Apoptotic cells derived micro/nano-sized extracellular vesicles in tissue regeneration


 Extracellular vesicles (EVs), products released by cells in multiple biological activities, are currently widely accepted as functional particles and intercellular communicators. From the orthodox perspective, EVs derived from apoptotic cells (apoEVs) are responsible for cell debris clearance, while recent studies have demonstrated that apoEVs participate in tissue regeneration. However, the underlying mechanisms and particular functions in tissue regeneration promotion of apoEVs remain ambiguous. Some molecules, such as caspases, active during apoptosis also function in tissue regeneration triggered by apoptosis,. ApoEVs are generated in the process of apoptosis, carrying cell contents to manifest biological effects, and possessing biomarkers to target phagocytes. The regenerative effect of apoEVs might be due to their abilities to facilitate cell proliferation and regulate inflammation. Such regenerative effect has been observed in various tissues, including skin, bone, cardiovascular system, and kidney. Engineered apoEVs are produced to amplify the biological benefits of apoEVs, rendering them optional for drug delivery. Meanwhile, challenges exist in thorough mechanistic exploration and standardization of production. In this review, we discussed the link between apoptosis and regeneration, current comprehension of the origination and investigation strategies of apoEVs, and mechanisms in tissue regeneration by apoEVs and their applications. Challenges and prospects are also discussed here.

Abstract: Extracellular vesicles (EVs), products released by cells in multiple biological activities, are currently widely accepted as functional particles and intercellular communicators. From the orthodox perspective, EVs derived from apoptotic cells (apoEVs) are responsible for cell debris clearance, while recent studies have demonstrated that apoEVs participate in tissue regeneration. However, the underlying mechanisms and particular functions in tissue regeneration promotion of apoEVs remain ambiguous. Some molecules, such as caspases, active during apoptosis also function in tissue regeneration triggered by apoptosis,. ApoEVs are generated in the process of apoptosis, carrying cell contents to manifest biological effects, and possessing biomarkers to target phagocytes. The regenerative effect of apoEVs might be due to their abilities to facilitate cell proliferation and regulate inflammation. Such regenerative effect has been observed in various tissues, including skin, bone, cardiovascular system, and kidney. Engineered apoEVs are produced to amplify the biological benefits of apoEVs, rendering them optional for drug delivery. Meanwhile, challenges exist in thorough mechanistic exploration and standardization of production. In this review, we discussed the link between apoptosis and regeneration, current comprehension of the origination and investigation strategies of apoEVs, and mechanisms in tissue regeneration by apoEVs and their applications. Challenges and prospects are also discussed here. by cells in multiple biological activities. EVs serve as information communicators of cells [1,2], which were formerly considered as cell debris. According to biogenesis and size, EVs can be roughly classified as exosomes (endosome origin, 30-100 nm in diameter), microvesicles (plasma origin, 50-1,000 nm), and apoptotic bodies (apoptotic cell origin, 1-5 µm) [3,4]. EVs of various sizes from apoptotic cells have been studied over the last several decades, and they are not just confined to apoptotic bodies. As a result, apoptotic vesicles are used to refer to all types of EVs derived from apoptotic cells (apoEVs) [5][6][7]. ApoEVs, as the products and information disseminators, seem to inherit the therapeutic role of apoptosis in tissue regeneration, changing death to another beginning of life [8,9]. Different from necrosis, apoptosis is unlikely to cause a traumatic defect or aberrant tissue shrinkage. As a result, apoptosis is well-known to play a role in tissue homeostasis and life cycle regulation, although there are still many unsolved mysteries in the field [10]. In recent years, tissue renewal, such as epithelium, skin, and liver renewal, has been widely described as one of the major biological effects mediated by apoptosis [11][12][13]. ApoEVs are products in apoptotic process. Traditionally, the uptake of apoEVs by phagocytes was thought to be the end of apoptosis, which biochemically signals the death of cells, but new studies suggest that their fates include transfer and reuse [14][15][16][17]. Apoptosis-like regeneration properties have also been demonstrated in the application of apoEVs, which come in a variety of sizes and are produced through different apoptotic pathways. Recent studies indicate that apoEVs might promote regeneration in tissue maintenance, including skin, bone, cardiovascular system, and kidney, by functioning bioactive effects through recruitment of targeted cells and delivery of biomolecules [18][19][20][21][22]. To better utilize and amplify the therapeutic properties of apoEVs, engineered methods implicating targeting, releasing, and retaining have been investigated [23,24]. Figure 1 demonstrates the production, isolation, functions, and applications of apoEVs. Culture medium of apoptotic cells was collected and centrifuged to obtain apoEVs. By recruiting targeting cells and delivering biomolecules, apoEVs promote cell proliferation and regulate inflammation, thus being applied in tissue regeneration. ApoEVs are modified to function in broader fields as well. Recent studies have discussed the functions and applications of apoEVs in immune regulation, inflammation mediation, mesenchymal stem cell (MSC) transplantation, and tumor progression [3,[25][26][27][28]. Nonetheless, few documents have comprehensively discussed the regenerative effect of apoEVs and analyzed the underlying mechanisms. In this review, the mechanisms of apoEVs in tissue regeneration are addressed, along with their applications in both medicine and engineering. We hope this work will provide an indicative view of the mechanisms in tissue regeneration by apoEVs, and inspiration for novel application of apoEVs in the future.

Apoptosis and regeneration 2.1 Apoptosis with caspases
Since the term "apoptosis" was proposed for the first time by John Foxton Ross Kerr and his colleagues at the University of Queensland [29], it has been proven to participate in various physiological processes, including development, aging, immune homeostasis, and tissue integrity maintenance [3,16,30]. Apoptosis, the most prominent pattern of programmed cell death, is characterized by several morphological changes, involving cell shrinkage, cytoplasmic and nuclear condensation, membrane blebbing, and the subsequent formation of apoptotic bodies.
Apoptosis, a conserved physiological process, can typically be triggered by intrinsic pathway and extrinsic pathway. A general summary is discussed here to clarify the processes of apoptosis. Caspases are fundamental participants in both pathways of apoptosis. The intrinsic pathway of apoptosis is closely associated with mitochondrial regulation, in which stimuli (for example, DNA damage) generate biochemical changes inside cells. Stimuli elicit the dissipation of mitochondrial transmembrane potential, thus forming mitochondrial permeability transition pore on the outer membrane and leaking proapoptotic mitochondrial contents. Cytochrome c is then released into the cytoplasm to induce oligomerization of apoptotic peptide activating factor 1, which recruits procaspase-9 and forms apoptosome, the activator of caspase-9. Subsequently, activated caspase-9 in turn cleaves response caspases, activating caspase cascade reactions [16,31,32]. The extrinsic pathway of apoptosis initiates activation of death receptors. Upon contact with extracellular ligands, mainly TNF and Fas, the death receptors on the cell membrane transfer death signals towards the cytoplasm. The binding of ligands and receptors initiates recruitment of Fas-associated death domain protein, causing autocleavage of procaspase-8. Activated caspase-8 then cleaves procaspase-3 to yield activated caspase-3 responsible for executing downstream degradation process [16,30,32].

Regeneration with apoptosis
The mechanism of apoptosis is much more sophisticated than the brief sketch above, concerning a series of interactions and cascade signaling pathways, in which caspases have a well-defined role. However, increasing evidence implies the function of caspases in tissue repair and regeneration [33]. Dead cells are replaced by compensatory proliferated cells to maintain tissue homeostasis. Apoptotic cells deliver mitogenic signals to trigger the compensatory proliferation of neighboring cells, which is so-called "apoptosis-induced compensatory proliferation (AiP)" [11,34].
Caspases, the initiator and effector of apoptosis, are the major activators of AiP [11]. In Drosophila, cell death was blocked by effector caspase inhibitor p53 even with high expression of initiator caspase Dronc, rendering Dronc functioning apart from apoptosis [35,36]. Actually, Dronc activated AiP via JNK signaling pathway to express Wingless and Decapentaplegic mitogens, triggering compensatory proliferation of surrounding cells [34,37]. In Dronc-induced AiP, extracellular reactive oxygen species (eROS) produced by NADPH oxidase Duox activated macrophage-like hemocytes, which in turn amplified the JNK signal by TNF ortholog Eiger, thus promoting epithelial cell growth [11,12]. Effector caspases DrICE and Dcp-1 induced AiP in eye tissue via Hedgehog signaling [36,38]. In some other species, such as Xenopus tadpole and Hydra, ROS are pivotal compartments for AiP. Profuse ROS at the wound activated Wnt/β-catenin signaling and downstream fgf20, initiating active cell proliferation in tail regeneration of Xenopus tadpole [39]. In Hydra, ROS produced immediately after injury elicited MAPK signaling [11] and Wnt3 [40] in head regeneration.
In mammals, regeneration due to AiP is also linked to caspases. In mice, the lack of caspase-3 or caspase-7 leads to deficiency in skin and liver regeneration [13]. Cleaved caspase-3 and caspase-7 initiate calcium-independent phospholipase A2 to upregulate the expression of arachidonic acid, which further mediates the production of prostaglandin E2, facilitating tissue repair and stem cell proliferation [13]. The activation of caspase-3 enlarged YAP-dependent organs, while inhibition of caspase-3 attenuated cell proliferation and diminished sebaceous gland [41]. Some mediate evidence also proves the function of caspases in regeneration. The self-renewal property of embryonic stem cells was remarkably inhibited when lacking caspase-3 [42]. Deficient caspase-3 in mice resulted in impaired differentiation of bone marrow stem cells and decreased bone density by activation of TGF-β/Smad2 signaling pathway [43,44], which was ameliorated by secretion from apoptotic MSCs [44]. In conclusion, participants in apoptosis, such as caspases and ROS, function in tissue renewal, coupling cell death and regeneration.
Here we demonstrated several representative morphologies of apoEVs ( Figure 2).

Biogenesis and characteristics
Nowadays, biogenesis of apoEVs is described as three sequential well-coordinated steps with corresponding morphological changes: plasma membrane blebbing, thin membrane protrusions formation, and fragmentation [8,25]. In membrane protrusions formation, different protrusions might be generated, including microtubule spike, apoptopodia, and beaded apoptopodia. However, for some specific cells, not every step is necessary for the biogenesis of apoEVs ( Figure 3). Apoptotic membrane blebbing, the earliest morphological change in apoptosis, is driven by cytoskeleton collapse and increased hydrostatic pressure [32]. Caspases and various protein kinases regulate this mechanism, preventing apoptosis when they are inhibited [31]. Rho-associated protein kinase 1, a target of active caspase-3, induces phosphorylation of myosin light chain (MLC), causing actomyosin contraction [32]. MLC kinase facilitates the assignment of degraded chromatin into blebs [48]. Serine/threonine LIM domain kinase 1 inhibits the actinbinding protein cofilin to promote actin polymerization [49]. This membrane blebbing is known as fundamental for membrane protrusions, but some cells undergo different membrane deformation in this process, indicating diverse mechanisms in addition to blebbing solely in apoEVs formation ( Figure 3). Apoptotic A431 epithelial cells formed a rigid membrane protrusion, the microtubule spike, in the process of apoptosis without blebbing [50]. Human Jurkat T cells generated apoptopodia following blebbing, which is regulated by the caspase-activated  [68]. (c) Live differential inference contrast microscopy image showing generation of apoEVs from THP-1 cells undergo UV-induced apoptosis. Scale bar, 5 or 10 µm as indicated [5].
four-pass transmembrane pannexin 1 channel, a negative regulator of apoptosis [51]. Recently, a novel "beads-on-astring" type of membrane protrusion in apoptosis was described in apoptotic human THP-1 cells and monocytes. This beaded apoptopodia forms abundant apoEVs (approximately 10-20) with diameters of 1-4 µm due to the formation of protrusions on the "string" with a length 8 times that of the dying cell, rendering it much more efficient than the previous 2 patterns of membrane protrusion [52]. After blebs and/or membrane protrusions formation, fragmentation is responsible for the drop of single apoEV ( Figure 3). While the specific mechanism or molecules mediating this process are currently unclear now, shear force in circulation or intercellular physical force of cell disassembly might be involved [8].
By the end of apoptosis, "find-me" and "eat-me" signals are indispensable for attraction and clearance of phagocytes [53], avoiding cell contents leakage and secondary necrosis [32]. The "find-me" signal is chemo-attractive gradient generated by released soluble molecules in apoptosis, such as ATP, UTP [54], and CX3CL1/fractalkine [55], which was also found to be associated with apoEVs [25,56]. Some receptor proteins in the membrane of phagocytes, the G protein-coupled receptor G2A for example, induce the recognition of "find-me" signals, even though the precise pathway of this process and quantified attractive gradient remain to be elucidated [57]. The most universally studied "eat-me" signal is the exposure of phospholipid phosphatidylserine (PtdSer) [55], anchoring on the surface of apoEVs. In the membrane of healthy cells, PtdSer is strictly confined to the inner side of the bilayer. While undergoing apoptosis, PtdSer flip from the inner side of the lipid bilayer of plasma membrane to the outer layer during blebbing, which is induced by caspase-activated scramblases, a set of phospholipid translocases [57]. The appearance of PtdSer on the outer layer in both apoptotic cells and apoEVs transmits the dying signal, inducing recognition and clearance of phagocytes. In addition, ICAM-3, a confirmed mediator of apoptotic cell clearance, was also present on the surface of apoEVs to attract macrophages [58]. Notably, apoEVs seem : Morphological changes in apoptotic cells in the biogenesis of apoEVs. It could be described in three steps with corresponding morphological changes: plasma membrane blebbing, membrane protrusions formation, and cell fragmentation. In step 2, different protrusions might be generated, including microtubule spike, apoptopodia, and beaded apoptopodia. Dotted arrows mean that apoEVs can be generated by some specific cells without step 1 or step 2.
to recruit phagocytes selectively, which might be due to content diversity caused by different parental cells and specific physiological and pathological processes. Adipocyte-derived apoEVs recruited macrophages rather than neutrophils [59]. However, neutrophils infiltration was observed in mice, in which larger apoEVs (1-3 µm in diameter) attracted neutrophils, whereas smaller ones (<1 µm) could not. Inflammatory chemokine IL-1 appeared in larger apoEVs but not in smaller ones to induce sterile neutrophil inflammation, which might originate from the nuclear IL-1α precursor that translocated into the nucleus during apoptosis [60]. The equilibrium between apoEVs and phagocytes is indispensable for homeostasis. Secondary necrosis might occur when massive apoptosis and subsequent excessive apoEVs containing inflammatory factors overwhelm the clearance ability of phagocytes or when the function or quantity of phagocytes is impaired in diseases.
In most studies, apoEVs were described as bioactive carriers transporting functional molecules [3,16,31], including DNA debris [61], RNA [21,[62][63][64], and proteins [65,66], yet cargo differences among each group of apoEVs exist widely. The proteomes of healthy and cirrhotic human biliary epithelial cells showed a huge divergence in the shotgun proteomics of apoEVs [65]. Furthermore, exosome-like apoEVs carried more active 20 S proteasome core than that in apoptotic bodies (apoBDs), which is responsible for their immunogenic activity [45]. More interestingly, DNA and RNA could not be packed into apoEVs of HL-60 cells simultaneously. Over 90% of apoEVs containing RNA were free of DNA, and vice versa [62]. Additionally, nuclear material was absent in beaded apoptopodia [52], suggesting that some bioactive molecules may enter apoEVs with tropism. Different groups of apoEVs possess diverse properties and functions due to their distinct biogenesis processes, and unequal contents that result from different parental cells and their previous physiological and pathological changes.

Production and identification
Technically, it is necessary for researchers to study apoEVs in a relatively uniform condition to acquire convincing information, whereas cells in a general culture condition are less likely to undergo apoptosis under control. Thus, some stimuli of apoptosis in specific physiological or pathological conditions have been applied in studies to produce apoEVs. Here we reviewed production and isolation methods in recent years studying apoEVs, along with their documented times, donor cells, and experimental models ( Table 1). The most universally used stimulus of apoEVs is staurosporine [18,19,23,64,67,68], a potent inhibitor of protein kinase C, which induces apoptosis by activating caspase-3 and suppressing AKT/MAPK pathway [69,70]. Serum starvation might affect cell cycle and promote apoptosis [71]. Exposure to gamma rays or ultraviolet radiation injured the stability of nucleic acids [67,72,73], which appeared when applying some chemotherapeutic drugs, such as cisplatin [67]. For some specific cells, the corresponding stimulus was applied. Alendronate was used to produce apoEVs from osteoclasts, due to endoplasmic reticulum stress in osteoclast precursors caused by alendronate [74].
For isolation of apoEVs, apoMVs, a subtype of apoEVs resembling microvesicles in size [4], were obtained in a centrifuge at 16,000-50,000g. ApoBDs and apoExos were isolated in a centrifuge at 1,000-5,000g and 100,000-200,000g, respectively. The size and morphology were observed and determined by dynamic light scattering analysis, scanning electron microscope, transmission electron microscope, or confocal microscope. Fluorescence-activated cell sorting (FACS) was used to quantify and purify apoEVs. Several molecules were detected to identify apoEVs. The most widely accepted biomarker is Annexin V, a calcium-binding protein of the annexin superfamily, which can interact with the "eatme" signal PtdSer. Thrombospondin and complement protein C3b, the products of membrane changes during apoptosis, are also well-accepted biomarkers. Cleaved caspase-3 and C1q were also applied in some cases. Nuclear granularity and hypochromicity were detected by propidium iodide, but their quantities were not equal and were even absent in some apoEVs [5,15,[18][19][20][21]23,44,63,64,67,68,72,73,75].
However, it is noteworthy that the current classification is based on size, while ignoring their biogenesis in cells and the contents of each group. Besides, apoEVs derived from different cells might differ in size. And apoEVs would collapse over time in vitro [5]. Actually, current studies are less likely to identify and analyze the subtype properties of apoEVs utilized in experiments, rendering disturbances from different groups being inevitable. It is pivotal to further analyze the contents and properties and uncover the key proportion of each group of apoEVs, which contributes to standardizing protocols for isolation and identification, thus obtaining rigorous scientific results with diminished confounding.

Functions of apoEVs in tissue regeneration
ApoEVs regulate cell function by transferring various abundant cargo, which involves many bioactive factors participating in tissue repair and regeneration. ApoEVs derived from different parental cells might differ in content and affect cell activities via diverse pathways. In tissue regeneration, beneficial effects of apoEVs have been studied and discussed. Figure 4 demonstrates the main mechanisms of apoEVs in tissue regeneration, including promoting the proliferation of neighboring cells by compensatory proliferation signals and remote cells by transferring therapeutic molecules, and suppressing inflammation by regulating macrophage polarization (Figure 4).

Promoting cell proliferation
Tissue integrity is maintained in the homeostasis of dynamic cell death and compensatory proliferation. Apoptosis has been implied to activate compensatory proliferation signaling, hence dying cells mediate compensatory proliferation in adjacent cells [11,13,30,34,76,77], which is also called AiP [11,34]. ApoEVs, the product of dying cells, seem to inherit such a compensatory proliferation function. ApoEVs produced by different apoptotic stimuli were reported to trigger neighboring cell proliferation in both transformed and primary cells upon contact via compensatory proliferation signaling [78], which is related to CrkI, a type of protein responsible for cell proliferation and cytokinesis [79,80] ( Figure 4). Epithelial stem cell-released apoEVs also triggered cell proliferation of standby healthy stem cells in a compensatory proliferation way, maintaining regeneration of the epithelium [75]. In addition to the near approach, apoEVs were also implicated in promoting cell proliferation in remote sites. ApoEVs from macrophages delivering miR-221/222 induced lung epithelial cell growth by suppressing the expression of cyclin-dependent kinase inhibitor 1B, a tumor suppressor controlling cell cycle at G1 [63] (Figure 4). Circulating apoEVs from MSCs transferring ubiquitin ligase RNF146 and miR-328-3p recovered the self-renewal and differentiation properties of distant impaired MSCs [44] ( Figure 4). Intriguingly but predictably, Wnt/β-catenin signaling pathway participated closely in cell growth and renewal promoted by apoEVs [44,75,78]. Focusing on miRNAs and proteins modulating Wnt signaling pathway in apoEVs could be a possible way to elucidate the mechanisms by which apoEVs promote tissue regeneration.

Regulating inflammation
Inappropriate inflammation impedes repair and regeneration progress in multiple diseases [81][82][83][84]. ApoEVs are capable of attenuating inflammation by modulating cell activities of inflammatory cells, thereby smoothing the tissue regeneration process. Suppression of inflammatory cytokines, including MCP-1 and IL-6, caused by upregulated TSG-6 was observed after treatment with apoptotic MSCs [85] (Figure 4). TSG-6 has been reported to reduce proinflammatory signals from macrophages [86] and inhibit the activation of macrophages [87]. Furthermore, apoEVs have been implicated in modulating macrophage polarization both in vivo and in vitro.
In mouse models, apoEVs from MSCs promoted polarization of anti-inflammatory M2 macrophages, whereby they increased the anti-inflammatory cytokines IL-10 and TGF-β and enhanced the migration and proliferation of fibroblasts [18] (Figure 4). Biomimic apoEVs were taken up more effectively by activated M1 macrophages and led to more intensive M2 macrophage polarization, ameliorating inflammation by upregulating anti-inflammatory cytokines [88]. In addition, apoEVs derived from activated T cells were likely to possess trophic property to inflammatory site similar to that of their parental cells, meanwhile promoting M2 macrophage polarization [23]. In short, apoEVs help to recover a healthy microenvironment of repair and regeneration by ameliorating excessive local inflammation. Besides above all, many other intricate mechanisms contribute to the regeneration function of apoEVs. ApoEVs from MSCs promoted the proliferation of endothelial cells [15] and endothelial progenitor cells [89], thus inducing angiogenesis. ApoEVs derived from different parental cells manifested specific functions in regeneration resembling their parental cells, which was observed in preosteoclasts and mature osteoclasts [19,20], cardiomyocytes [90,91], and proximal tubular cells [67]. From this perspective, stem cells might be a promising origin of apoEVs due to their intrinsic regenerative properties. However, it is difficult to draw a definite conclusion from these scattered studies, since there is little convincing result indicating the direct correlation between the regenerative property of apoEVs and specific mechanisms.

Therapeutic effects of apoEVs in regeneration
In recent years, researchers have probed and utilized the regeneration ability of apoEVs in numerous tissue regeneration processes. Here we summarize and discuss the promotion of skin, bone, heart, and kidney repair and regeneration by apoEVs ( Figure 5).

ApoEVs in skin regeneration
ApoEVs produced by MSCs facilitated wound healing as MSCs [18,85,92]. ApoEVs induced M2 macrophage polarization, the anti-inflammatory subtype of macrophages, and sequentially promoted the viability of fibroblasts, accelerating cutaneous wound repair [18] ( Figure 5). Furthermore, apoEV mimics carry miR-21/curcumin boosted M2 polarization [23], indicating that the membrane of apoEVs and these therapeutic factors cooperated in skin regeneration. Taking advantage of the macrophage targeting property, apoEVs delivering antibiotic vancomycin relieved intracellular methicillin-resistant Staphylococcus aureus infection of macrophages [24]. Remaining dormant in macrophages, intracellular Staphylococcus aureus escape from immune system and antibiotics, causing persistent cutaneous infection, which results in impaired wound healing and skin regeneration [93]. Moreover, after the apoptotic induction of TNF-α, apoEVs enveloping more IL-1RA were released by MSCs, promoting gingival and skin regeneration through Fas/Fap-1/Cav-1 axis [92]. Reduced hypertrophic scar formation in the process of skin regeneration was also observed after treatment with TSG-6 secreted by apoptotic human MSCs [85]. TSG-6, a protective factor in the inflammatory response, has been reported to be one of the dominant mediators in the regenerative ability of MSCs [94].

ApoEVs in bone regeneration
ApoEVs produced by osteoclasts also manifested similar biological effects as parental cells [19,20]. Preosteoclastderived apoEVs promoted angiogenesis, while mature osteoclast-apoEVs promoted osteogenesis, similar to perspective parental cells both in bioinformatics analysis and in vivo [19]. Mechanistically, mature osteoclast-derived apoEVs induced the differentiation of osteoblasts that mediate osteogenesis by activating receptor activator of NF-κB ligand reverse signaling [20,95], coupling bone resorption and formation ( Figure 5). Notably, MSCs-released apoEVs rescued the osteopenia phenotype in both apoptosis-deficient mice and ovariectomized mice by restoring stem cell properties of impaired MSCs ( Figure 5). The uptake in circulation of ubiquitin ligase RNF146 and miR-328-3p carried by apoEVs activated Wnt/β-catenin pathway [44], a conservative signaling pathway crucial in development and tissue regeneration [96,97]. These results demonstrate the indispensable role of apoEVs in regulating bone remodeling and maintaining bone homeostasis.

ApoEVs in cardiovascular system
ApoEVs derived from different donor cells provide varied biological benefits for cardiovascular system protection and regeneration. Endothelial apoEVs deliver miR-126 ameliorated atherosclerosis and plaque in vivo by reducing macrophage infiltration and apoptotic cells. Endothelial apoEVs upregulated the expression of chemokine CXCL12 and mobilized Sca-1 + progenitor cells, thus inhibiting CXCR4-dependent atheroprogression [21]. Similarly, an apoEV bio-mimic liposome enhanced the stability of atherosclerotic plaques, which means there is a lower possibility of plaque rupture and subsequent thrombosis by improving macrophage-induced inflammation, to protect cardiovascular system [88] ( Figure 5). In another aspect, MSC-derived apoEVs enhanced angiogenesis and cardiac function recovery in myocardial infarction mice, which was due to recipient endothelial cells activating autophagy [15]. Moreover, apoEVs derived from cardiomyocyte regulated the differentiation of cardiac stem cells to cardiomyocyte precursors and improved the contractility of myocardium [90,91], while apoEVs released by fibroblast produced no such effect [90] ( Figure 5).

ApoEVs in renal regeneration
In renal regeneration, a novel type of apoEVs containing CrkI has been described, inducing mitosis and proliferation of nearly 100% analyzed recipient cells including renal parietal epithelial cells [78]. Podocyte-derived apoEVs containing CrkI might promote regeneration of injured tubular epithelial cells [22] ( Figure 5). Surprisingly, apoEVs released by the first apoptosis in proximal tubular cells (PTCs) caused by cisplatin mediated the secondary apoptosis of recipient PTCs, while apoEVs produced by this secondary apoptosis stimulated tubular cell proliferation [67]. These controversial effects of apoEVs derived from different apoptosis processes might be explained by propagation of toxic effects of cisplatin, or apoptotic stimulus of Fas ligand that probably expressed on apoEVs [98]. Nevertheless, specific comparison experiments concerning apoEVs induced by cisplatin and other stimuli, and the interaction and homeostasis between apoptosis and regeneration of PTCs need to be addressed further. In other words, apoEVs from diverse origins may have distinct therapeutic effects or play different roles in different biological processes.

Engineered apoEVs
EVs have been reported to possess tropism targeting a specific tissue or cell [99][100][101][102][103], known as the "homing effect" [101][102][103]. Moreover, EVs show natural advantages as drug delivery systems over synthetic nanoparticles, with the abilities to cross biological barriers efficiently and interact with the plasma membrane via ligand/receptor responses, while possessing low immunogenicity and toxicity [100,[104][105][106][107]. These intrinsic properties of apoEVs qualify them as optimal bioactive materials to be modified for tissue regeneration, causing more precise tissue targeting, enhanced inflammation modulation, and longer drug retention. Thus, apart from the direct utilization of natural apoEVs, modified apoEVs or biomimetic apoEVs are also a concern. The unique identifying signal of apoEVs for phagocytes has piqued the interest of researchers in harnessing their targeting ability. ApoEV bio-mimic liposomes were constructed by connecting PtdSer onto the surface of liposomes [88], inspired by the "eat-me" signal from exposed PtdSer of apoEVs. The PtdSer-modified liposomes effectively escaped from clearance by reticuloendothelial system owing to PEGylation [88,108]. Compared with liposomes, the modified liposomes that targeted M1 macrophages were engulfed more effectively as expected. As a sign of suppressed inflammation, inhibited M1 polarization and motivated M2 polarization were observed from increased anti-inflammatory cytokines such as IL-4 and IL-10 and decreased proinflammatory cytokines such as IL-1β, IL-6, and TNF-α [88] ( Table 2). In addition, apoEVs derived from some kinds of cells might inherit the particular targeting property of their parental cells [109], such as leukocytes. Taking this into consideration, apoEVs released by T cell were manipulated to target inflammatory sites and then macrophages [23]. In this research, natural membranes of T cell derived apoEVs without cargos were obtained, and then encapsulated with mesoporous silica nanoparticles (MSNs) that were preloaded with anti-inflammatory molecules. These chimeric apoEVs showed CD44 and Mac-1 protein expressions comparable to T cells, which were responsible for recognition between inflamed endothelium and leukocytes [110,111], thus displaying inflammation targeting capacity both in vitro and in colitis models ( Table 2). Promoted macrophage-specific phagocytose and M2 polarization were observed as well [23]. Similarly, cancer cell derived apoEVs were reconstructed for targeting the same type of cancer cells, as well as targeting macrophages [24]. Combined with their targeting property dictated by their membrane markers and cargo content [105], the natural lipid membrane potentialized apoEVs as drug delivery systems. Co-bathing in sonication is one of the common strategies to load therapeutic molecules into apoEVs [23] ( Table 2). To boost the efficiency and stability of drug loading, therapeutic molecules could be preloaded into MSNs, an outstanding nanoplatform as a drug carrier [112,113], by electrostatic interactions [23] or undergoing a freeze-thaw process bathing with apoEVs [24]. By a mechanically extruding technique, apoEVs were processed into a relatively uniform size of 80-150 nm and obtained an enhanced encapsulation efficiency of vancomycin [24] ( Table 2). Otherwise, genetic manipulation can also be concerned. Transient transfection into donor cells of therapeutic miRNA could also be transferred into apoEVs successfully [68], providing another cargo loading approach ( Table 2).
Another drug delivery strategy is loading apoEVs into materials or scaffolds with biocompatibility, not concerning the targeting property or membrane superiority, but aiming to achieve longer drug retention and controlled release. When embedded into PF-127 hydrogel, apoEVs showed the fastest skin regeneration [18] ( Table 2). Although a concentration time curve is lacking in this research, a compelling retention time and promoted regeneration ability of other EVs combined with biomaterials have been provided [114][115][116]. In general, the combination of apoEVs and biomaterials is an attractive method with the potential for sustainable therapy.

Challenges, prospects, and conclusion
ApoEVs, the products, and meanwhile the information disseminators, seem to inherit the therapeutic role of apoptosis in tissue regeneration, changing death to another beginning of life. Different biogenesis pathways of apoEVs seem to be linked to cell types. To obtain information on the reason why different biogenesis patterns of apoEVs occur, investigating the functions of cells undergoing diverse apoptotic ways and respective apoEVs might be a possible direction. Moreover, morphological diversities in different biogenesis patterns of apoEVs might contribute to the clearance of apoptotic cells. For example, suppressed blebbing in Jurkat T cells caused inhibited uptake by macrophages in cell disassembly [117]. The interaction between microtubule spikes of A431 epithelial cells and phagocytes was reduced when formation of microtubule spikes was inhibited [50]. Although a conclusive explanation for such diverse patterns still needs to be clarified, diversity among the contents of apoEVs originating from different sources exists. For example, apoEVs undergoing "beads-on-string" apoptosis contain no nuclear material, unlike other groups of apoEVs [52]. In addition, research has seldom compared the underlying discrepancies in different production protocols or differences among subtypes of apoEVs. In limited literature that analyzed the content diversity of apoEVs, different attractive abilities for neutrophil were demonstrated in various groups of apoEVs. The group containing inflammatory chemokines caused secondary inflammation, which we will not anticipate to occur in tissue regeneration. This means that it is not that the more apoEVs there are, the better the regenerative effect, but the appropriate quantity and prompt processing of phagocytes, along with the optimal group of apoEVs, or the specific contents of apoEVs in another word, matter more. Figuring these puzzles out is fundamental to probe into the particular portion of apoEVs functioning directly or mediately as key therapeutics in different diseases, thus further formulating instructive protocols to study or manufacture in the future, aiming at specific pathological situations.
In conclusion, apoptosis, the silent closure of one cell, and the initiation of other cells, play pivotal roles in tissue integrity maintenance. ApoEVs may promote tissue regeneration, including tissue of skin, bone, cardiovascular system, and kidney, by activating targeted recipient cells or transferring biomolecules (DNA, RNA, and proteins) to facilitate proliferation and regulate inflammation, both locally and remotely. Nevertheless, the reason why apoEVs could be generated in different patterns in vivo, and the connection between different biogenesis patterns and the properties of respective apoEVs remain unclear. Controlling the distribution and release of biomolecules or drugs carried by apoEVs is also important for therapeutic usage, for which manipulation by engineered means and combination with biomaterials can be optional. There still is a long way to utilize the regenerative and therapeutic functions of apoEVs, many comprehensive studies need to be finished.