Extracellular Vesicles from Human Adipose-Derived Mesenchymal Stem Cells: A Review of Common Cargos

In recent years, the interest in adipose tissue mesenchymal cell–derived extracellular vesicles (AT-MSC-EVs) has increasingly grown. Numerous articles support the potential of human AT-MSC-EVs as a new therapeutic option for treatment of diverse diseases in the musculoskeletal and cardiovascular systems, kidney, skin, and immune system, among others. This approach makes use of the molecules transported inside of EVs, which play an important role in cell communication and in transmission of macromolecules. However, to our knowledge, there is no database where essential information about AT-MSC-EVs cargo molecules is gathered for easy reference. The aim of this study is to describe the different molecules reported so far in AT-MSC- EVs, their main molecular functions, and biological processes in which they are involved. Recently, the presence of 591 proteins and 604 microRNAs (miRNAs) has been described in human AT-MSC-EVs. The main molecular function enabled by both proteins and miRNAs present in human AT-MSC-EVs is the binding function. Signal transduction and gene silencing are the biological processes in which a greater number of proteins and miRNAs from human AT-MSC-EVs are involved, respectively. In this review we highlight the therapeutics effects of AT-MSC-EVs related with their participation in relevant biological processes including inflammation, angiogenesis, cell proliferation, apoptosis and migration, among others. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s12015-021-10155-5.


Introduction
"Extracellular vesicle" (EV) is defined by the International Society for Extracellular Vesicles (ISEV) as the "generic term for particles naturally released from the cell that are delimited by a lipid bilayer and cannot replicate, i.e. do not contain a functional nucleus" [1,2]. These particles contain a significant variety of proteins and RNAs that play important roles in cellcell communication and in transmission of macromolecules between cells [3][4][5][6]. As this feature makes EVs a potential therapeutic approach for various diseases, interest in EV research has significantly increased over the last decade [4,7]. Importantly, the profile of EV cargo depends on the cell type of origin [8]. In this sense, although a wide range of mammalian cells release EVs [4,9], mesenchymal stem cells (MSC) are considered one of the most prolific producer cell types [10]. These vesicles are involved in the paracrine properties of MSCs [11][12][13].
MSCs can be harvested from different tissues, such as bone marrow (BM), adipose tissue (AT), dental pulp, and umbilical cord, among others [14,15]. BM and AT are the most common sources of MSC for use in research [16][17][18][19]. Although BM-MSCs were the first identified MSC [20] type and have been extensively studied [21], AT-MSCs present remarkable advantages by comparison, including higher stability in culture conditions and lower senescence ratio [21]. In addition, the amount of MSC that can be obtained from this tissue, which is usually treated as waste material and discarded [22,23], is significantly greater than that obtained from BM aspirates [21].
The interest in AT-MSC-EVs has increasingly grown, due to the wide range of AT sources and their relatively easy accessibility [9]. AT-MSC-EVs have been isolated not only from human cells, but also from mouse [24][25][26][27][28][29][30][31][32], rat [33,34], pig [35][36][37][38], and rabbit [39,40] cells. The main objective of most published studies on AT-MSC-EVs was to evaluate their potential use as a new therapeutic approach to treat various diseases. Moreover, several of these publications did include an analysis of the molecules transported by the EVs, which is especially relevant to understanding their mechanism of action beyond their observable effects. Taken together, these studies have confirmed the presence of 591 proteins and 604 microRNA (miRNA) in the AT-MSC-EVs. Nevertheless, evaluation of effects of the molecules identified in the cargo focused solely on the disease or tissues under study. However, independent of the specific therapeutic use, the human AT-MSC-EVs are compositionally identical. Therefore, we anticipate that a review collecting together all available information about AT-MSC-EVs cargo and their function will be extremely useful for researchers working in this field.
ISEV recently published a guideline encouraging researchers to report their data to these field-specific databases to detect different studies describing the same molecules [1]. Thus, there is a great need for a well-organised review that collects all relevant information regarding molecules identified so far in AT-MSC-EVs cargo, and their biological activities. This will facilitate future research in this area. Currently, there are two online databases collecting the identified molecules in cargos of EVs derived from different cell types: http:// microvesicles.org [41] (formerly http://www.exocarta.org [42]), and http://evpedia.info [43] (link currently unavailable). Both databases are good, reliable sources of information; however, the information available on AT-MSC-EVs cargo is still limited compared to that available on other cell types, such as T cells or prostate cancer cell EV cargos. Thus, this review will provide an updated source not only of identified AT-MSC-EVs cargo molecules, but also their functions and potential therapeutic applications.
Given the growing interest in the MSC-EVs, especially in those derived from AT, the purpose of this study is to provide the AT-MSC research community with a systematic review of publications reporting the cargo of AT-MSC-EVs, including an analysis of their molecular functions and the biological process in which they are involved.

Methods
A systematic literature search was conducted in the medical databases Pubmed and Web of Science, using the keywords "extracellular vesicles", "exosome", "adipose mesenchymal stem cells", "cargo", "protein" and "miRNA" without setting a time limit (last searched 6th September 2020). 112 articles published between 2006 and 2020 (inclusive) were reviewed. 48 of these articles were related to human AT-MSC-EV, and 17 to AT-MSC-EVs in other species. The remaining articles were about EVs in general and MSC-EVs from other sources. This study has included both articles that used the nomenclature recommended by ISEV ("EV") [1] and those which used the terms "exosomes" and "microvesicles". Given the number of publications that have used these terms during the past decades [2], we considered that the exclusion of them could lead to the loss of relevant information. In addition, although the isolation methods of EVs could have an impact on the cargo composition, it was not an exclusion criterion since there is no single optimal separation method [1].
Different nomenclatures such as adipose stem cells, adipose stromal cells, or adipose-derived stem cells, have been used to identify AT-MSCs. The keyword "adipose mesenchymal stem cells" allowed us to find articles in which authors used several of these nomenclatures. However, we may have missed some information due to this great variety of terms, and this may be a limitation of the present study.
Information regarding proteins (10 articles) and RNA (16 articles) detected in human AT-MSC-EVs was collected in two databases created in Excel (Microsoft Office Excel 2013; Microsoft Corporation, Redmond, WA, USA). Although an article was found in which the lipid content of human AT-MSC-ECs was measured, no more information about lipids was reported. Therefore, it was not possible to include a database of lipids in this review.
To standardise the data and facilitate the recognition of identified proteins, we used the recommended name and identifier code proposed by the Universal Protein Knowledgebase [44] (UniProtKB). This database includes additional information about the short and alternative names for some proteins, which allowed us to identify proteins described by certain authors with these terms. UniProtKB host institutions are the European Bioinformatics Institute (EMBL-EBI), the Swiss Institute of Bioinformatics, and the Protein Information Resource.
For RNA, we used the name of mature micro RNAs (miRNAs) and the code of identification recommended by the RNAcentral database [45] (https://rnacentral.org/). This database is coordinated by EMBL-EBI and integrates information from 41 Expert Databases out of the 53 which constitute the RNAcentral Consortium. In addition, we used the miRBase database [46][47][48][49][50][51] to classify miRNAs by gene families. miRBase is one of the Expert Databases integrated in the RNAcentral database, and is managed by the University of Manchester. This database also includes information about the previous nomenclature of some miRNAs, which allowed us to correlate the previous miRNA name used by certain authors with the current recommended terminology.
Messenger RNA (mRNA) [52], transfer RNA (tRNA), small ribosomal RNA (rRNA), small nuclear RNA (snRNA), small nucleolar RNA (snoRNA) and small cytoplasmic RNA (scRNA) are also present in AT-MSC-EVs [53,54]. However, there is less information available on these, therefore, it was possible to include the list of the main tRNAs and mRNA present in AT-MSC-EVs, but not the other types of RNA.
Finally, the web-based tool QuickGO [55] (https://www. ebi.ac.uk/QuickGO/), also managed by EMBL-EBI, was used to search the gene ontology (GO) terms of molecular functions and biological processes of detected proteins and miRNAs. An ontology consists of a set of specific concepts with welldefined relationships between them. The GO was developed by the GO Consortium, as a tool to unify the terminology used to describe the functions of genes and gene products [56].
In the present review, we comprehensively describe the GO annotations of molecular functions and biological processes of each type of cargo reported in human AT-MSC-EVs.
So far, 591 proteins have been identified (Table 1). Nevertheless, taking into account both the name and the gene or NCBI Reference Sequences mentioned in the articles, it was not possible to connect the proteins C-peptide, HCR/ CRAM-A/B [52,65], INSL3, macroglobulin [65], CA 19-9, MSHa, PPARg2, TGF-beta 5 and TRA-1-60/TRA-1-81, Pepsinogen I [52] with an UniprotKB code conclusively (Table 1). The presence of the protein families annexin, HSP 70 and HSP 90 has also been described [12] (Table 1). However, as the specific members of these three families were not reported, it was not possible to include them in the GO analyses.
The detailed molecular functions enabled by each protein are collected in Table 1S. The results showed that 577 proteins contribute to different molecular functions described by 710 GO terms. For the BMP-binding endothelial regulator protein, carcinoembryonic antigen-related cell adhesion molecule, coagulation factor XIII B chain and kremen protein 2, no GO annotations were found.
The main molecular functions enabled by the AT-MSC-EVs proteins are described by specific child terms (more specific terms) of binding: protein binding (80%), metal ion binding (20%), cytokine activity (18%), identical protein binding (17%), and signaling receptor binding (15%) (Fig. 1). Therefore, binding seems to be the most relevant molecular function of AT-MSC-EVs. The number of AT-MSC-EVs proteins involved in each molecular function is variable. Most described molecular functions are enabled by a limited number of proteins (less than 10), and only 11.6% of the functions are enabled by 10 or more proteins.They are related by specific terms of four molecular functions: binding, catalytic activity, structural molecule activity and molecular transducer activity (Fig. 2).

Therapeutic Approaches of AT-MSC-EV Proteins
These results illustrate the role of AT-MSC-EVs in cell-cell communication [3][4][5][6], and the promising therapeutic effects observed in different research fields. Regarding the musculoskeletal system, AT-MSC-EVs have shown protective effects against cartilage degeneration, promotion of cell proliferation and migration of osteoarthritis chondrocytes, and antisenescence effects in osteoarthritis osteoblasts in vitro and in vivo [66,78]. They have also shown protective properties on muscle damage in an in vivo model of hindlimb X-linked interleukin-1 receptor accessory protein-like 2* IRPL2_HUMAN Q9NP60 IL1RAPL2 [65] *The referred article used alternative or short names **The specific member of this family detected has been not described ***Name and gene referred by the article cited ischemia and in an in vitro model of ischemia/reperfusion [52]. These effects may be a consequence of the presence of proteins such as lactotransferrin, C-X-C motif chemokine 16, protein Wnt-5a, and transforming protein RhoA, which are involved in positive regulation of chondrocyte proliferation, positive regulation of cell migration, regulation of inflammatory response and regulation of osteoblast proliferation, respectively. The complete list of proteins involved in these processes is reported in Table 2S. With regard to cardiology and vascular system, AT-MSC-EVs are involved in a wide range of biological processes, including heart development, contraction and morphogenesis, positive regulation of cardiac muscle cell proliferation and hypertrophy, regulation of cardiac muscle cell apoptotic process and proliferation, blood vessel maturation, remodeling and morphogenesis, regulation of blood vessel diameter and angiogenesis, among others (Table 2S). Hence, numerous proteins detected in AT-MSC-EVs could account for the protective effects observed in cardiac function and cardiomyocytes after their injection in an in vivo model of myocardial infarction [79] . In addition, the effects of AT-MSC-EVs in angiogenesis have been also studied in vitro and in vivo [60,72,80]. Proteins detected in AT-MSC-EVs such as IL-1 alpha and apelin receptor are proangiogenic, while SPARC is antiangiogenic (Table 2S).
Human AT-MSC-EVs also have an inhibitory effect on vein graft neointima formation, as observed in a mouse model of vein grafting [81]. This effect correlated with decreased macrophage infiltration, attenuated inflammatory cytokine expression, and reduced activation of MAPK and phosphatidylinositol-3 kinase signaling pathways [81]. EV proteins potentially involved in these processes are thrombospondin-1 (inflammatory response), IL-4 (negative regulation of macrophage activation), growth factor receptor-bound protein 2 (regulation of MAPK cascade) and MAP kinase 1 (regulation of phosphatidylinositol 3-kinase signaling) (Table 2S).
The effects of AT-MSC-EVs proteins in the vascular system may also be related to the cardio-renal protection observed in a deoxycorticosterone acetate-salt hypertensive animal model [82]. Thus, the administration of AT-MSC-EVs in this in vivo model protected against renal damage, preserved renal function, reduced inflammatory response, prevented fibrosis in the kidney and in cardiac tissue, and conserved normal blood pressure [82]. The administration of AT-MSC-EVs also showed a renal protective effect in an in vivo model of acute kidney injury [83]. Proteins detected in AT-MSC-EVs such as integrin alpha-3, IL-4, IL-10, collagen alpha-2(I) chain or periostin could be implicated in these outcomes (Table 2S).
In this review, we present a comprehensive analysis of miRNAs currently identified in human AT-MSC-EVs. 489 miRNAs from 255 gene families were classified. The mir-515 and mir-10 families have the greatest numbers of miRNAs (Table 2). However, there was no information available about which gene families the other 115 miRNAs belonged to. In addition, hsa-miR-320a-3p and hsa-miR-375-3p were identified by the sequence and the precursor reported by Reza et al. [54], since the actual names used in the reference, hsa-miR-320a and hsa-miR-375, respectively, were not found for mature miRNA in any of the databases. Hsa-miR-1273a [54,66] was included in the miRBase database as a dead miRNA entry. It was eventually removed due to lack of consistency between the patterns of mapped reads from RNA-sequencing experiments and the gene being processed as a miRNA. hsa-miR-1274a, hsa-miR-1274b, hsa-miR-1300 and hsa-miR-720 [65] were also included in the miRBase database as dead miRNA entries. They were removed because it is likely that they are fragments of tRNAs and mRNA. This could be the reason for their absence from  (Table 2). Other special cases included hsa-miR-548aa and hsa-miR-548 t-3p [66] there is a specific entry for each one in the miRBase database, however, both entries showed the same sequence and RNAcentral link. Therefore, in the present review they are treated as the same miRNA. The same applies to hsa-miR-199b-3p and hsa-miR-199a-3p [53,65,66,72].
The variety of miRNAs present in AT-MSC-EVs may play a role in the different therapeutic effects based on the paracrine properties of MSC [13]. Regardless, to confirm the involvement of miRNAs in these effects, it is necessary to take into consideration not only the presence of a specific miRNA, but also other factors such as concentration, structure, and availability of accessory proteins [13].
Only 199 miRNA showed GO annotations for molecular function when using the QuickGO database [55]. The molecular functions enabled by these miRNAs are mRNA binding involved in post-transcriptional gene silencing (95%), mRNA 3'-UTR binding (22%), RNA polymerase II complex binding (6%), single-stranded RNA binding and high-density lipoprotein particle binding (2% each), protein binding, transcription regulatory region sequence-specific DNA binding and sequence-specific single stranded DNA binding (1% each) (Fig. 5). All of these functions are specific child terms of the binding function (Fig. 6) which is also the most relevant molecular function of AT-MSC-EV proteins, as previously described. The specific molecular functions enabled by each miRNA are detailed in Table 3S.
The number of miRNAs with GO annotations of biological processes in QuickGO [55] was 212. These miRNAs take part in biological processes described by 577 different GO terms. The biological processes in which the greatest number of miRNA are involved are: negative regulation of gene expression, response to stimulus, regulation of cellular process, developmental process, locomotion, signaling, and cell communication (Fig. 7). The specific miRNAs involved in each process are detailed in Table 4S. 89% of these miRNAs are involved in gene silencing (Fig. 8). Other relevant GO terms in which a large number of miRNAs are included are miRNA mediated inhibition of translation (28%) negative regulation of gene expression (17%), negative regulation of angiogenesis (14%), negative regulation of inflammatory response (13%) and negative regulation of cell migration involved in sprouting angiogenesis (11%) (Fig. 8).

Lipids
The third type of molecule transported by EVs is lipids [3,4]. The lipid composition of EVs has been less studied than that of proteins or miRNAs [8]. Thus, the number of lipid entries (639) in the Vesiclepedia database [41] is notably lower than the number of protein and miRNA entries (349,988 and 10,520, respectively). None of these lipid entries are related to AT-MSC-EVs or any other MSC-EVs. The total lipid content of AT-MSC-EVs has been analysed by Bari et al. [58], using the Nile Red assay. However, to our knowledge, there is no detailed information about the different types of lipids present in AT-MSC-EVs.

Modification of Cargo Components to Improve their Potential Effects
Different cell culture conditions and pre-treatments have been used to modify the profile of human AT-MSC-EV cargo, with the aim to improve its effects in skin flap survival [59,86], angiogenesis [60,61,64,80], immune response [71,87], bone regeneration [77] and cancer [118,119]. To this purpose, human AT-MSCs have been exposed to oxidative stress [59,86], hypoxic [61,80] or inflammatory culture conditions [71,87], stimulation with platelet-derived growth factor (PDGF) [60,65] and basic fibroblast growth factor (bFGF) Fig. 7 Simplified outline of the main biological processes in which the miRNA detected in EVs derived from human AT-MSC are involved. For a complete review of the relationships between gene ontology terms see the chart view in the web-based tool QuickGO (https://www.ebi.ac.uk/QuickGO/) [64] and transfected with lentiviral particles with different miRNAs [77,118,119].
Hypoxic culture conditions also induce the release of larger EVs according to Han et al. [61], although other authors claim that there are no significant differences in size [80]. The EVs collected from AT-MSC cultured under hypoxic conditions (5% O 2 ) seemed to enhance angiogenic properties in cultured human umbilical vein endothelial cells and in an in vivo model of fat grafting [61,80]. The results of these studies showed that the amount of the surface marker CD44 was significantly lower in hypoxic EVs [80], while VEGF-A, EGF, FGF-4, VEGFR-2, VEGFR-3, C-C motif chemokine 8 and 13 were increased under these culture conditions [61].
EVs contents are also different after AT-MSC exposure to inflammatory cytokines. In EVs secreted by INF-γ-stimulated AT-MSC, indoleamine 2,3-dioxygenase mRNA was detected, although its presence did not significantly improve their potential to control activated T cell proliferation, in comparison with those derived from unstimulated AT-MSC [87]. However, when AT-MSCs were pretreated with both INF-γ and TNF-α, the enriched EVs induced the polarization of macrophages to the M2 phenotype [71]. Under this proinflammatory culture condition, AT-MSC-EVs cause differences in the expression of 81 different miRNAs [71] (Table 3).
Finally, AT-MSC-EVs have been transfected with lentiviral particles to produce EVs enriched in miRNA 375 [77], miRNA-125b [119] and miRNA 101 [118]. The miRNA-375-enriched EVs promoted bone regeneration in an in vivo model of calvarial defects. AT-MSC-EVs enriched in miRNA-125b [119] and miRNA 101 [118] induced a reduction in cell proliferation of hepatocellular carcinoma cells and inhibited osteosarcoma cell invasion and migration in vitro, respectively. In addition, miRNA-101-enriched EVs also induced inhibition of osteosarcoma metastasis in a lung metastasis model in vivo [118].

Conclusions
There is an increasing interest in the study of EVs as new therapeutic options in several research fields, due to their role in different biological processes, including cell proliferation, apoptosis, angiogenesis, inflammation and immune response, among others. Their potential is based upon the molecules transported inside these particles. Therefore, both molecule identification and an understanding of the molecular functions and biological processes in which they are involved are essential to advance this area of research. To the best of our knowledge, the presence of 591 proteins and 604 miRNAs in human AT-MSC-EVs has been described. The most important molecular function enabled by them is the binding function, which supports their role in cell communication. Regarding the biological processes, the proteins detected are mainly involved in signal transduction, while most miRNAs take part in negative regulation of gene expression. The involvement of both molecules in essential biological processes such as inflammation, angiogenesis, cell proliferation, apoptosis and migration, supports the beneficial effects of human AT-MSC-EVs observed in both in vitro and in vivo studies, in diseases of the musculoskeletal and cardiovascular systems, kidney, and skin.
Interestingly, the contents of AT-MSC-EVs can be modified by cell stimulation and different cell culture conditions, such as oxidative stress or hypoxia, to engineer a cargo selection with improved antigenic, anti-inflammatory or immunosuppressive effects. Moreover, it is also possible to enrich specific miRNAs in the cargo via transfection of AT-MSC with lentiviral particles. These modifications have enhanced the positive effects in skin flap survival, immune response, bone regeneration and cancer treatment. This phenomenon opens new avenues to examine the therapeutic potential of AT-MSC-EVs.

Declarations
Ethical Approval Not applicable.
Consent to Participate Not applicable.

Consent to Publish Not applicable.
Competing Interests The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.