The incidence of bone-related disorders is continuously growing as the aging of the population in developing countries continues to increase. Although therapeutic interventions for bone regeneration exist, their effectiveness is questioned, especially under certain circumstances, such as critical size defects. This gap of curative options has led to the search for new and more effective therapeutic approaches for bone regeneration; among them, the possibility of using extracellular vesicles (EVs) is gaining ground. EVs are secreted, biocompatible, nano-sized vesicles that play a pivotal role as messengers between donor and target cells, mediated by their specific cargo. Evidence shows that bone-relevant cells secrete osteoanabolic EVs, whose functionality can be further improved by several strategies. This, together with the low immunogenicity of EVs and their storage advantages, make them attractive candidates for clinical prospects in bone regeneration. However, before EVs reach clinical translation, a number of concerns should be addressed. Unraveling the EVs’ mode of action in bone regeneration is one of them; the molecular mediators driving their osteoanabolic effects in acceptor cells are beginning to be uncovered. Increasing the functional and bone targeting abilities of EVs are also matters of intense research.
1. Introduction
Bone is a dynamic organ that is constantly remodeling to ensure a constant net bone mass within an organism. This is achieved by two opposite and balanced phases: bone formation and bone resorption, carried out by osteoblasts and osteoclasts, respectively. Moreover, bone has a self-repairing ability, and therefore when an injury occurs, the damaged part regains its original structure and mechanical strength by the activation of the bone healing process. However, under certain situations, such as critical-sized bone defects (defined as those that will not heal spontaneously during the life of a patient), bone loss due to bone-related diseases, lack of vascularization, infections, and tumors, this ability is not enough and requires clinical intervention
[1]. In the case of osteoporosis (OP), a highly prevalent metabolic disease affecting more than 200 million of patients worldwide, each year more than 8.9 million fractures are reported globally
[2]. In the EU, the cost of OP in 2019, including pharmacological intervention, entailed more than €56 billion, doubling that needed in 2010
[3][4].
Bone tissue is highly demanded in clinics, and after blood, is the second most transplanted tissue worldwide
[5]. Thus, currently, the most successful intervention to treat bone defects is still bone grafting, a strategy first outlined in the early 1900s
[6]. Bone autografts are the gold standard in bone regeneration procedures since they avoid rejection from the patient’s immune system. Unfortunately, autologous bone supply is limited and the need of additional surgery for bone extraction increases the risk of infections and morbidity. Another option is the use of allografts (from a human donor) or xenografts (from large animals like pigs or bovines), which nevertheless entail some risk of pathogen transmission; more importantly, these bone implants have shown poor bone regeneration abilities
[7]. All in all, there is an urgent need to discover new, effective therapies to boost bone regeneration to satisfy the growing world population (progressively more aged) affected by bone conditions.
On this basis, the field of bone tissue engineering has emerged, focused on developing “bone substitutes” that mimic the bone tissue features, usually formed by a 3D scaffold and bone-relevant cell types, which are able to promote osteogenic differentiation in host tissue without any adverse inflammatory response
[8]. The features of mesenchymal stem cells (MSCs), such as their capacity to undergo osteogenic differentiation, immunomodulatory potential, and trophic effects, make them quite attractive components for these bone constructs
[8][9]. In fact, the use of MSCs-based therapies in clinics is gaining interest in the field of bone regeneration due to the clinical improvements exhibited by patients affected by bone-related diseases, such as Osteogenesis Imperfecta (OI) after MSCs administrations
[8][10][11][12]. However, there are still some limitations to the clinical translation of MSCs, such as the large number of cells that are required as well as the high cellular heterogeneity, even within populations of MSCs from the same donor. In addition, other factors, such as cell culture conditions, cell source, and donor age, have determined the variable outcomes shown in different clinical trials using MSCs
[13]. Regarding the mode of action of MSCs therapy in bone regeneration, quite revealing findings point to the paracrine mechanisms elicited by these cells, rather than the initially expected cell engraftment and subsequent osteogenic differentiation
[10][14]. Moreover, a set of experiments performed in OI animal models, suggested that the extracellular vesicles (EVs) secreted by MSCs could be mediating the recovery of bone phenotypes observed in OI patients subjected to MSCs therapy
[15].
EVs are small, lipid membrane delimited particles secreted by most cell types and present in several biological fluids, such as blood and urine, that play a key role in cell-to-cell communication
[16]. This paracrine crosstalk is mediated by the EVs’ cargo, an array of bioactive molecules including proteins, lipids, and nucleic acids, that interestingly exhibit a parent-cell-specific signature
[17][18]. Thus, through specific surface molecule interaction, EVs can be uptaken by target cells and modify their biology/fate
[19]. Attending to their size, the EVs population can be divided into exosomes (diameters of 30–200 nm), microvesicles (diameters of 200–1000 nm), and apoptotic bodies (diameters > 1000 nm)
[19][20]. The biogenesis of EVs is a complex process and currently it is accepted that each EV subtype can be originated by two predominant pathways: the endosomal pathway, through multivesicular endosome fusion, or by the outward budding and fission of the plasmatic membrane
[21]. The fact that EVs carry functional molecules that can modulate target cell responses opens the possibility of using EVs as next-generation drug delivery platforms, a vision strongly supported by their low immunogenicity. Thus, EVs are known to escape from immune clearance when systemically administered, due to the expression of surface molecules, such as CD47, which mediates the “don’t eat me” signal that blocks phagocytosis
[22]. Bone-relevant cell types have been shown to secrete EVs that regulate bone homeostasis, and in addition, recent evidences suggest that EV therapy is at least as efficient as cell therapy in eliciting bone regeneration in large bone defect animal models
[19][23][24]. Thus, the use of EVs as osteoanabolic delivery systems could be a reliable clinical approach for bone regeneration
[25].
2. EV Sources for Bone Regeneration and Mechanisms of Action
From a clinical perspective, most research in the field of bone regeneration and EVs, which has grown exponentially in the last years, has focused on either exosomes or microvesicles. However, these EV populations are not fully characterized, in spite of the remarkable efforts made by the International Society of Extracellular Vesicles (ISEV) to determine the minimal information for studies regarding EVs (Minimal information for studies of extracellular vesicles 2018 (MISEV2018): a position statement of the International Society for Extracellular Vesicles and update of the MISEV2014 guidelines, doi:10.1080/20013078.2018.1535750). Following these recommendations, herein will use EVs as a broad term encompassing all the vesicles secreted by cells
[26].
Bone resident cells, including bone-cell types (MSCs, osteoblasts, osteoclasts, and osteocytes) as well as other cell populations present in the bone microenvironment, such as endothelial cells (ECs) and macrophages, secrete EVs that mediate the continuous bone remodeling process (
Figure 1)
[23]. Interestingly, the cargo contained in the EVs reflects the biological function of parent cells. Thus, osteoclast-secreted EVs can inhibit osteoblast activity and therefore suppress bone formation, whereas the EVs secreted by osteoblastic-lineage cells, such as MSCs and osteoblasts, enhance osteoblast differentiation in vitro and promote bone regeneration in vivo
[27][28]. Moreover, osteoblast-derived EVs also can be uptaken by osteoclasts, but the consequences triggered in these cells are conflicting. Thus, the promotion of osteoclasts’ differentiation in vitro and in vivo has been described, but also the inhibition of osteoclastogenesis, and therefore of bone resorption
[29][30][31]. Current evidence, mainly obtained from basic and preclinical experimentation, strongly suggest that the specific cargo of EVs determines the signaling triggered in recipient cells. In this line, miRNAs contained in EVs are known to be transferred to target cells as a mechanism of genetic exchange between cells, playing a key role in regulating bone homeostasis
[32]. Actually, a specific, dysregulated miRNAs signature has been described in plasma/serum EVs of some pathological conditions, including bone disease. This is quite valuable in silent pathologies, such as OP, which has no clinical manifestation until a fracture occurs, being that these circulating miRNAs, proposed as potential biomarkers, are capable of predicting the risk of fracture
[33][34].

Figure 1. Illustration of the bone-relevant cell types known to secrete EVs that promote bone regeneration. Pro-osteogenic and pro-angiogenic EVs, determined by their specific cargo, are produced by cell types present in bone microenvironment, such as mesenchymal stem cells (MSCs), osteoblasts, osteocytes, macrophages, and endothelial cells (ECs). Pro-osteogenic EVs stimulate MSCs and osteoblasts differentiation, inducing the bone formation process, while pro-angiogenic EVs elicit the formation of new blood vessels in bone tissue. Both processes are essential to conduct a successful regeneration of bone tissue. The figure was created with
BioRender.com (accessed on 1 December 2021).
Herein will discuss the current knowledge about the bone-relevant cell types that secrete osteoanabolic EVs and the current identified mechanisms, mainly driven by miRNAs (and some proteins, to a lesser extent) (Table 1), by which these EVs exert their function in target cells.
Table 1. Main cell sources that secrete osteoanabolic EVs and the identified cargo mediating bone regeneration in different mice models of bone disease.
| EVs Source |
Bioactive Cargo |
Disease Model |
Target Molecule-Pathway |
Target Process |
Ref. |
| Bone marrow (BM)-MSCs |
miR-335 |
Bone fracture |
VAPB-WNT/β-CATENIN |
↓ Osteoclastogenesis & ↑ Osteogenesis |
[35] |
| miR-25 |
Bone fracture |
SMURF1-RUNX2 |
↑ Osteogenesis |
[36] |
| NID1 |
Femoral defects |
Myosin-10 |
↑ Angiogenesis |
[37] |
| mir-29a |
Wild type mice |
VASH1 |
↑ Angiogenesis & ↑ Osteogenesis |
[38] |
| Aged BM-MSCs |
mir-128-3p |
Bone fracture |
SMAD5 |
↓ Osteogenesis |
[39] |
| Umbilical cord (UC)-MSCs |
miR-1263 |
Disuse osteoporosis (OP) |
MOB1-HIPPO |
↓ Apoptosis |
[40] |
| miR-21 |
Glucocorticoid-induced osteonecrosis of the femoral head (GIONFH) |
PTEN-PI3K/AKT |
↓ Apoptosis |
[41] |
| miR-365a-5p |
GIONFH |
SAV1-HIPPO |
↑ Osteogenesis |
[42] |
| miR-3960 |
Senile OP |
unknown |
↑ Osteogenesis & ↑ Osteoclastogenesis |
[43] |
| CLEC11A |
Ovariectomized (OVX)-OP, Disuse OP, Senile OP |
unknown |
↑ Osteogenesis & ↓ Osteoclastogenesis |
[44] |
| Hypoxia-UC-MSCs |
mir-126 |
Bone fracture |
SPRED-1 |
↑ Angiogenesis |
[45] |
| ECs |
miR-126 |
Distraction osteogenesis |
SPRED-1 |
↑ Osteogenesis & ↑ Angiogenesis |
[24] |
| miR-155 |
OVX-OP |
Spi1, Mitf, Socs1 |
↓ Osteoclastogenesis |
[46] |
3. Novel Strategies to Improve the Bone Regenerative Potential of EVs
The bone regenerative potential of EVs, especially if they come from MSCs and ECs, is an undeniable fact nowadays, supported (as been mentioned above) by a considerable number of basic and preclinical studies
[47]. Nevertheless, the use of EVs as an advanced therapy for bone regeneration has been hampered mainly by two observations inherent to EV biology. First, the osteoanabolic potential of EVs is far from being optimum to achieve complete bone regeneration, and therefore these osteoanabolic abilities should be improved; second, osteoanabolic EVs do not mainly target bone tissue. Thus, upon intravenous administration in mice, EVs show a rapid (within the first hour) tissue distribution, accumulating mainly in the spleen, liver, lung, and kidneys
[48][49]. Even so, there are also studies indicating the accumulation of MSCs-derived EVs in bone tissue, although to a lesser extent
[50]. On the contrary, EVs coming from osteoclasts, known to negatively regulate bone formation by targeting and inhibiting osteoblasts, have shown acceptable intra-osseous accumulation in injected mice
[28]. Therefore, current attempts pursuing the production of EVs with the maximum bone regenerative potential mainly rely on the enhancement of the EVs’ osteoanabolic abilities as well as on their bone cell targeting and, therefore, bone tissue.
3.1. Enhancing the Osteoanabolic Potential of EVs
MSCs undergoing osteogenic differentiation, especially those isolated from bone marrow (BM), have demonstrated acceptable bone regeneration properties due to two facts: their EVs show increased bone targeting potential and exhibit osteoanabolic-specific cargo
[51][52][53]. Hence, it is not surprising that the vast variety of investigations focus on enhancing the osteogenesis of parent MSCs in order to achieve innate EVs with maximum osteoanabolic and bone targeting abilities.
3.1.1. Preconditioning of Parent Cells
Preconditioning of MSCs’ culture conditions, either by the addition of exogenous molecules (cytokines, growth factors, drugs) or by the optimization of physical factors (hypoxia or shear stress), has been proposed, as these strategies induce a robust osteogenic differentiation in MSCs in order to obtain highly osteoanabolic EVs (
Figure 2)
[54]. Along this line, the mimicking of the bone healing signaling milieu, such as that occurring in the inflammatory phase upon bone injury, has been demonstrated to be effective. Thus, when priming AT-MSCs with TNF-α, a specific pro-inflammatory molecule, the secreted EVs showed enhanced abilities in promoting the proliferation and osteogenic differentiation of human primary osteoblastic cells. Interestingly, an increase in WNT3a protein, a known inducer of osteogenesis, was detected in the cargo of these EVs
[55].

Figure 2. Preconditioning strategies to enhance the osteoanabolic potential of EVs. Pre-treatment of MSCs with inflammatory factors or histone deacetylase inhibitors enhance their osteogenic differentiation, whereas hypoxia conditions elicit pro-angiogenic responses in these cells, leading to the secretion of pro-osteogenic or pro-angiogenic EVs, respectively. Mechanistically, the mimicking of the bone inflammatory microenvironment after bone injury triggers the expression of the pro-osteogenic protein WNT3a in MSCs, which in turn, is enriched in the EVs secreted by these cells. The inhibition of histone deacetylases, such as via the use of thrichostatin A (TSA), elicits an epigenetic reprogramming of MSCs, ensuring an open conformation of chromatin and promoting the transcription of pro-osteogenic genes. The hypoxia simulation in MSCs, achieved by low oxygen cell culture or by chemical compounds (for instance dimethyloxaylglycine (DMOG)), induces the activation of the HIF-1α transcription factor, which drives the cell responses to hypoxia, among them being hypoxia-induced angiogenesis. The figure was created with
BioRender.com (accessed on 1 December 2021).
3.1.2. Engineering of Parent Cells
MSCs can be genetically modified to increase the expression of certain pro-osteogenic molecules with the assumption that, this way, their secreted EVs would also be enriched in those induced molecules, enhancing their osteoanabolic properties (
Figure 3). In fact, this hypothesis has been validated by overexpressing pro-osteogenic miRNAs in MSCs, such as miR-375 and miR-101. The authors demonstrated that EVs could be enriched in these miRNAs when overexpressed in parent MSCs, without affecting distinctive features of EVs such as morphology, size, and the expression of surface proteins CD9 and CD63, which are used as EV markers. Moreover, these EVs improved the osteogenic differentiation of MSCs and enhanced bone regeneration in animal models of bone defects
[56][57].

Figure 3. Genetic engineering as an approach to enrich EVs with osteoanabolic factors. The induced expression of known osteoanabolic miRNAs, proteins, or inhibitors of anti-osteogenic miRNAs in MSCs, by using expression vectors or direct transfection approaches of these molecules, yields EVs enriched in these molecules. The figure was created with
BioRender.com (accessed on 1 December 2021).
3.2. Directing EVs to Target Bone Tissue
EV therapy to treat skeletal conditions can be delivered by local or intravenous administration, and each one has their specifications and advantages/disadvantages. The local administration, most suitable for concrete bone fractures or defects, ensures the bone targeting of EVs, but requires the concomitant use of scaffolds, such hydrogels, in order to maintain the EVs in the site of injury. On the contrary, when considering global skeletal conditions, such as OP or OI, the intravenous administration of EVs is the considered option; nevertheless a major drawback of this administration route is the low bone tropism that EVs show. Some innovative strategies are currently under intensive research in order to achieve and enhance the bone targeting of EVs; especially promising are those that directly modify the EVs’ surface with specific bone-targeting molecules (Figure 4).

Figure 4. Functionalization of EVs’ surface to improve bone targeting. The surface modification of EVs with molecules showing affinity for bone cells has been described. This is the case of specific aptamers, DNA/RNA molecules with affinity for a desired target, and in this case MSCs and osteoblasts. Anti-resorptive drugs, such as bisphosphonates (BPs), which show high affinity for the mineralized bone matrix, have also been covalently bound to the surface of EVs. Both approaches have demonstrated increased bone targeting of functionalized EVs. The figure was created with
BioRender.com (accessed on 1 December 2021).
4. Conclusions
The increasing knowledge about EV biology has strengthened the idea that EVs hold great potential to be applied with therapeutic purposes, mainly due to their ability to transfer diverse bioactive molecules modifying the fate of recipient cells. Thus, EVs may offer a promising “cell free” advanced therapy as next-generation biocompatible vehicles delivering therapeutic factors.
However, before EVs move forward to the clinic, it is mandatory to address several requirements that challenge their claimed therapeutic abilities, including standardization and scalability production, their full molecular characterization, and bioengineering improvements that increase their therapeutic potency. Moreover, the inherent biology of the target tissue plays a key role in the success of EV-based therapies. When intended for bone regeneration purposes, EV therapeutics have to overcome two main limitations, both matters of intense research: the osteoanabolic properties of EVs, which should be enhanced in order to achieve robust, in vivo bone regeneration and the limited tropism for bone tissue that the osteoanabolic EVs show upon administration. Therefore, to achieve bone regeneration, the ideal EVs should combine features aiming to counteract these two limitations.
EVs isolated from a wide range of bone-relevant cells have demonstrated osteoanabolic potential. However, the fact that the majority of studies only rely on EVs isolated from a single cell type hinders the comparison of their osteoanabolic capacity. Therefore, the systematic analysis of different EVs isolated from different cell types abdcomparing their osteogenic capacity should be a prerequisite to identify those EVs, or their combination, with the maximum osteoanabolic potential. This knowledge will come along with the understanding of the mode of action of EVs, and to achieve it, essential requirements should be considered, such as deciphering the molecular players driving the downstream signaling of EVs in target cells. Accordingly, comprehensive multi-omic technologies have enabled a deep characterization of EV cargo, but the identification of those molecular drivers in EVs conducting bone regeneration is just beginning to emerge. So far, the majority of the current research has identified several single molecules, especially miRNAs and some proteins, as drivers of the EVs downstream regulation in the recipient cells. However, considering that EVs carry an array of molecules and that EVs from different cell sources achieve the induction of bone regeneration, it is more likely for a synergistic collaboration of different molecules in target cells to occur, as opposed to a single upstream molecular regulator. In fact, recent evidence point to this observation: Lee and collaborators reported that AT-MSCs-EVs attenuated bone loss in OVX mice by the simultaneous transfer of proteins and miRNAs targeting osteoclasts. Thus, the inhibition of osteoclastogenesis elicited by AT-MSCs-EVs and the subsequent restoration of bone mass in OVX mice was mediated by the transfer of osteoprotegerin (OPG), a decoy receptor for RANK ligands that inhibits osteoclasts differentiation, and miR-21-5p and let-7b-5p, which reduced osteoclast differentiation
[50]. Liu and coworkers also identified a multi-component pro-osteogenic miRNAs cargo in BM-MSCs-EVs: let-7a-5p, let-7c-5p, miR-328a-5p, and miR-31a-5p. These miRNAs were shown to synergistically mediate the osteoanabolic properties of BM-MSC-EVs by promoting the activation of the canonical BMP signaling pathway
[58].
The increasing knowledge about the most suitable EV cell source and the bioengineering approaches under development will address the aforementioned limitations facilitating the development of EV-based therapeutics that will transform the pharmaceutical scene for bone regeneration. Currently (as of December 2021), there is one clinical trial testing EVs, specifically exosomes, as therapeutic drugs applied for a bone disease: a phase I trial evaluating intra-articular injections of a single dose allogenic MSCs-derived exosomes for knee osteoarthritis (ExoOA-1; NCT05060107). It is anticipated that, as different approaches demonstrate improvements in the osteoanabolic potential and bone-targeting abilities of EVs, there will be increasing clinical trials evaluating the safety and potential of this advanced therapy for bone regenerative purposes in the not-so-distant future.