Bone Regeneration Improves with Mesenchymal Stem Cell Derived Extracellular Vesicles (EVs) Combined with Scaffolds: A Systematic Review

Simple Summary Extracellular vesicles (EVs) have been recently considered one of the main characters for liquid biopsy (biomarkers) and therapeutic application. Particularly, the therapeutic application of EVs is involved in the bone regeneration, thanks to the regulation of immune environments, enhancement of angiogenesis, differentiation of osteoblasts and osteoclasts, and promotion of bone mineralization. In the past 15 years, researchers have focused on the application of EVs derived from different types of mesenchymal stem cells (MSCs) in the field of bone regenerative medicine. This systematic review aims to analyze in vitro and in vivo studies that report the effects of EVs combined with scaffolds in bone regeneration. A methodical review of the literature was performed on PubMed and Embase from 2012 to 2020. Sixteen papers were analyzed; of these, one study was in vitro, eleven were in vivo, and four were both in vitro and in vivo studies. This analysis shows a growing interest in this upcoming field, with overall positive results. Promising in vitro results have been discussed in terms of bone regeneration and pro-angiogenetic processes. The positive in vitro findings were confirmed in vivo, with studies showing positive effects on several critical-size defects. However, some aspects remain to be elucidated, like the different effects induced by EVs and secretome, the most suitable cell source, the EV production protocol and concentration, and the clinical use that may benefit from this new biological approach. Abstract Scaffolds associated with mesenchymal stem cell (MSC) derivatives, such as extracellular vesicles (EVs), represent interesting carriers for bone regeneration. This systematic review aims to analyze in vitro and in vivo studies that report the effects of EVs combined with scaffolds in bone regeneration. A methodical review of the literature was performed from PubMed and Embase from 2012 to 2020. Sixteen papers were analyzed; of these, one study was in vitro, eleven were in vivo, and four were both in vitro and in vivo studies. This analysis shows a growing interest in this upcoming field, with overall positive results. In vitro results were demonstrated as both an effect on bone mineralization and proangiogenic ability. The interesting in vitro outcomes were confirmed in vivo. Particularly, these studies showed positive effects on bone regeneration and mineralization, activation of the pathway for bone regeneration, induction of vascularization, and modulation of inflammation. However, several aspects remain to be elucidated, such as the concentration of EVs to use in clinic for bone-related applications and the definition of the real advantages.


Introduction
In recent years, health problems involving the musculoskeletal system, mainly due to osteoporosis, tumors, and fractures [1], have been widely studied. Different therapeutic approaches are used to induce bone regeneration; however, it has been shown that the physiological processes do not occur in some conditions like critical bone defects, which still remain an unmet clinical need [2] both in orthopedic [3] and maxillofacial [4] surgery. The impossibility to regenerate critical-bone defects is mainly due to the low intrinsic ability to regenerate, as the quantity of bone needed to fill the defect is too large or the bone tissue is compromised by particular situations like osteonecrosis, tumors, or congenital abnormalities [5]. Currently, the gold standard method is the reconstruction with revascularized, bone-containing free flaps. These microvascular reconstructions provide optimal results thanks to the vitality of the bone tissue employed. On the other hand, they are technically demanding, requiring high expertise, a considerable time for harvesting and a non-negligible dose of handicraft skills that need several years to be developed [6]. Moreover, the donor site morbidity, although potentially minimized in expert hands, can be considered a further unavoidable drawback of this technique [6]. In this condition, an autologous bone grafting or a cells/scaffold construct implantation is necessary to fill or cover a defect depending on the size and quality of the adjacent soft tissues [7,8].
Bone tissue regeneration procedures have been introduced into clinical practice mainly based on the combination of bone tissue engineering materials with mesenchymal stem cells (MSCs) and/or growth factors [8,9].
MSCs represent the most promising cell population for clinical application in bone and joint diseases, and have been well demonstrated in preliminary clinical studies. MSCs can be derived from different sources like bone marrow and adipose tissue, but also from oral tissues (dental pulp, periodontal ligament, and gingival) [10][11][12]. Particularly, MSCs promote tissue repair thanks to their migration ability to reach injured tissues and by exerting immunomodulatory and trophic effects [13,14]. Moreover, MSCs can be expanded in adequate quantities for potential therapeutic applications, and show a self-renewal capacity and ability to differentiate in vitro into osteoblasts, adipocytes, and chondroblasts [15,16]. However, the efficacy of these approaches may be limited by regulatory issues [17,18].
In order to overcome these limitations, researchers (in the past 15 years) have focused on the application of extracellular vesicles (EVs) derived from different types of MSCs in the field of bone regenerative medicine [19][20][21].
EVs have been studied comprehensively in disease therapy and tissue regeneration as they are the primary paracrine executors in signaling communication between cells [22].
EVs derived from MSCs have been shown to exert a compatible regeneration potential compared with MSCs. Thus, EVs are even more appealing than stem cell transplantation as a potential alternative for tissue engineering due to several advantages, including good biosafety, stability, and efficient delivery [23]. EVs circulate in the human body and are present in most biological fluids [24,25].
EVs mainly differ in their dimensions/origin and are classified in exosome, microvesicles (MVs), and apoptotic bodies (ABs) (Figure 1a). Exosomes (<150 nm) are developed from the fusion of multivesicular bodies with the cytoplasmic membrane. MVs (<1mm) are formed through budding of the membrane and shuttle local cytosolic biomolecules. Larger ABs (1-5 µm) are released during the cell apoptotic process and contain cell debris, organelles, and nuclear particulates derived from karyorrhexis [26].
EVs have been proven to play a role in cellular signaling as immunomodulatory messengers [27]. EVs also contribute to improving bone regeneration by increasing angiogenesis [28]. Studies have revealed that EVs maintain the balance of bone metabolism by promoting the differentiation of osteoclasts, osteoblasts, and MSCs [29]. Additionally, EVs participate in bone mineralization, an essential process in bone regeneration [30]. EVs contain different types of biomolecules, such as microRNA (miRNA), short interfering RNA (siRNA), long non-coding RNA (lnRNA), proteins, and cytokines [31,32], and to facilitate their delivery were combined with different scaffolds (Figure 1b).
Among the scaffolds used in bone regeneration, hydrogels can be adapted to have properties closer to the natural extracellular matrix (ECM). In particular, Liu et al. [33] found that a biodegradable hydrogel encapsulating small exosomes can still produce the expected therapeutic effects when applying them directly or near to the treated area [33]. Multi responsive self-healing chitosan based-hydrogels have been extensively studied [34][35][36][37][38][39].

Data Source
The use of EVs in both in vitro and in vivo studies for bone regeneration has been systematically reviewed. This search was performed on PubMed and Embase databases from 2012 to 2020, using the following search terms: exosom* OR microvesicle* OR vesicle* OR ectosom* OR secretory* OR embedded-vesicles* OR released vesicle*) AND (bone* OR bone tissue*) AND (regeneration* OR tissue regeneration* OR inflammation* OR Tissue inflammation*) AND (hydrogel* OR biomaterial* OR scaffold*).

Study Selection Process
The screening process of the papers was conducted independently by two reviewers (F.R. and G.L.) by following the PRISMA guidelines. First, the resulting records were screened by title and abstract. After that, the selected manuscripts were assessed considering the in vitro and in vivo studies on the use of EVs and exosomes in bone regeneration. Articles written in other languages or not studying the effect of EVs or not exploiting their potential effect in the bone were excluded. The reference lists of the selected papers were also screened by the reviewers. The flowchart reported in Figure 2 illustrates the systematic review process.

Data Extraction and Synthesis
Relevant data from the selected studies were summarized and analyzed according to the aim of the present manuscript.
Particularly, the cell source of EVs, target cell types, production method, and study design were evaluated [47]. For the in vivo studies, the animal model and the method of bone regeneration were also considered together. In particular, in vitro effects were evaluated in terms of their effect on bone mineralization and their proangiogenic ability; the in vivo effects were evaluated in terms of bone regeneration and mineralization, activation of pathway for bone regeneration, induction of vascularization, and modulation of inflammation.

Results
One hundred and twenty papers from PubMed and sixty-seven from Embase were found according to the systematic strategy. Sixteen papers were analyzed after the removal of duplicates. Between them, four were in vitro and in vivo, and eleven were in vivo studies. The following paragraphs summarize all of these studies.
(ii.) Pizzicanella J et al. [49] investigated the ability of PEI complexed with hPDLSCs to induce the osteogenic differentiation of hPDLSCs grown on a collagen membrane. In fact, they demonstrated that this system based on collagen membrane plus hPDLSCs and PEI may help to induce bone regeneration. (iii.) Gandolfi et al. [50] showed the ability of exosomes secreted by hAD-MSCs combined with PLA-based scaffolds to trigger the osteogenic commitment of hAD-MSCs, improving their osteogenic properties. Particularly, they used two formulations of PLA+calcium silicates (CaSi)+dicalcium phosphate dihydrate (DCPD), namely: PLA-10CaSi-10DCPD and PLA-5CaSi-5DCPD. Exosomes, encapsulated on the surface of the scaffolds, the improve gene expression of major markers of osteogenesis such as collagen type I (COL1), osteopontin (OPN), osteonectin (ON), and osteocalcin (OCN). The experimental scaffolds enriched with exosomes, in particular PLA-10CaSi-10DCPD, increased the differentiation of MSCs from the osteogenic lineage. (iv.) Benton Swanson W et al. [47] provided strong evidence that osteogenic hDPSCsderived exosomes facilitate pro-mineralization cues to drive local stem/progenitor cells towards osteogenic lineage on PLLA in vitro.

Effect on Proangiogenic Ability
Pizzicanella J et al. [49] also investigated the ability of PEI complexed with hPDLSCs grown on a collagen membrane to induce the vascularization of bone defects, thanks to its capacity to increase the levels of vascular endothelial growth factor (VEGF) and vascular endothelial growth factor receptor 2 (VEGFR2), which that was shown to play an important role in osteogenesis and bone regeneration. Particularly, PEI-EVs, up-regulating the osteogenic genes and increasing the protein levels of BMP2/4, activated an osteogenic response.

Effect on Activation of Pathway for Bone Regeneration
(i.) Zhang et al. [56] provided evidence that the exosomes secreted by hiPS and scaffold based on tricalcium phosphate can effectively promote bone repair and regeneration in a rat model of calvarial bone defects through the activation of the phosphoinositide 3-kinases/protein kinase B (PI3K/Akt) signaling pathway on BMSCs [56]. (ii.) Cell-free bone regeneration was demonstrated by Zhai et al. [57], who revealed that scaffolds without cells can induce bone regeneration as efficiently as the hMSCseeded exosome-free scaffolds. Particularly, osteogenic exosomes can be identified from pre-differentiated stem cells and thus used to replace stem cells in tissue regeneration. In fact, exosomes contain upregulated osteogenic miRNAs and thus trigger PI3K/Akt and mitogen-activated protein kinase (MAPK) osteogenic differentiation pathways [57]. (iii.) The MAPK pathway was shown to also be activated in hADSCs by hDPSC-EVs as a cell-free biomaterial in a model of the mandibular defects in rat [58].

Effect on Inflammation and Cytokines
Another interesting work demonstrated that the association of chitosan hydrogel and dental pulp stem cell-derived exosomes can effectively treat periodontitis, accelerating the healing of alveolar bone and the periodontal epithelium in mice [55] and reducing inflammatory cytokines. The suppression of periodontal inflammation is mediated by macrophages converted from a pro-inflammatory phenotype to an anti-inflammatory phenotype in the periodontium of the mice [53].

Discussion
Direct cell-cell contact or the transfer of secreted molecules allows intercellular communication, which is an essential hallmark of multicellular organisms. In the last two decades, a third mechanism for communication between cells has emerged, and it involves the intercellular transfer of EVs. EVs are considered to be one of the main characters for liquid biopsy (biomarkers) and therapeutic application [38]. Particularly, the therapeutic application of EVs is involved in the bone regeneration, thanks to the regulation of immune environments, enhancement of angiogenesis, differentiation of osteoblasts and osteoclasts, and promotion of bone mineralization [54].
Various miRNAs and proteins are present in EVs derived from osteoblasts, osteoclasts, osteocytes, monocytes, macrophages, and dendritic cells, and are involved in enhancing or inhibiting the osteogenic activity, as reported by a recent review [63]. In particular, EVs play some role in the calcification of cartilage, bone, and dentin [29]. It has been shown that EV mediated-miRNA is a potential target of high mobility group AT-hook 2 (HMGA2) [64], glycogen synthase kinase-3β/β-catenin [65], or Wnt/β-catenin [66], and is well known to play a role in osteogenic differentiation. Moreover, EVs are also enriched in proteins like tissue non-specific alkaline phosphatase (TNAP), nucleotide pyrophosphatase phosphodiesterase (NPP1/PC-1), annexins (ANX; principally annexins II, V and VI) and phosphatidyl serine (PS) relative to the membranes from which they are derived, matrix metalloproteinases (MMPs), proteoglycan link proteins and actin, a variety of integrins, and PHOSPHO-1 [30], which are all involved in bone remodeling.
The main finding of this review is that EVs associated with different scaffold types could efficiently improve bone regeneration by enhancing angiogenesis through the activation of specific pathways.
Angiogenesis plays an important role in bone growth and regeneration [68][69][70], and is also promoted by EVs from both human [49,59,60,67] and rat [53,54,61] MSCs associated with different scaffold types. Interestingly, it has been demonstrated that angiogenesis and osteogenesis was also observed in critical-sized bone defect repair in osteoporotic rats using exosomes secreted from iPSC-MSCs [60], suggesting their positive action also when used in pathological conditions. The important role of HIF-1α for promoting osteogenesis and vascular regeneration has been shown [71,72]. Both Ying et al. [61] and Zhang et al. [62] evaluated, for the first time, the role of exosomes to carry mutant HIF-1a in repairing critical-sized calvarial defects, confirming its important role in the promotion of bone regeneration and neovascularization.
Another feature of EVs carried by different scaffold types is the up-regulation of different osteogenic genes, like RUNX2, OCN, OPN, TUFT1, TFIP11, and COL1A1 [49,51,53], as well as the expression of well-known bone inducing factors like BMP2/4 and TGFβ [49,51]. Vascularization, mineralization, and bone regeneration are strictly linked, and all of these steps are necessary for obtaining correct bone regeneration. It is well known that the induction of ECM followed by mineralization is a fundamental step during bone healing, and the establishment of osteocyte concomitant is associated with the peak expression of genes that are typical markers of mature osteoblasts. These include, but are not limited to, RUNX, bone sialoprotein, OCN, OPN, and BMP pathways, but also local growth factors such as bone TGF-β1/2 [69,70].
Some studies [56][57][58] have focused on the activation of the pathways involved in bone regeneration, such as PI3K/AKT and MAPK signaling pathways. These signaling pathways play important roles in the osteogenesis of the hMSCs and could help to confirm the osteogenic ability of the exosomes. The PI3K/Akt pathway may have a role in the exosome induced pro-osteogenic effects on MSCs, as this signaling cascade has been reported to play important roles in osteoblast differentiation and bone growth. Particularly, an interplay between the PI3K/AKT signaling cascade and BMP-2 gene transcription that regulates osteoblastogenesis has been demonstrated [73]. The MAPK signaling is a key player in bone development and skeletal homeostasis, particularly in osteoblast differentiation [74]. EVs have been found to increase the proliferation of osteoblasts and MSCs through the MAPK pathway [75]. To activate the PI3K/Akt and MAPK signaling pathways, osteogenic exosomes induce osteogenic differentiation by using their cargos, including upregulated osteogenic miRNAs (Hsa-miR-146a-5p, Hsa-miR-503-5p, Hsa-miR-483-3p, and Hsa-miR-129-5p) [45].
As demonstrated by Shen et al. [53], EVs reduce inflammatory cytokines. In fact, exosomes have emerged as potent stimulators of immune responses and as potential biomarkers and therapeutic agents for autoimmune disorders, even if their precise functions and potential in autoimmune diseases are not fully understood [27]. An important role in modulating the phenotype of macrophages has been recently reported using cell-derived exosomes containing miRNAs such as miR-223 [76] and miR-182 [77]. In fact, one of the important mechanisms of the immunomodulatory effects of stem cell exosomes is the exosome-mediated transfer of miRNAs [53]. Among miRNA, Shen et al. [53] focalized on miR-1246, an osteogenic miRNA demonstrating that facilitates the conversion of macrophages from a pro-inflammatory to an anti-inflammatory phenotypes [53], and Zhai et al. [45] also confirmed the presence of other osteogenic miRNAs in the exosome cargos.
It has been reported that no significant effects were observed using free exosome treatment, because of its rapid excretion from the site of application [78]. In fact, exosomes diffused out from the defect rapidly [79]; for this reason, a three-dimensional matrix that can support cell infiltration and vascularization is critical to support tissue neogenesis [80]. Scaffolds are 3D porous substrates promoting the cell-biomaterial interactions, permitting transport of gases, nutrients, and factors for cell survival, proliferation, and differentiation [81].
Particularly, special attention has been given to hydrogels, as they offer the possibility to generate well-defined 3D biofabricated tissue analogs to the natural ECM, and they have been identified to be proangiogenic during tissue healing [82].
Hydrogels have excellent comprehensive properties and are expected to become a new type of bone graft material, as their effects of promoting bone repair are more significant after the addition of MSC-derived exosomes [59]. Moreover, bone regeneration using hydrogels has been detected starting from 7 days after implantation in rats [62], and generally has been evidenced in 6 weeks after implantation [55,62].
Nevertheless, calcium phosphate based bioceramics are the most widely used osteoinductive materials, including TCP [57,83]. In fact, classical porous β-TCP scaffolds have a good bone conduction performance and bone repair mechanism, which has been evidenced 2 months after the implantation in animal models by Whang et al. [52], Zhang et al. [56], and Qi et al. [60], while this was shown for almost 3 months by Ying et al. [61].
Osteoconductive materials, able to support new bone formation on their surfaces, can also be introduced into non-biological materials (e.g., metal, ceramics, and synthetic polymers) by using various strategies such as coating or composite materials [45].
For example, although titanium is generally considered to be not osteoconductive, bone conduction has been obtained by using an appropriate surface treatment of the titania layer [57,84]. In fact, Zhai et al. [57] described the peculiarities of this kind of material, stating that titanium materials are biocompatible and non-toxic after implantation; they own good mechanical strength to support the bone; and their structures have optimal porosity for cell attachment, migration, and proliferation .
Osteoconductivity of synthetic polymers such as PLA and PCL scaffolds has been realized by developing composite materials with a calcium phosphate coating [50,85]. Interesting results have been obtained using PCL after 1 and 2 months in critical size defects of rat models, where an increase of both bone and vessels formation has been achieved [53].
PEI was often used for the engineering of EVs [48,49,51]. PEI, a complexed nucleic acid, is a well-known polymer useful for promoting the endosomal content release without the need for an additional endosomolytic agent [86]. Moreover, PEI has been demonstrated to have a high affinity to PLA scaffolds, activating this material [48,87]. Both Diomede et al. [48,51] and Pizzicanella et al. [49] evidenced an increase in bone content, mineral contents, and BMP, as well as angiogenesis, 6 weeks after the implantation of PLA-based scaffolds combined with EVs. All these data evidence the positive effects of scaffolds combined with EVs in bone regeneration; however, further studies are necessary to dissect if the mechanism of action is mainly dependent by the scaffold or EV types that were combined. Future research in this field is necessary to understand the therapeutic mechanism of EVs with scaffolds, also considering larger animal models (i.e., sheep) that are closer to humans. This systematic review shows that the most used cell sources are currently oral MSCs and bone marrow MSCs, followed by hucMSCs and hiPS. However, there is a lack of information about the differences between the vesicles derived from different stem source, as well as among the EVs types used, their size, and the isolation procedures, which clearly evidence the limitations of these studies. The most investigated EV types (Figure 1a) are exosomes, followed by EVs and secretoma. The studies show that both EVs and exosomes exert similar osteogenic and mineralization effects, leaving the question on the most suitable approach still open. A critical aspect has been represented by the isolation procedures, as there is no standardized method to isolate EVs, and therefore different products could be used that could represent a critical issue for clinical applications.
Finally, the proper dosage of EVs is another key factor to be considered. Generally, 1, 20, and 50 µg/mL are used, but not all works indicated their concentrations. Moreover, the time points evaluated for bone regeneration in rat and mice models was variable, ranging from 1 to 3 months.
Nevertheless, most of the studies reported a dose-dependent tissue repair of the EVs from MSCs, and particularly high concentrations in vitro (10-100 µg/mL) are preferred [88,89].
However, the lack of standardization and thus the presence of heterogeneous products does not allow for identifying the best EVs concentration for an optimal effect in terms of bone regeneration. Further efforts should investigate the protocols to optimize the EV production and concentration for bone-related applications and for the definition of the related advantages.
These promising results support the potential of this new biological approach, opening new future perspectives for stem cell-based therapy. However, it is crucial to have further efforts to standardize the reporting of the methodology. In this light, new research is required for the identification of the proper cell source; the best preparation protocol for EV isolation; and the most suitable scaffolds in bone regeneration, also in larger animal models in order to facilitate clinical translationality.

Conclusions
Increasing interest towards EVs as a cell free bone regeneration method has been underlined in this systematic review, with overall positive findings. Promising in vitro results have been discussed in terms of bone regeneration and pro-angiogenetic processes, as summarized in the cartoon (Figure 3). The positive in vitro findings were confirmed by in vivo studies, showing positive effects on several critical-size defects. However, some aspects remain to be elucidated, like the different effects induced by EVs and secretome, the most suitable cell source, the EVs production protocol and concentration, and the clinical use that may benefit from this new biological approach.