApoEVs: See Fig. 3 for details.

In GVDH mice, apoptosis BMMSCs (Apo-BMMSCs) induced efferocytosis by recipient macrophages and produced indoleamine 2,3-dioxygenase (IDO) ameliorating the infiltration of immunotoxic cells [32]. IDO is an immunosuppressive molecule that promotes the conversion of tryptophan to kynurenine and locally reduces the concentration of the essential amino acid tryptophan, which has been shown to reduce T-cell proliferation, thereby inhibiting T-cell-mediated immune responses [32, 102].

In allergic asthmatic mice, efferocytosis of Apo-BMMSCs induced alterations in the metabolic and inflammatory pathways of alveolar macrophages, i.e., a shift in the M2 phenotype and secretion of IL-10, which acted as an immunosuppressor to reduce the number of pulmonary eosinophils and decrease the severity of the disease [29]. Macrophages are characterized by remarkable plasticity and versatility, capable of switching from one phenotype to another, i.e., M1 phenotype to M2 phenotype [103]. M1 macrophages often promote the production of reactive oxygen species (ROS) and nitrogen intermediates, and the production of inflammatory factors such as IL-1 beta(β), TNF-α, and IL-6, which contribute to pro-inflammatory, bactericidal, and anti-tumour effects [103]. And M2 macrophages can produce IL-10, transforming growth factor-beta (TGF-β), vascular endothelial growth factor (VEGF), etc., which is conducive to the regulation of immunity, inhibition of inflammation, and promotion of angiogenesis [103]. He et al. found that Apo-BMMSCs ameliorated acute liver injury by binding to the macrophage surface Mer tyrosine kinase (MerTK) receptor via cell-surface PS and undergoing an M2 phenotypic shift by recruiting Ly-6C^high monocyte-derived macrophages partly via chemokine receptor C–C motif chemokine receptor 2 (CCR2) as well as producing IL-10 [33]. Interestingly, direct administration of PS liposomes mimicking Apo-BMMSCs exerted similar protective effects [33].

Release of soluble factors by ApoMSCs

In addition to efferocytosis, ApoMSCs can also directly secrete immunosuppressive factors to exert therapeutic effects. In an in vitro cellular assay, human ADMSCs were exposed to anti-Fas activating antibody and recombinant granzyme B to induce apoptosis and their transcriptional analysis revealed that, among a series of up-regulated genomes present in Apo-ADMSCs, immunomodulatory genes were the most prominently up-regulated, which mainly contained: IL-6, leukemia inhibitory factor, tumor necrosis factor alpha-induced protein 6, and cyclooxygenase 2 encoding gene prostaglandin-endoperoxide synthase 2, followed by the apoptotic genome [37]. Specifically, the authors found that Apo-ADMSCs were induced to produce a series of immunosuppressive molecules after caspase activation under the regulation of the NF (nuclear factor)-kappa(κ)B pathway, among which PGE2 was essential for Apo-ADMSCs to directly inhibit T cells activation [37]. This was similarly demonstrated in mice with acute colitis, where the addition of induced Caspase 9 MSCs with a suicidal apoptosis gene could inhibit T-cell activation by secreting PGE2 to promote weight regain, improved survival, and decreased inflammation, such as a decrease in serum levels of TNF-α and IL-6, in the diseased mice [19]. In rabbits with proliferative scarring, direct injection of H₂O₂-induced Apo-BMMSCs produced more anti-inflammatory factors, such as TSG-6, through the caspase 3 activation pathway, which effectively down-regulated pro-inflammatory/pro-fibrotic mediators, such as monocyte chemotactic protein-1, macrophage inflammatory protein-1 beta, IL-6, and TGF-β1, in the tissues, and ameliorated the formation of scarring [35].

Apoptotic cell-induced alterations in macrophage polarization of the M2 phenotype have been demonstrated in various types of apoptotic cells, but correlates secreted by ApoMSCs gave them a further predominant advantage [104,105,106]. In ischemic muscle injury mice, after induction of apoptosis by three different cell types such as BMMSCs, murine fibroblast cell line NIH 3T3 and murine macrophage cell line RAW 264.7, apoptotic supernatants of BMMSCs and murine fibroblast cell line NIH 3T3 significantly down-regulated TNF-α produced by macrophages activation, whereas apoptotic supernatants of murine macrophage cell line RAW 264.7 required 8–16 times more cells to produce the same efficacy as the other two [107]. This may be related to the different expression content of functional proteins in the cells [107]. In the apoptotic supernatant of BMMSCs, various proteins associated with the polarization of M2 macrophage anti-inflammatory phenotype, such as plasminogen activator inhibitor 1 (PAI-1), macrophage-colony stimulating factor (M-CSF), and IL-33, were highly expressed [107]. Although there are fewer comparative studies in this area, they still illustrate the excellent therapeutic potential of ApoMSCs.

Release of MSC-ApoEVs

Immunomodulatory capacity of MSC-ApoEVs

Recent studies have shown that MSC-ApoEVs, as metabolic secretions of ApoMSCs, played a similar role in inducing macrophage phagocytosis to exert positive anti-inflammatory and immunosuppressive effects. It has been found that BMMSC-ApoEVs expressed ‘eat’ me signaling: calreticulin on their surface, which attracted phagocytosis by hepatic macrophages [91]. This effectively restored hepatic macrophage homeostasis and allowed them to exert anti-inflammatory effects [91]. It played an important role in high-fat-induced type 2 diabetic mice, where the infusion of BMMSC-ApoEVs accumulated in the liver effectively improved glucose tolerance, attenuated insulin resistance, and ameliorated hepatic steatosis, among others [91]. The anti-inflammatory effect of MSC-ApoEVs was also demonstrated in acute endometrial injury mice. In situ treatment of hUCMSC-ApoEVs combined with hyaluronic acid hydrogel enhanced macrophage recruitment and phagocytosis and increased the number of M2 macrophages, resulting in high levels of IL-10 production [94]. It effectively increased the endometrial thickness and the number of endometrial glands, reduced the fibrotic area, improved the endometrial tolerance and ultimately promoted the restoration of fertility [94]. See Fig. 3 for details.

Fig. 3

Therapeutic mechanisms of MSC-ApoEVs

(1) Immunomodulatory effects: MSC-ApoEVs can regulate the activation of macrophages, neutrophils, and T cells. (2) Rescue damaged MSCs: functional proteins and miRNAs carried by MSC-ApoEVs can promote self-renewal and differentiation of endogenously damaged MSCs. (3) Promote apoptosis of diseased cells: induction of apoptosis of diseased cells by Fas/FasL. (4) Promote angiogenesis: ECs undergo proliferation, migration, and autophagy through functional proteins or miRNA and PD1/PDL1 from MSC-ApoEVs and thus angiogenesis is promoted.

In vitro, BMMSC-ApoEVs obtained after TGF-β1/IFN-γ-licensed BMMSCs-induced apoptosis effectively suppressed lipopolysaccharide (LPS)-stimulated inflammatory responses in THP-1 [76]. This may be related to the fact that BMMSC-ApoEVs were partially phagocytosed by pro-inflammatory macrophages through the phosphatidylinositol 3-kinase pathway, effectively inhibiting the secretion of pro-inflammatory factors such as TNF-α and IL-1β [76]. In a similar study, it was found that BMMSC-ApoEVs may inhibit inflammation in pro-inflammatory macrophages via the AMPK/SIRT1/NF-κB pathway, as demonstrated in pro-inflammatory bone marrow-derived macrophages induced by porphyromonas gingivalis derived LPS [108].

More and more studies have shown that MSC-ApoEVs are indispensable for regulating macrophage anti-inflammatory polarisation in terms of the microRNAs they carry. Li et al. found that miRNA (miR)-21–5p carried in ADMSCs-ApoEVs induced M2 polarisation in macrophages and enhanced skin wound healing by targeting the kruppel like factor 6 (KLF6) gene [109]. Similarly, recently Mao et al. found that miR-20a-5p carried in ADMSCs-ApoEVs effectively ameliorated diabetic rat-induced skin injury by promoting macrophage M2 polarization by regulating the JAK-STAT1 signaling pathway [80].

The immunomodulatory mechanism of MSC-ApoEVs, in addition to its significant role in regulating the macrophage M1/M2 phenotypic transition, still played an important potential in regulating other pro-inflammatory behaviors of macrophages such as the occurrence of pyroptosis and anaerobic glycolysis. During skin wound repair in type 2 diabetes mice, a large number of macrophages located in the injured area underwent pyroptosis due to oxidative stress, resulting in intense inflammation [61]. In situ treatment with hUCMSC-ApoEVs can effectively improve wound healing by decreasing macrophage NOD-, LRR-, and pyrin domain-containing protein 3 (NLRP3) inflammasome production, which is upstream of pyroptosis [61]. Jiang et al. found that hUCMSC-ApoEVs-expressed programmed cell death 1 ligand 1 (PDL1) caused macrophages to undergo a metabolic switch from glycolysis to oxidative phosphorylation (OXPHOS) through the Erk pathway downstream of the PDL1- programmed cell death protein 1 (PD1) axis, thereby shifting from pro-inflammatory to anti-inflammatory state [85]. The application of hUCMSC-ApoEVs effectively ameliorated the inflammation of lung tissues in acutely lung-injured mice [85].

MSC-ApoEVs played an important role in regulating T cell activation. In vitro, BMMSC-ApoEV derived from TGF-β1/ IFN-γ-licensed MSCs inhibited T cell proliferation and reduced activated CD69⁺ T cells [76]. However, the induction of Tregs was promoted and the expression of immunomodulatory CD73⁺ T cells was maintained [76]. Efferocytosis induced by apoptotic cells in macrophages often produces anti-inflammatory mediators that inhibit the immune response and which play an important role in suppressing T-cells activation [76]. However, it has recently been shown that the expression of PS on the surface of MSC-ApoEVs can directly inhibit T-cells activation [92]. Previous studies have shown that PS specific to the surface of apoptotic cells often binds directly or indirectly to phagocytic receptors such as the TIM (T-cell immunoglobulin domain and mucin domain) family and TAM (TYRO3、AXL and MERTK) family to exert anti-inflammatory effects [101]. Wang et al. found that after in vivo infusion of BMMSC-ApoEVs in systemic lupus erythematosus mice, in the absence of macrophages, BMMSC-ApoEVs inhibited T cell activation through direct contact between surface PS and TIM-3 receptors on the surface of activated CD4⁺ T cells [92]. This effectively contributed to the reduction of IFN-γ and IL-2 levels from serum and the improvement of the corresponding clinical symptoms in lupus and arthritis [92].

MSC-ApoEVs had great potential in regulating neutrophil activation [74]. It has been claimed that MSC-ApoEVs can alter the form of neutrophil death, thereby inhibiting the activity [74]. In septic mice, neutrophils underwent a unique form of cell death, namely neutrophil extracellular traps (NETs) consisting of citrullinated histone H3 and myeloperoxidase (NETosis) [74]. Intravenous administration of BMMSC-ApoEVs switched the mode of neutrophil death from NETosis to apoptosis via the Fas/FasL pathway [74]. This effectively inhibited the infiltration of neutrophils and monocytes into distal organs, thereby improving the survival rate of septic mice [74] Interestingly, BMMSC-ApoEVs and NETosis can be attracted to each other by electrostatic interactions [74]. Specifically, positively charged NETs, which were mainly concentrated in the bone marrow and spleen, could be effectively regulated by attracting negatively charged BMMSC-ApoEVs to contact them [74].

MSC-ApoEVs rescued damaged endogenous MSCs

Most adult organs contain regenerative stem cells, which are essential for maintaining tissue homeostasis and promoting repair after injury [110]. Apoptosis-induced compensatory proliferation-induced regenerative repair plays a crucial role in the proliferation of neighboring cells, especially neighboring stem cells [111]. Previous studies have shown that the specific mechanism of apoptosis-induced compensatory proliferation (AICP) was mainly dependent on caspase activation and driven by a paracrine mechanism: secretion of mitogens [112]. Recently, it has been shown that mitogenic signals generated by apoptosis-induced compensatory proliferation may be utilized by neighboring MSCs in the form of ApoEVs, thus exerting a reparative effect. Brock et al. found that in zebrafish epidermis, MSC-ApoEVs containing Wnt8a were generated in a caspase-activated manner after the apoptosis of basal stem cells [113]. Phagocytosis of MSC-ApoEVs was performed by neighboring MSCs to activate the Wnt signaling pathway and induce the proliferation of MSCs in the epidermis [113]. Recently, it was found that zebrafish epidermal basal stem cells induced apoptosis can generate MSC-ApoEVs carrying macrophage inhibitory factor (MIF) on their surface [114]. It could promote the proliferation of neighboring epidermal basal MSCs by regulating the MIF/CD74/ERK signaling axis [114].ROS is usually regarded as deleterious to the cells, however, ROS plays an important role in AICP and tissue regeneration [115, 116]. On this basis, it has been found that low concentrations of BMMSC-ApoEVs could promote the production of beneficial ROS by recipient cells and promote BMMSCs proliferation, migration, and osteogenic differentiation through activation of JNK signaling [89]. In an in vitro cellular study, it was found that dental pulp stem cells (DPSC)-ApoEVs may promote DPSCs osteogenic differentiation by activating the ERK1/2 signaling pathway [83].

As MSC-ApoEVs, they themselves carried MSCs regulatory-related factors to play important roles in the self-renewal and differentiation ability of damaged endogenous MSCs, such as microRNA. Among them, miR-1324, miR-328-3p, and miR-145a-5p carried by MSC-ApoEVs have a major role in the promotion of bone injury diseases [71, 75, 93]. In Zhu et al.‘s study, the amelioration of bone damage may be related to the release of miR-1324 from BMMSC-ApoEVs [75]. Specifically, miR-1324 inhibited the expression of the target gene Sorting Nexin 14 (SNX14), which activated the SMAD 1/5 pathway in endogenous BMMSCs [75]. This effectively promoted BMMSCs osteogenesis and inhibited osteoclast differentiation [75]. In another study, Liu et al. found that BMMSC-ApoEVs could promote self-renewal and differentiation by activating the Wnt/β- catenin pathway in endogenous MSCs [93]. This may be related to the presence of RNF146 and miR-328-3p in BMMSC-ApoEVs to downregulate Axin1 [93]. Activation of the wnt/β-catenin pathway also upregulated FasL expression in endogenous MSCs, contributing to osteoclast apoptosis [93]. In a recent study, Zhang et al. found that BMMSC-ApoEVs can downregulate TGF-β/Smad 2/3 signaling in OVX MSCs via miR-145a-5p [71]. miR-145a-5p also activated Wnt/β-catenin-linker protein signaling through downregulation of TGF-β receptor 2 (TGF-βR2) and Dickkopf 1 (DKK1) [71]. These promoted the self-renewal and differentiation of endogenous MSCs and effectively ameliorated osteoporosis in mice [71]. However, it has also been reported that the specific Wnt/β-catenin signaling was not as effective in improving osteoporosis in mice. It has also been reported that hsa-miR-4485-3p, which is specifically present in BMMSC/ADMSC-ApoEVs, inhibits osteogenic differentiation and promotes adipogenic differentiation of MSCs by regulating the AKT signaling pathway [117]. Taking advantage of this property, customized MSC-ApoEVs knocking down hsa-miR-4485-3p demonstrated better bone regeneration therapeutic effects [117].

In addition to miRNAs in MSC-ApoEVs, several functional protein molecules played important roles in regulating endogenous stem cell injury. In a model of aging-associated bone loss, Lei et al. found that young BMMSC-ApoEVs carried high levels of Ras-related GTP binding protein 7 (Rab 7) that acted on receptor-aged MSCs to restore autolysosomal lysosome formation [118]. It activated cellular autophagy, whereas promoted regeneration and osteogenic and lipogenic differentiation in MSCs [118]. This was confirmed in the Li et al. study, BMMSC-ApoEVs could carry a large number of Ras proteins to activate the Ras/Raf1/Mek/Erk pathway [119]. It promoted osteogenesis and bone formation of MSCs in vivo and in vitro, and effectively improved osteoporosis in mice [119]. MSC-ApoEVs have been shown to promote skin MSCs regeneration in skin injury, but some studies have shown that embryonic stem cells-derived ApoEVs (ESC-ApoEVs) showed better efficacy than hUCMSC-ApoEVs [88]. It may be related to the fact that ESC-ApoEVs expressed more functional nuclear proteins from their parental MSCs, such as SOX2, which can activate the Hippo signaling pathway of skin MSCs [88].

Recently, it has also been shown that ESC-ApoEVs were significantly better than hUCMSC-ApoEVs in the treatment of senile osteoporosis [120]. This may be related to the fact that ESC-ApoEVs carried high levels of functional proteins associated with the maintenance of mitochondrial homeostasis [120]. Specifically, protein mass spectrometry analysis showed that ESC-ApoEVs highly expressed TCOF1, a nucleolar factor that regulates nucleolar transcription, compared with hUCMSC-ApoEVs [120]. In vitro, TCOF1 in ESC-ApoEVs rescued mitochondrial dysfunction in senescent BMMSCs by upregulating the mitochondrial protein FLVCR1 [120]. However, this was reversed by the knockdown of TCOF1 in ESC-ApoEVs using siRNA, and FLVCR1 levels were also decreased in senescent BMMSCs [120].

Angiogenesis promoted by MSC-ApoEVs

Angiogenesis is an essential step in tissue repair due to the ability of blood vessels to transport nutrients and oxygen to support cells at the wound site [121]. Recently, more and more studies have shown that MSC-ApoEVs play an influential role in the therapy of angiogenic diseases. Angiogenesis often contributes to skin wound healing, and in skin-injured rats, ADMSC-ApoEVs effectively induced endothelial cells(ECs) to proliferate, migrate, and undergo angiogenesis, thereby reducing the area of wound injury and accelerating wound healing [87]. However, compared with normal MSC-ApoEVs oxygen stress-treated MSC-ApoEVs (Oxi-MSC-ApoEVs) were more effective in promoting angiogenesis at the wound site, which may be related to the fact that Oxi-MSC-ApoEVs contained a large amount of miR-210-3p [84]. miR-210-3p promoted ECs migration and activation by activating AKT signaling [84].

The dental pulp is a vessel-rich tissue in stable root canals, and in pulp-in-situ injured Beagle dogs, ApoEVs derived from DPSC -ApoEVs could be phagocytosed by ECs to transfer mitochondrial Tu translation elongation factor (TUFM) [81]. TUFM could specifically activate the endogenous ECs’ autophagy pathway via the autophagy pathway of the transcription factor EB (TFEB), which promotes angiogenesis [81]. Ultimately, accelerated revascularization promoted efficient regeneration of the dental pulp, which was rendered ineffective by the addition of apoptosis inhibitors [81]. This mechanism was demonstrated in mice with ischemic heart disease, i.e. myocardial infarction [34]. Direct intramyocardial injection of BMMSC-ApoEVs induced autophagy in ECs after being phagocytosed by recipient ECs, thereby enhancing angiogenesis and recovery of cardiac function [34]. This may be related to the fact that donor BMMSC-ApoEVs promoted TFEB-mediated lysosomal function, which upregulated autophagy in recipient ECs to promote angiogenesis [34]. It also involved AKT-mediated activation of the VEGF signaling pathway and NO generation [34].

The PD1/PDL1 axis has been reported to regulate glycolytic metabolism during inflammation, however, recent studies have found that it also has great potential in regulating ECs glycolytic metabolism [82, 122]. In ischemic retinopathy mice, ECs at the retinopathy site highly expressed PDL1 due to hypoxia, whereas ApoEVs derived from SHED-ApoEVs superficially expressed high levels of PD1 [82]. SHED-apoVs could inhibit ECs’ anaerobic glycolysis via the PD1-PDL1 pathway to promote angiogenesis [82].

Apoptosis promoted by MSC-ApoEVs in pathogenic cells

As BMMSC/ADMSC-ApoEVs, the apoptosis-typical receptor Fas carried on their surface could activate platelet coagulation by binding to the platelet surface FasL, thereby rescuing hemophilia A mice with factor VIII knockout [78]. However, the surface of MSC-ApoEVs, in addition to carrying the apoptosis-associated factor Fas, was highly expressed in FasL, which may be related to the inheritance of properties from parental MSCs, which played an important role in the induction of apoptosis [123, 124]. In addition to the previously mentioned alteration of neutrophil death pattern by BMMSC-ApoEVs through Fas/FasL, in multiple myeloma mice, BMMSC-ApoEVs utilized their FasL to directly activate the calcium channel-induced Fas membrane transport pathway in multiple myeloma cells to cause apoptosis of multiple myeloma cells and extended the survival rate of multiple myeloma mice [69, 74] It has even been found that BMMSC-ApoEVs could promote apoptosis of myoblasts and then cause them to release myoblast-derived ApoEVs, which could release creatine through activated Pannexin 1 channels and eventually increased the fusion of myoblasts to promote muscle regeneration [70].

Maintenance of homeostasis of other tissues by MSC-ApoEVs

Surely, some of the functional molecules carried by MSC-ApoEVs have also shown remarkable efficacy in the treatment of other diseases such as aging, and in the maintenance of organ homeostasis such as the liver, ovaries, ears, and teeth.

It has been demonstrated that BMMSC/hUCMSC/ESC-ApoEVs can inherit nuclear proteins from their parental MSCs, which was also confirmed by Huang et al. They found that MSC-ApoEVs could inherit a variety of DNA repair-related proteins from parental MSCs, including PARP1, Rad51, Rad52, XRCC4-like factor, Proliferating Cell Nuclear Antigen, damage-specific DNA binding protein 1, X-ray Repair Cross Complementing Protein 1, and Recombinant Prominin 1, which were higher than viable MSC-EVs [72]. Among them, PARP1 played a key role in MSC-ApoEVs-mediated DNA repair, which could effectively ameliorate premature senescence and DNA damage induced by lethal dose irradiation in apoptosis-deficient mice [72].

Defective apoptosis resulted in hepatic injury in mice, however, infusion of BMMSC/hUCMSC-ApoEVs effectively protected hepatocyte health and maintained hepatic homeostasis, which may be related to the presence of membrane proteins on the surface of MSC-ApoEVs that can interact with hepatocytes [125]. Specifically, MSC-ApoEVs were endocytosed by hepatocytes via Gal(Galactose) /GalNAc (N-acetylgalactosamine) -ASGPR (asialoglycoprotein receptor) and formed ApoEVs-Golgi complex (AGC) by binding to the hepatocyte Golgi surface receptor syntaxin 16(STX16) via membrane surface vesicle-associated membrane protein 3(VAMP3) [125]. The AGC promoted microtubule acetylation by regulating α-microtubule protein N-acetyltransferase 1, thereby facilitating the cytoplasmic division of hepatocytes [125]. This effectively promoted liver regeneration and effectively saved acute liver failure [125].

Naturally, MSC-ApoEVs also played an instrumental role in the maintenance of ovarian homeostasis [73]. It has been shown that apoptosis defects caused the development of polycystic changes and oocyte damage in ovary mice, while BMMSC-ApoEVs effectively ameliorated follicular dysfunction [73]. This was attributed to the fact that BMMSC-ApoEVs inherited the Wnt membrane receptor inhibitor protein RNF43 from parental MSCs [73]. Specifically, BMMSC-ApoEVs could deliver the functional protein RNF43 to the theca cells (TCs) through a characteristic vesicle-cell membrane integration mechanism. And then interacted with the Wnt receptor frizzled-4 (FZD4), which could effectively down-regulated aberrantly activated Wnt/β-catenin signaling in the TCs [73]. The restored Wnt/β-catenin signaling down-regulated reactively activated DKK1 levels in the oocyte regions, thereby effectively rescuing the down-regulated Wnt/β-catenin signaling in the oocytes [73]. This played an important role in promoting ovarian folliculogenesis [73]. The functional pathway of the NPPC/cGMP/PDE3A/cAMP cascade in follicles was also restored [73]. In addition to apoptosis-deficient mice, the amelioration of ovarian dysfunction by apoptotic follicles was demonstrated in dehydroepiandrosterone-induced polycystic ovarian syndrome mice and in mice with aging-associated ovarian dysfunction [73].

Furthermore, MSC-ApoEVs have tremendous potential for the treatment of hearing impairment [85]. Proteomic analysis of hUCMSC-ApoEVs revealed that they were rich in various oxidoreductases and signal transduction and activators of transcription 3 (STAT3), a transcription factor that responds to various cytokines or growth factors [85]. It was found that hUCMSC-ApoEVs could reduce oxidative stress and ameliorate noise-induced hearing loss in mice by promoting nuclear translocation of activated forkhead box o3 (FOXO3a) and upregulation of superoxide dismutase 2 via STAT3 in damaged cells [85].

In periodontitis mice, gingival sulcus injection of hypoxia-induced BMMSC-ApoEVs better induced targeted phagocytosis of BMMSC-ApoEVs by osteoclasts since both BMMSC-ApoEVs and osteoblasts express dendritic cell-specific transmembrane protein (DC-STAMP) (a membrane fusion regulator) [38]. Meanwhile, miR-223-3p (a key regulator of osteoclast differentiation and bone resorption) in BMMSC-ApoEVs effectively ameliorated alveolar bone destruction by targeting Itgb1 to attenuate osteoclast formation and reduced bone resorption [38].

Challenges and prospects

Most of the existing reports suggested that apoptosis of transplanted MSCs in vivo may be induced by the microenvironment in which they are transplanted, such as hypoxia, inflammatory factors, nutritional deficiencies, cytotoxic immune cells, microorganisms, etc [36]. However, based on the currently available data, it is still difficult to know the clear mechanism, which limits the exploration of both in vitro expansion methods and in vivo therapeutic mechanisms for apoptosis-dependent MSCs therapy. In other studies, MSCs pretreated with inflammatory factors or hypoxia could also exert favorable therapeutic effects. In vitro, IFN-γ-stimulated hUCMSCs could effectively inhibit NK cells activation and resist NK-mediated cytotoxicity through the production of IDO [126]. In hindlimb ischemic mice, hypoxia preconditioned MSCs activated the HIF-1α/GRP78/Akt signaling axis to repair mouse injury [127]. Surprisingly, some studies have shown that pretreatment with the inflammatory factor IFN-γ upregulated apoptosis-related genes such as Fas and interferon regulatory factor 1 [39, 128]. Hypoxia has also been used in many studies as a method to induce apoptosis [38, 79]. This evidence suggested that the same mechanism of downstream driving effect may exist between the significant therapeutic effects of inflammatory factors and hypoxic preconditioning and the efficacy produced by cell-induced apoptosis [18]. However, it is also necessary to note that in many cases ApoMSCs therapy relies on tissue-resident macrophages, but the state of macrophages such as chemotaxis and numbers are difficult to predict in different diseases, which is a factor to be considered in the use of apoptotic MSCs therapy.

In this review, we also explored MSC-ApoEVs, the administration of which may be a favorable cell-free alternative treatment due to the inheritance of important molecules and properties from the parental MSCs. Although viable MSC-EVs have been found to be therapeutically effective, available technologies are still limited in the number of viable MSC-EVs that can be produced in vitro [129]. In contrast, MSC-ApoEVs are readily available and have a high production rate. It has been found that BMMSC-ApoEVs were produced ten times more than viable BMMSC-EVs [92]. Furthermore, BMMSCs produced 245 times more BMMSC-ApoEVs in vitro compared to the same number of T cells [10, 92]. Yang et al. found that the same number of ADMSCs produced six times and four times more ADMSC-ApoEVs particles and proteins, respectively than viable ADMSC-EVs within a 12-hour period [65]. However, MSC-ApoEVs therapy has been an emerging therapeutic modality in the last few years, and we have yet to thoroughly investigate the complexity and heterogeneity of its contents and the stability of its amplification in vitro. We found that some studies have used gels such as PF-127, gelatin methacrylate, and modified gelatin such as the synthesis of norbornene-modified gelatin, and the synthesis of tetrazine-modified gelatin to slow down the degradation rate of MSC-ApoEVs when mixed with them [64, 65, 79]. Even the tissue engineering approach of constructing 3D-printed extracellular matrix environments in vitro further improved the efficacy of MSC-ApoEVs [79]. The adoption of such engineered MSC-ApoEVs offers more possibilities for its developmental applications.

Questions remain regarding the therapeutic options for ApoMSCs and MSC-ApoEVs. In vitro, it has been found that it was ApoMSC, but not MSC-ApoEVs, that were able to produce PGE2 to inhibit T-cell activation, furthermore, in vivo MSCs were shown to undergo apoptosis by means of cell-to-cell contact [37]. It has been shown that the anti-inflammatory effects of apoptotic cells during efferocytosis are closely related to the metabolites they produce at different stages of efferocytosis. Even though receptors that induce efferocytosis are present on the surfaces of both ApoMSCs and MSC-ApoEVs, some differences in the therapeutic mechanisms of these two types of cells may remain, which will require a contextual analysis of the utilization of ApoMSCs/ MSC-ApoEVs in future therapies.

Conclusions

There is growing evidence of the potential therapeutic and even drug delivery capacity of ApoMSCs and MSC-ApoEVs. In this review, we concluded the therapeutic mechanism of ApoMSCs, found that they played significant roles in immunomodulation, such as inducing macrophage polarization into anti-inflammatory phenotype and producing immunosuppressive secretome, and tissue regeneration, such as through secretion of MSC-ApoEVs carrying regenerative molecules, as well as in the promotion of angiogenesis, inducing apoptosis of diseased cells, and maintaining tissue homeostasis. This evidence provides new ideas for investigators of MSCs-based therapies. Among them, we found that MSC-ApoEVs are broadly promising for therapeutic applications due to their inheritance of important properties from their parental MSCs, as well as their easy availability and high yield. However, we have not yet clarified the specific mechanism of apoptosis of transplanted MSCs in vivo, which limits the exploration of both in vitro expansion methods and in vivo therapeutic mechanisms that rely on ApoMSCs for therapeutic purposes. Moreover, the therapeutic application of ApoMSCs and MSC-ApoEVs is still in the early stages, and further research and exploration of the differences between the two and their respective advantages are still necessary.