Lambert et al. Stem Cell Research & Therapy (2019) 10:361 https://doi.org/10.1186/s13287-019-1460-1 RESEARCH Open Access Stem cells from human cardiac adipose tissue depots show different gene expression and functional capacities Carmen Lambert1, Gemma Arderiu1*, Maria Teresa Bejar1, Javier Crespo1, Maribel Baldellou1, Oriol Juan-Babot1 and Lina Badimon1,2,3* Abstract Background: The composition and function of the adipose tissue covering the heart are poorly known. In this study, we have investigated the epicardial adipose tissue (EAT) covering the cardiac ventricular muscle and the EAT covering the left anterior descending artery (LAD) on the human heart, to identify their resident stem cell functional activity. Methods: EAT covering the cardiac ventricular muscle was isolated from the apex (avoiding areas irrigated by major vessels) of the heart (ventricular myocardium adipose tissue (VMAT)) and from the area covering the epicardial arterial sulcus of the LAD (PVAT) in human hearts excised during heart transplant surgery. Adipose stem cells (ASCs) from both adipose tissue depots were immediately isolated and phenotypically characterized by flow cytometry. The different behavior of these ASCs and their released secretome microvesicles (MVs) were investigated by molecular and cellular analysis. Results: ASCs from both VMAT (mASCs) and the PVAT (pASCs) were characterized by the expression of CD105, CD44, CD29, CD90, and CD73. The angiogenic-related genes VEGFA, COL18A1, and TF, as well as the miRNA126-3p and miRNA145-5p, were analyzed in both ASC types. Both ASCs were functionally able to form tube-like structures in three-dimensional basement membrane substrates. Interestingly, pASCs showed a higher level of expression of VEGFA and reduced level of COL18A1 than mASCs. Furthermore, MVs released by mASCs significantly induced human microvascular endothelial cell migration. Conclusion: Our study indicates for the first time that the resident ASCs in human epicardial adipose tissue display a depot-specific angiogenic function. Additionally, we have demonstrated that resident stem cells are able to regulate microvascular endothelial cell function by the release of MVs. Keywords: Epicardial adipose tissue, Perivascular adipose tissue, Ventricular myocardium adipose tissue, Adipose stem cells, Microvesicles, Angiogenesis Background investigated for its potential to repair/regenerate the in- Cardiovascular diseases (CVDs) are the leading cause of jured tissue within the infarct zone [1–4]. death and disability worldwide, and coronary artery dis- Adipose tissue (AT) has long been considered an energy ease makes up the greatest proportion of those deaths. storage and an endocrine organ. However, in the last de- Acute myocardial infarction leads to an irreversible loss cades, adipose tissue has also been considered a rich of proper cardiac function; implantation of adult stem source of mesenchymal cells and is currently the focus of cells into the ischemic damaged myocardium has been interest in the field of inducible spontaneous regeneration and cell therapy. Adipose-derived stem cells (ASCs) are * Correspondence: garderiu@santpau.cat; lbadimon@santpau.cat easily obtained, show a strong capacity for ex vivo expan- 1 Cardiovascular-Program ICCC, IR-Hospital Santa Creu I Sant Pau, IIB Sant Pau, sion and differentiation to other cell types (as cardiomyo- C/Sant Antoni Ma Claret 167, 08025 Barcelona, Spain cytes [5] or endothelial cells [6, 7]), release a large variety Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 2 of 12 of angiogenic factors, and have immunomodulatory prop- (PVAT)] and those residents in the EAT directly superim- erties. All these properties have encouraged its use to in- posed on the cardiac muscle (VMAT) covering the apex duce ischemic tissue recovery. However, their clinical use of the heart. The contribution to these two stem cell reser- of bone marrow stem cells has been hampered by the rec- voirs to cardiac cell function and to cardiac endogenous ognition of the loss of function effects of aging, cardiovas- repair processes after an ischemic injury is unknown. As cular risk factors (CVRFs), and metabolic disorders [8]. the coronary arteries and their major branches are imbed- Indeed, regarding spontaneous endogenous repair, our ded in PVAT, this creates a perfect environment for the group has shown that the presence of CVRFs, such as type local interaction between ASCs present in this fat and the 2 diabetes mellitus, negatively affects the pluripotency and coronary vessels it surrounds. Angiogenesis is essential for self-renewal capacities of adipose stem cells (ASCs) [9], the repair of wounded or ischemic organs, and insufficient and ASCs obtained from obese patients display an im- angiogenic switch in an ischemic heart can limit revascu- paired angiogenic potential [10, 11]. Indeed, the spontan- larization, healing, and regeneration [22]. eous regenerative capacity for ASC self-renewal seems to Here, we hypothesized that EAT from two different lo- be regulated by the anatomical WAT depot. cations in the human heart may contain a repository of Besides the commonly known CVRFs that are known to stem cells (ASCs) with different pro-angiogenic functions. elevate the risk of suffering CVDs, excess adiposity is con- sidered a CVRF for CVD [12]. The relationship between Materials and methods obesity and the development of any CVDs lies not only in Sample collection/patients the amount of fat, but also in its localization. AT is highly VMAT and PVAT were obtained from the excised heterogeneous tissue due to its different types (white, hearts of 30 patients undergoing cardiac transplant sur- brown, and “brite”) and its different body localizations gery at the Hospital de Sant Pau i la Santa Creu (subcutaneous, visceral, epicardial, and perivascular, etc.). (Table 1). Only samples from 3 patients were excluded, Composition, structure, and function of adipose tissue are 2 of them due to internal cell contamination and the highly dependent on individual metabolic factors that we other 1 due to endocarditis. Tissue was obtained with in- are still learning. Ectopic AT covering different organs formed consent of patients. The protocol was approved may have different anatomic and functional characteristics by the Research Ethics Committee of our hospital and and due to its proximity to various internal organs may was conducted in accordance with the Declaration of exert organ-specific regulatory functions [13]. Helsinki. Patients until the moment of transplant surgery AT that directly surrounds the heart is known as epi- were treated as recommended by clinical practice guide- cardial adipose tissue (EAT). EAT, a thoracic fat depot lines. Left anterior descending (LAD) coronary arteries that exists on the surface of the myocardium and is con- were collected and PVAT isolation. Arteries were fixed tained entirely beneath the pericardium, has acquired in- in 4% paraformaldehyde and stored at − 80 °C embedded creasing interest because of its proximity to the in OCT until histological analysis. myocardium and coronary arteries and their branches [12]. EAT has been reported to exert an endocrine role Table 1 Clinical characteristics of the study patients by deregulation of adipokine secretion [14, 15] and to Variables Patients have direct effects on local inflammation and coronary N 30 atherosclerosis [16–18]. EAT is also a source of ASCs, Age (years) 54 ± 3 and it has been reported that these cells may have a Sex (M/W) 19/11 higher cardiomyogenic potential as compared to pericar- 2 dial and omental ASCs subtypes [19]. BMI (kg/m ) 27.4 ± 1.0 ASCs not only have interest for their capacity to differ- Obesity (%) 33.3 entiate into differentiated cells, as cardiomyocytes, but also Overweigh (%) 20 for their capacity to stimulate angiogenesis promoting Preserved FE (%) 20 progenitor cell differentiation and paracrine proangiogenic CVRFs 2±1 and immunomodulatory effects [20]. The proangiogenic Smoking (%)* 23 effects of conditioned medium derived from ASCs support this paracrine view of stem cell function [4]. Previous HTA (%)** 47 studies indicated that microvesicles (MVs) released by DM (%)** 17 ASCs mediate their proangiogenic effects by cell-to-cell DLP (%)** 57 communication stimulating formation and stabilization of Values are expressed as mean ± SD or as percentages, when indicated vessel formation [21]. However, little is known about epi- CVRFs cardiovascular risk factors, HTA hypertension, DM diabetes mellitus, DLP dyslipidemia cardial ASCs, in particular, those located closely to the *Five years ex-smokers were considered non-smokers vessels [AT around vessels is known as perivascular AT **Treated as per guidelines Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 3 of 12 Simultaneously, epicardial adipose tissue directly overly- cells were characterized as stem cells by using the following ing the ventricular myocardium at the apex of the heart criteria: adherence to plastic, cell surface antigen phenotyp- (VMAT) was collected. All studies were run in parallel ing, and differentiation into multiple cell lineages. All ana- with samples from both locations from the same patients. lyses were performed between passages 3 and 4. A Both VMAT and PVAT were carefully dissected and fro- simplified scheme of the procedure is shown in Fig. 1. zen in liquid N2 and kept at − 80 °C until used or placed For cell cytometry characterization, cell surface antigen in tubes with 10 mL of Dulbecco’s modified Eagle’s phenotyping was performed on SVF and ASCs obtained medium (DMEM; Gibco, Life Technologies, Grand Island, from SVF at passage 3 (P3). The following cell surface epi- NY, USA) supplemented with 1% penicillin/streptomycin topes were marked with anti-human antibodies: CD105, (P/S, Gibco, Life Technologies, Grand Island, NY, USA) to CD44, CD29, CD90, CD73, CD45, and CD14. Cells (1 × be immediately processed to isolate ASCs. 105) at passage 3 or 1 mL of the SVF was suspended in flow cytometry buffer (PBS, 0.1% BSA, 0.1% sodium azide) ASC isolation and characterization and incubated for 30 min at 4 °C with the corresponding PVAT and VMAT tissues were washed with sterile antibodies. After that, the reaction was stopped by adding phosphate-buffered saline (PBS) supplemented with 1% of 500 μL of flow cytometry buffer or 250 μL of Quicklysis P/S. Tissue was digested into a type I collagenase solution reagent (Cytognos) in the case of the SVF. Quicklysis was (1 mg/mL; Sigma-Aldrich, St. Louis, MO, USA) and incu- incubated for 15 min at room temperature to eliminate bated for 1 h in a 37 °C pre-warmed orbital shaker. Collage- erythrocytes, and the reaction was stopped by adding nase activity was neutralized with the same amount of fetal 250 μL of flow cytometry buffer. bovine serum (FBS; Biological Industries, Kibbutz Beit- Cellular events (at least 30,000 in the case of ASCs Haemek, Israel) and the suspension filtered through a 100- and between 10,000 and 60,000 in the case of the SVF) μm strainer to eliminate remaining tissue fragments. Then, were acquired and analyzed by fluorescence-activated the suspension was centrifuged at 1200 rpm for 10 min to cell sorting using Coulter EPICS XL flow cytometer separate the adipocytes and to obtain the stromal vascular (Beckman Coulter) running Expo32 ADC XL 4 color fraction (SVF). Isolated SVF cells were counted and sepa- software (Beckman Coulter). rated into two aliquots to be analyzed by flow cytometry Adipogenic and osteogenic differentiation was also eval- and to be plated onto 25-cm2 culture flasks. After 24 h, uated for further characterization as previously described non-adherent cells were removed and the medium re- [11]. Briefly, cells were seeded in 6-well and 24-well plates placed. Cells were expanded in a humidified environment for the differentiation assay and for RNA isolation or cell at 37 °C with 1% O2 and 5% CO2 and maintained at sub- staining, respectively. Cells were allowed to grow to sub- confluent levels prior to phenotypic profile analysis. The confluence (approximately 80% of confluence) and then Fig. 1 Study design. Ventricular myocardium adipose tissue (VMAT) and perivascular adipose tissue (PVAT) were obtained during heart transplant surgery. ASCs were immediately isolated and phenotypically characterized by flow cytometry Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 4 of 12 allowed to differentiate for 21 days with the corresponding collected and centrifuged at 1200 rpm for 10 min to re- conditional medium or normal medium without supple- move cellular debris and kept at − 80 °C until used. ments as a negative control (Additional file 1: Table S1). After 21 days, RNA from the differentiated cells was iso- Microvesicle isolation lated, and also, cells were stained with Oil Red to detect ASC-derived microvesicles (MVs) were isolated by ultra- adipogenic differentiation and Alizarin red to detect osteo- centrifugation of P3 cell supernatants as previously de- genic differentiation. scribed [23]. Briefly, fresh supernatants were firstly centrifuged at 900g for 15 min to eliminate cell debris and Growth kinetics of ASCs then at 20,000g for 45 min to isolate the MVs as a pellet. To determine the growth kinetics of ASCs, 2 × 104 cells MV concentration was determined by flow cytometry. For per well were seeded in ten wells of 6-well plates. Cells that, MVs were washed with a PBS-citrate buffer and cen- from two duplicate wells were harvested and counted trifuged again at 20,000g for 30 min. MVs were extracted every other day for 10 days (cells were counted on days with PBS-citrate buffer and incubated with Annexin V 2, 4, 6, and 10). ASC number was plotted against the (CF Blue ANXV, Immunostep) and anti-TF antibody number of days cultured. Population doubling time (FITC-conjugated 4508CJ, Sekisui). Samples were then di- (PDT) was calculated using the formula PDT = Te/[(log luted with Annexin V binding buffer (BD Bioscience) to N2 − log N1)/log2], where Te is time (days) in the expo- stop the reaction and then analyzed on a FACSCantoll™ nential growth phase, N1 is the number of cells at the flow cytometer (BD Bioscience). The formation of MVs beginning, and N2 is the number of cells at the end of after cell activation is initiated by the increase of intracel- the experiment. lular calcium resulting in an elementary rearrangement of the phospholipid asymmetry with translocation of phos- MTS viability/proliferation analysis phatidylserine from the inner to the outer surface leaflet Cell proliferation was determined by 3-(4,5-dimethylthiazol- of the plasma membrane as a consequence of activation of 2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfopheny)-2H- scramblase and floppase/ABC1 and inhibition of translo- tetrazolium (MTS) assay (CellTiter 96 Aqueous One Solu- case/flippase activities. The presence of phosphatidylserine tion cell proliferation assay kit; Promega). For this assay, at the surface of the MP membrane enables the use of the 15 × 103 of ASCs or 1 × 104 human microvascular endothe- Annexin V for MVs detection by flow cytometry. MVs’ lial cells (HMEC-1) were seeded in triplicates into 96-well gate limits were established following two criteria: (1) cali- plates. ASCs were cultured for 24 h and then treated with bration using a Flow-Check Size Range Calibration Kit 10 μL of MTS. HMEC-1 were cultured for 24 h then treated (Polysciences) and (2) using an in vitro platelet-derived with ASC-derived MVs for other 24 h, and finally, 10 μL of microparticle population as a positive control. The lower MTS per well was added and incubated for an additional 2 detection limit was placed as a threshold above the elec- h, while MTS tetrazolium is reduced to formazan (490 nm tronic noise of our flow cytometer. absorbance) by the metabolically active cells. The absorb- ance was then quantified with the spectrophotometer Spec- Cell migration tramax 250 and analyzed with the SoftMax software HMEC-1 were used for the cell migration assays. Briefly, (Molecular Devices). Formazan production was directly re- 2.3 × 104 eASCs were seeded into 100-mm dish and cul- lated with the number of cells alive in the culture. tured with MCDB 131 medium supplemented with 10% of FBS for 48 h to allow cells to secrete MVs. The day Tube-like formation assay after, 2.5 × 105 HMEC-1 cells were seeded into a Tube-like formation assay in three-dimensional basement Culture-Insert 2 well dish (Idibi) and kept with MCDB membrane culture was employed to assess the angiogenic 131 supplemented with 10% of FBS overnight. MVs from potential of ASCs. Briefly, 1 × 105 cells were seeded in a the eASC supernatant (eMVs) were isolated, and before Matrigel-coated plate and incubated with endothelial cell performing the experiment, the insert was removed by growth medium (EGM-2 BulletKit, Lonza, USA) for 24 h using sterile tweezers and the dish washed with PBS to to allow the formation of tube-like structures. The total remove cell debris. Cells were treated with 600 μL of (A) tube-covered area was quantified using ImageJ software conditional medium from ASCs after 48 h of culture, (B) (National Institutes of Health, USA). conditional medium from ASCs without eMVs after 48 h of culture, and (C) eMV-enriched medium after 48 of ASC-conditioned medium collection culture. In all conditions, medium was supplemented Cells at passage 3 or 4 were allowed to grow until reach- with 2% of FBS. Cell migration and wound repair were ing sub-confluence, then cells were washed exhaustively controlled every 2 h for 10 h. Wound areas were ana- with PBS to remove FBS and serum-free medium was lyzed using ImageJ software. Protein, RNA, and micro- added. After 48 h, the ASC-conditioned medium was RNA were isolated from the ASCs; microRNA from the Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 5 of 12 eMVs and RNA; and microRNA from the HMEC-1 cells The level of significance was set at P < 0.05. All analyses after 24 h of cell migration. were conducted with StatView software. Gene expression analysis Results Total RNA was isolated from ASCs in silica membrane Sample characterization columns with the Qiagen RNesasy Mini Kit (Qiagen) ac- Heart adipose tissue and LAD coronary arteries were cording to the manufacturer’s instructions. collected at the operating room during heart transplant- MirVana miRNA isolation kit (Life Technologies) was ation surgery. LAD coronary arteries were classified ac- used to extract microRNA from the cells, and miRNeasy cording to the AHA classification from intimal Serum/Plasma Kit for the microRNA isolation from thickening (IT) to type VI and total occlusion (TO) MVs, according to the manufacturer’s instruction. (Additional file 1: Figure S2A-H). ASCs were isolated RNA and microRNA quantity was determined with immediately after tissue dissection, harvested, and char- Nanodrop ND-1000 spectrophotometer (Nanodrop Tech- acterized by their adherence to plastic, their ability to nologies). Isolated total RNA was reverse-transcribed into differentiate into multiple cell lineages, and the expres- cDNA using the High Capacity cDNA Archive kit (Ap- sion of different cell surface markers. ASCs were positive plied Biosystems, Foster City, CA, USA) and microRNA to CD29, CD44, CD73, CD90, and CD105, and charac- with the TaqMan advanced miRNA assay (Life Technolo- teristic mesenchymal stem cell surface markers, and gies). Gene expression analysis was carried out by quanti- negative for the hematopoietic marker CD45 and the tative PCR using TaqMan® Gene Expression assays monocyte marker CD14 (Fig. 2a). These results show (Applied Biosystems; Additional file 1: Table S2) and the that the adipose tissue covering the human heart is a re- Applied Biosystems Prism 7900HT Sequence Detection pository for stem cells. System (Applied Biosystems) according to manufacturer’s instructions. Gene expression data are expressed as target Proliferation and viability of ASCs from the different gene mRNA expression relative to the correspondent adipose tissue depots housekeeping gene expression. The proliferative rate of ASCs obtained from VMAT (mASCs) was always higher than that of ASCs obtained Western blot analysis from PVAT (pASCs) (Fig. 2b); however, the main change Protein was extracted from total cell lysates by using RIPA was observed during the first 4 days; thus, the calculated buffer (50 mM Tris-HCl pH 8, 150 mM NaCl, 1% NP-40, doubling time of mASCs and pASCs was almost the same 0.5% sodium deoxycholate, 0.1% SDS) or from 48 h cell during days 6 and 8 (Fig. 2c). Interestingly, the cellular via- supernatant. Protein concentrations were measured with the bility MTS assay demonstrated a significantly higher via- Pierce BCA Protein Assay Kit (ThermoScientific). Twenty- bility in mASCs than in pASCs (P = 0.001; Fig. 2d) after five micrograms of protein was resolved by 1-DE under 24 h in culture. reducing conditions onto 10% SDS-PAGE gels and electro- transferred to nitrocellulose membranes. After blocking for In vitro angiogenic-related gene expression of apex EAT- non-specific binding with 5% of bovine serum albumin and PVAT-derived ASCs (BSA; MP Biomedical) or Blotto, the membranes were incu- In order to evaluate the angiogenic potential of mASCs bated with primary antibodies, including TF (4501-Sekisui and pASCs, different angiogenic-related genes were Diagnostics), endostatin (ab109660-abcam), and VEGF measured. pASCs showed a significantly higher mRNA (ab51745-abcam). Band detection was performed using a expression of VEGFA compared with apex-derived ASCs chemiluminiscent substrate dye (SuperSignal® West Dura Ex- (N = 15; P = 0.04; Fig. 3a) and a significantly reduced tended Duration Substrate, Thermo Scientific, Waltham, mRNA level of COL18a1 (N = 15; P = 0.02; Fig. 3b). In- MA, USA) and a molecular imager ChemiDoc XRS System, stead, pASCs and mASCs showed similar mRNA expres- Universal Hood II (BioRad, Hercules, CA, USA). Band quan- sion levels of TF (N = 12; P = 0.24; Fig. 3c), FGF2 (N = 6; tification was performed with Image Lab 4.0 software P = 0.46; Fig. 3d), and TIE2 (N = 7; P = 0.31; Fig. 3e). (BioRad Laboratories, Hercules, CA, USA). Protein load was Protein levels of TF and endostatin were also measured normalized with total protein staining, and normalization be- by western blot, and a non-significant higher level of TF tween different membranes was performed with a common on pASC compared to mASC (P = 0.093; Additional file 1: pool in every gel, as previously described [24]. Figure S2A) was observed. Endostatin levels were mea- sured both in cells and secretome, but no differences were Statistical analysis observed between tissues (Additional file 1: Figure S2B-C). Non-parametric Wilcoxon or Mann-Whitney analyses The ability of both ASCs to form tube-like structures were performed to analyze the differences between tis- was investigated. Both mASCs and pASCs had a similar sues. Data are expressed as mean ± SEM unless stated. capacity to induce tube-like formation in Matrigel cultures Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 6 of 12 Fig. 2 ASC characterization. a Flow cytometry analysis. Surface mesenchymal stem cell markers were analyzed from the stromal vascular fraction (SVF) and from cells at passage 3. Data is presented as percentage of positive cells. b Growth curves of mASC and pASV. Cells were counted every day after and plotted in a line curve graph. c Population doubling time (PDT) of mASC and pASC measured every 2 6 for 10 days. d MTS viability assay (**P < 0.01) (Additional file 1: Figure S3A and B, N = 9; P = 0.21). The and released in the microvesicles. Next, we analyzed ability to form tubes by pASCs derived from LADs whether these MVs were also positive for TF, because with mild atherosclerosis (N = 4; IT to grade III) and previous studies in our group have described that TF- pASCs derived from LAD with severe atherosclerosis rich microvesicles increase angiogenesis in ischemic tis- (N = 5; grade IV occlusion to TO) was not signifi- sue [25]. The mMVs from mASCs of non-obese and cantly different (Additional file 1: Figure S3C). non-diabetic patients (total or Annexin V positive) con- tain significantly higher TF amount than those mASCs Effect of obesity and diabetes on eASCs released from obese non-diabetic patients (P = 0.004 and P = 0.02, microvesicle function respectively; Fig. 4a). mMVs from mASCs of obese non- Because of the impact of obesity in adipose tissue accu- diabetic patients also contain less TF than mMVs from mulation and fat characteristics, we investigated whether mASCs of obese diabetic patients (P = 0.02; Fig. 4a). mASC-derived microvesicles (mMVs) could exert a We then analyzed the microRNA (miRNA) compos- regulatory effect on human microvascular endothelial ition of those eMVs. We analyzed two different miRNAs cell proliferation. First, we observed that mASCs from that have been implicated in ASCs and the angiogenic obese non-diabetic patients release less mMVs than process as miRNA 126-3p and miRNA 145-5p [26, 27]. those ASCs from obese diabetic patients (P = 0.06; We found that mMVs from mASCs of non-obese non- Fig. 4a). Interestingly, obese and diabetic mASC-released diabetic patients had a significantly higher content of mMVs have significantly higher levels of Annexin V miRNA 126-3p than those mMVs from mASCs of obese than those released by non-obese non-diabetic patients diabetic patients (P = 0.05; Fig. 4b). Non-significant (P = 0.03; Fig. 4a). Annexin V binds to phosphatidylser- change was observed on miRNA 145-5p (Fig. 4c). ine that is usually exposed in the outer leaflet of the To evidence the effects of mASC-secreted mMVs on plasma membrane of cells that are activated or apoptotic human microvascular endothelial cell function, we first Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 7 of 12 Fig. 3 Boxplot diagram showing ASC genomic profile. a Differential protein expression of VEGFA in mASC and pASC. b Differential protein expression of COL18a1 in mASC and pASC. c Differential protein expression of TF in mASC and pASC. d Differential protein expression of FGF2 in mASC and pASC. e Differential protein expression of TIE2 in mASC and pASC (*P < 0.05; **P < 0.01) performed a MTS viability assay and no significant effects CM (P = 0.01) and mMV-free medium (P = 0.01), imply- on cell viability after treatment with mMVs was observed; ing effects independent of MVs but associated with sol- however, HMEC cells treated with mMVs from non-obese uble factors. non-diabetic patients showed a non-significant improve- These results indicate that obesity and diabetes impact ment on HMEC viability (Additional file 1: Figure S4A). on the pro-angiogenic capacity of resident stem cells The effect of mASC-secreted mMVs on HMEC-1 mi- and explain the low spontaneous regenerative efficacy of gration showed that, after 10 h of incubation, those stem cells in patients with metabolic syndromes. wounds treated with mMV-free secretome were signifi- cantly less able to repair the wounded area compared with Discussion those treated with mMV-rich medium (P = 0.05; Fig. 4d), The EAT has emerged as a depot with a distinctive bio- with the highest effect seen when HMEC-1 cells were logical significance in cardiac function. Moreover, it is a treated with mMVs released by mASC from obese dia- reservoir of ASCs, with yet undefined effects on the po- betic patients (Additional file 1: Figure S4B-D). tential regeneration of ischemic tissue. Here, we show To evidence the effect of mMVs on receptor HMEC-1 that ASCs contained in different EAT locations obtained cells, we analyzed the mRNAs and the miRNAs expres- from patients with advanced heart failure requiring sions in the receptor cells that had been incubated with transplant have distinct functional properties. the supernatants of EAT-ASCs. We observed no change The relationship of the adipose tissue with CVD has on TF expression (Fig. 5a) nor in miRNA 126-3p (Fig. 5c) been widely studied [28, 29]. Around 80% of the heart is and miRNA 145-5p (Fig. 5d), but a significant increase covered by visceral ectopic adipose tissue depots, which in on VEGFA (Fig. 5b) expression induced by both total lean subjects is considered cardioprotective because of the Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 8 of 12 Fig. 4 mASC microvesicle (mMV) secretion. a–c Influence of obesity and diabetes on eASC-derived MVs. a Number of mMVs measured by flow cytometry (total, Annexin V-positive, TF-positive, and Annexin V/TF-positive mMVs). b miR126 relative gene expression. c miR145 relative gene expression. d Healing rate line diagram of HMEC-1 cells treated with mASC total medium, mMV-depleted medium, or eMV-rich medium (*P < 0.05; **P < 0.01) Fig. 5 Effect of mASC-derived MVs on HMEC-1. Genomic changes of HMEC-1 cells produced by the treatment with mASC total medium, mMV- depleted medium, or mMV-rich medium. a HMEC-1 VEGFA gene expression. b HMEC-1 TF gene expression. c HMEC-1 miR126 gene expression. d HMEC-1 miR145 gene expression (*P < 0.05) Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 9 of 12 secretion of anti-inflammatory and anti-atherogenic adi- genes in angiogenesis is well established [39–41]; in fact, pokines. However, excess and thickness of fat around the FGF-2 induces the formation of new capillary-like tu- heart, usually measured by imaging techniques, have been bules on HMEC cells [42]. However, in line with our re- associated with CVD [16, 30]. Although imaging has sults, in previous studies, no differences were observed brought a growing interest in the presence of heart EAT, when comparing the expression of FGF-2 in ASC from it has not been possible to functionally study this adipose different sources, but changes on these proteins were tissue as widely as it has been done with other more ac- obtained as a result of the age of the donor [43]. TIE-2 cessible depots [31]. is related with the ANGPT system [44]; thus, a deep The ectopic visceral adipose tissue covering the heart study on this pathway is needed for the better under- can be sub-classified by location and as perivascular fat standing of this new source of epicardial ASC. and fat directly overlying the myocardium, with blood Few studies have analyzed the presence and function supplies directly derived from the coronary arteries and of heart adipose tissue-derived stem cells, neither in ani- their branches [32, 33]. The function of the stem cells mal [45, 46] nor in human [19] models, and in none of found in these adipose tissue depots has not been stud- them, their angiogenic potential has been analyzed and ied in-depth. This study has focused on the identifica- compared. Previous studies in our group have shown tion, isolation, and comparison, for the first time, of the how ASCs derived from the subcutaneous adipose tissue, ASCs covering the LAD (PVAT) and the ASCs covering independently of the BMI, have no ability to form tubes the ventricular myocardium at the apex (VMAT) of in a Matrigel-based experiment [11], but here, we dem- hearts explanted during heart transplant operations. onstrate for the first time that both mASCs and pASCs, These ASCs were characterized by their adherence to when seeded in a 3D Matrigel surface and cultured with plastic as well as by the expression of different stem cell EGM-2 medium, are able to form capillary-like ring markers (CD90, CD105, CD44, CD29, and CD73) and structures, independently of cardiac fat location and of the absence of the leukocyte marker CD45 and the in- atherosclerosis severity in the LAD indicating a proan- nate immune system marker CD14. Thus, both fat loca- giogenic function. More studies are needed to analyze tions have a significant amount of resident stem cells. the influence of different co-morbidities on heart adi- We have observed that mASCs have a higher prolifera- pose tissue and its derived ASC. tive rate and growth kinetics than pASC, which might Angiogenesis is mediated not only by VEGFA [47] and suggest that mASCs are cells with an active participation COL18A1 [48], but also by many other genes, such as in local regenerative activity and homeostasis. Due to fibroblast growth factor (FGF), angiopoietin I and II the angiogenic activity of adipose tissue [34], we investi- (ANGPT I-II) [41], and NOTCH [49] among others, and gated the genomic angiogenic profile of these cells. Even target cells include endothelial cells (EC) and vascular though there are many proteins involved in the angio- smooth muscle cells (VSMCs) [50, 51] were involved. genic process, VEGF is undoubtedly the most important In fact, a coordinated reaction between different cellu- angiogenic factor. It is a potent mitogen for endothelial lar types and metabolic factors is needed for the correct cells (ECs), involved both in the development of the vas- development of the angiogenic process [52]. It would cular system and the induction and maintenance of the also be of great interest to analyze the adipokines se- neovasculature in tumors. VEGF binds to any of the creted by these EATs and resident ASCs to further tyrosine kinase receptors (VEGFR1 or VEGFR2), which characterize both epicardial heart depots and their po- induce the phosphorylation of the tyrosine residues, thus tential implication on myocardial self-regeneration. activating intracellular signaling pathways involved in Spontaneous regeneration of damaged tissue may be ac- EC proliferation, migration, and survival [35, 36]. On the complished not only by ASC differentiation and also by other hand, COL18A1 is a gene that encodes the anti- ASC-released microvesicles (MVs) and exosomes [53, 54]. angiogenic protein endostatin in its C-terminal domain MVs are small particles with a diameter ranging from 0.10 [37]. Endostatin inhibits EC proliferation and migration to 1.00 μm released by many cell types, including ASCs and induces EC apoptosis by blocking the binding of [55]. The potential of MVs released by ASCs to improve VEGF with the tyrosine receptor KDR (VEGFR-2) [38]. HUVEC migration, proliferation, and angiogenesis has We found that pASCs express a higher amount of been reported [56, 57]. Many studies support the concept VEGFA, as well as a reduced expression of COL18A1, of a proangiogenic effect of MVs depends on their com- compared with the mASCs, suggesting a coordinated position, conferring to them a dynamic storage pool of regulation of angiogenesis through these two proteins by bioactive molecules that play essential roles in cell-cell EAT-derived ASCs, which is increased on pASC, in dir- crosstalk. In this way, MVs can provide intercellular com- ect contact with the LAD compared with mASC. By munication by the delivery of miRNAs to influence tran- contrast, no differences by location were achieved in scription and altering genetic processes. miRNA126-3p is other genes as TF, TIE2, and FGF-2. The role of these one of the more important miRNAs in angiogenesis [58], Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 10 of 12 and it has been described to be present in MVs. In line of mASC-derived microvesicles on endothelial cell migra- with these results, we have also demonstrated that heart tion that could represent a promising strategy to stimulate mASCs from non-obese and non-diabetic heart failure pa- spontaneous regeneration and repair in rarefaction areas of tients release mMVs, rich in miRNA126-3p, that improve the damaged myocardium. Tissue activation and microvascular endothelial cells (HMEC-1) viability and mobilization of resident ASCs could be a potential means are more effective than mASC-conditioned medium to reduce the progression of disease in failing hearts. (CM), or mMV-depleted medium in inducing angiogen- esis. mMVs also increase HMEC-1 migration rate, inde- pendently on the presence of CVRFs, and HMEC-1 Supplementary information Supplementary information accompanies this paper at https://doi.org/10. wounds in vitro are closed faster when treated with mMVs 1186/s13287-019-1460-1. from obese diabetic patients. In this regard, it has been re- ported that MVs released from subcutaneous ASCs of Additional file 1: Figure S1. A-H) LAD coronary arteries classification. obese subjects have impaired pro-angiogenic potential and According to the American heart association, LAD coronaries were classi- fied from intimal thickening (IT) to type VI and total occlusion (TO). Fig- harbor reduced VEGF, MMP-2, and miR-126 content ure S2. Change in protein levels measured by western blot. A) Tissue than those released by ASCs of non-obese subjects, redu- factor protein expression in mASC compared with pASC. B) Endostatin cing their capability to promote EC migration and tube- protein expression in mASC compared with pASC. C) Endostatin protein expression in mASC supernatant compared with pASC supernatant. Fig- like structure formation [26]. In addition, previously, we ure S3. Tube-like formation assay. A) Representative image of 3D- have found that TF-rich MVs increase neovascularization matrigel-tube formation in mASC and pASC. B) Box plot comparing the in ischemic tissue [23], and here, we observed that mASCs tube length from the mASC and the pASC. C) Bar chart comparing the tube formation potential of pASC from non-occluded LAD coronaries and have high levels of TF that may induce the angiogenic from occulted LAD coronaries. Figure S4. Influence of different CRF on properties of these MVs and are decreased in obese and mMVs proliferation potential. A) MTS viability assay. Change on HMEC-1 diabetic patients. HMEC-1 treated with eASC supernatant viability when treated with mMVs from patients with different number of CVRFs. B) Healing rate line diagram of HMEC-1 cells treated with mASC (CM, MV-free medium or MV-rich medium) increases total medium, mMVs depleted medium or mMVs rich medium from non- the expression of VEGFA, suggesting a potential way of obese and non-diabetic patients. C) Healing rate line diagram of HMEC-1 action of ASC extracellular particles (not only MVs, but cells treated with mASC total medium, mMVs depleted medium or mMVs rich medium from obese and non-diabetic patients. D) Healing rate line also exosomes and other secreted factors). TF induces diagram of HMEC-1 cells treated with mASC total medium, mMVs de- angiogenesis by ETS1 transcription factor [59], and VEGF pleted medium or mMVs rich medium from obese and diabetic patients is a target gene of ETS1 [60]. However, more experiments (*P=0.05). Table S1. Adipogenic and osteogenic conditional medium composition. Table S2. Taqman gene expression assays. on this field should be carried out to elucidate the precise mechanism of action of each different extracellular par- ticle depending on ASC origin. Abbreviations ASCs: Adipose stem cells; CM: Conditional medium; CVD: Cardiovascular disease; CVRF: Cardiovascular risk factor; eASCs: Epicardial adipose stem cells; Limitations EAT: Epicardial adipose tissue; HMEC-1: Human microvascular endothelial As a limitation, we have to report the low number of cells-1; LAD: Left anterior descendent; MVs: Microvesicles; mASCs: Ventricular cases in this study, but the number of heart transplants myocardium adipose stem cells; mMVS: mASC-derived microvesicles; pASCs: Perivascular adipose stem cells; PVAT: Perivascular adipose tissue; in one single hospital is low. Organ donations are dis- SVF: Stromal vascular fraction; VMAT: Ventricular myocardium adipose tissue; tributed among patient waiting lists in many different VSMC: Vascular smooth muscle cell hospitals. The methods and technologies reported here have to be performed immediately with hearts directly Acknowledgements obtained by research investigators waiting at the surgical The technical support of Olaya García is gratefully acknowledged. The authors would like to thank the heart failure service of the Sant Pau i la room to receive the excised heart. At the annex of the Santa Creu Hospital for providing the samples and clinical data. The authors operating room, the hearts are sampled and taken from declare no conflict of interest. the operating room within minutes in the appropriated buffers to the cell biology laboratory for further process- Authors’ contributions ing. Therefore, only in-hospital transplants could be CL designed and conducted the research, collected the samples, analyzed used. Many other genes and pathways would be analyzed the data, performed the statistical analysis, and wrote the manuscript. GA designed and supervised the research, analyzed the data, performed the if more hearts were available. statistical analysis, and wrote the manuscript. MTB designed and conducted the research, collected the samples, analyzed the data, performed the Conclusions statistical analysis, and wrote the manuscript. JC performed, analyzed, and interpreted the flow cytometry experiment. MB designed and conducted the In the present study, we demonstrate for the first time that research, collected the samples, and wrote the manuscript. OJB designed hearts from patients with advanced heart failure requiring and conducted the research, collected the samples, performed the transplant have ASCs in their different heart adipose tissue histological analysis, and wrote the manuscript. LB designed and supervised the research, analyzed the data, performed the statistical analysis, and wrote depots and that these ASCs have distinct functional proper- and revised the manuscript. All authors had primary responsibility for the ties. Additionally, we have demonstrated the potential role final content and approved the final version of the manuscript. Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 11 of 12 Funding 13. Calabuig Á, Barba J, Guembe MJ, Díez J, Berjón J, Martínez-Vila E, et al. Grasa This work was supported by grants from the Spanish Ministry of Science, epicárdica en la población general de mediana edad y su asociación con el Innovation and Universities [SAF2016-76819-R to L.B.]; Institute of Health síndrome metabólico. Rev Española Cardiol. 2017;70:254–60 Available from: Carlos III, ISCIII TERCEL RD16/00110018, CB16/11/0041 to LB, and P17/01321 http://linkinghub.elsevier.com/retrieve/pii/S0300893216303451. to GA; FEDER “Una Manera de Hacer Europa”; the Secretary of University and 14. Karastergiou K, Evans I, Ogston N, Miheisi N, Nair D, Kaski JC, et al. Epicardial Research, Department of Economy and Knowledge of the Government of adipokines in obesity and coronary artery disease induce atherogenic Catalonia [2017SGR1480]; and “CERCA Programme/Generalitat de Catalunya” changes in monocytes and endothelial cells. Arterioscler Thromb Vasc Biol. Spain. CL is the recipient of a pre-doctoral grant from the Spanish Ministry of 2010;30:1340–6. Economy and Competitiveness of Science [BES-2014-068269]. We thank FIC- 15. Baker AR, da Silva NF, Quinn DW, Harte AL, Pagano D, Bonser RS, et al. Human Fundacion Jesús Serra, Barcelona, Spain, for their continuous support. epicardial adipose tissue expresses a pathogenic profile of adipocytokines in patients with cardiovascular disease. Cardiovasc Diabetol. 2006;5. Availability of data and materials 16. Patel VB, Shah S, Verma S, Oudit GY. Epicardial adipose tissue as a metabolic All data generated and materials are included in the manuscript. transducer: role in heart failure and coronary artery disease. Heart Fail Rev Heart Failure Reviews. 2017;22:889–902. Ethics approval and consent to participate 17. Gruzdeva O, Borodkina D, Uchasova E, Dyleva Y, Barbarash O. Localization of Tissue was obtained with informed consent from patients. The protocol was fat depots and cardiovascular risk. Lipids Health Dis. 2018;17:218 Available approved by the Research Ethics Committee of the Sant Pau i la Santa Creu from: https://lipidworld.biomedcentral.com/articles/10.1186/s12944-018-0856-8. Hospital and was conducted in accordance with the Declaration of Helsinki. 18. Lazaros G, Antonopoulos A, Antoniades C, Tousoulis D. The role of epicardial fat in pericardial diseases. Curr Cardiol Rep. 2018;20:40–7. Consent for publication 19. Wystrychowski W, Patlolla B, Zhuge Y, Neofytou E, Robbins RC, Beygui RE. Not applicable Multipotency and cardiomyogenic potential of human adipose-derived stem cells from epicardium, pericardium, and omentum. Stem Cell Res Ther. 2016;7: Competing interests 84–96. The authors declare that they have no competing interests. 20. Rubina K, Kalinina N, Efimenko A, Lopatina T, Melikhova V, Tsokolaeva Z, et al. Adipose stromal cells stimulate angiogenesis via promoting Author details progenitor cell differentiation, secretion of angiogenic factors, and 1 Cardiovascular-Program ICCC, IR-Hospital Santa Creu I Sant Pau, IIB Sant Pau, enhancing vessel maturation. Tissue Eng - Part A. 2009;15:2039–50. C/Sant Antoni Ma Claret 167, 08025 Barcelona, Spain. 2Ciber CV, 28029 21. Lopatina T, Mazzeo A, Bruno S, Tetta C, Kalinina N, Romagnoli R, et al. The Madrid, Spain. 3Cardiovascular Research Chair UAB, Barcelona, Spain. angiogenic potential of adipose mesenchymal stem cell-derived extracellular vesicles is modulated by basic fibroblast growth factor. J Stem Received: 29 July 2019 Revised: 16 October 2019 Cell Res Ther. 2014;4:26–38. Accepted: 17 October 2019 22. Carmeliet P. Angiogenesis in life, disease and medicine. Nature. 2005; 438(7070):932–6. References 23. Arderiu G, Peña E, Badimon L. Angiogenic microvascular endothelial cells 1. Orlic D. Adult bone marrow stem cells regenerate myocardium in ischemic release microparticles rich in tissue factor that promotes postischemic heart disease. Ann N Y Acad Sci. 2003;996:152–7. collateral vessel formation. Arterioscler Thromb Vasc Biol. 2015;35:348–57. 2. Kawamoto A, Tkebuchava T, Yamaguchi JI, Nishimura H, Yoon YS, Milliken C, 24. Cubedo J, Padro T, Badimon L. Glycoproteome of human apolipoprotein A- et al. Intramyocardial transplantation of autologous endothelial progenitor I: N- and O-glycosylated forms are increased in patients with acute cells for therapeutic neovascularization of myocardial ischemia. Circulation. myocardial infarction. Transl Res. 2014;164:209–22. 2003;107:461–8. 25. Arderiu G, Peña E, Aledo R, Juan-Babot O, Badimon L. Tissue factor regulates 3. Asahara T, Isner JM. Endothelial progenitor cells for vascular regeneration. J. microvessel formation and stabilization by induction of chemokine (C-C Hematotherapy Stem Cell Res. 2002;11(2):171–8. motif) ligand 2 expression. Arterioscler Thromb Vasc Biol. 2011;31:2607–15. 4. Vilahur G, Oñate B, Cubedo J, Béjar MT, Arderiu G, Peña E, et al. Allogenic adipose- 26. Togliatto G, Dentelli P, Gili M, Gallo S, Deregibus C, Biglieri E, et al. Obesity derived stem cell therapy overcomes ischemia-induced microvessel rarefaction in reduces the pro-angiogenic potential of adipose tissue stem cell-derived the myocardium: systems biology study. Stem Cell Res Ther. 2017;8:1–15. extracellular vesicles (EVs) by impairing miR-126 content: impact on clinical 5. Planat-Bénard V, Menard C, André M, Puceat M, Perez A, Garcia-Verdugo JM, applications. Int J Obes. 2016;40:102–11. et al. Spontaneous cardiomyocyte differentiation from adipose tissue stroma 27. Arderiu G, Peña E, Aledo R, Juan-Babot O, Crespo J, Vilahur G, et al. cells. Circ Res. 2004;94:223–9. MicroRNA-145 regulates the differentiation of adipose stem cells toward 6. Cao Y, Sun Z, Liao L, Meng Y, Han Q, Zhao RC. Human adipose tissue-derived microvascular endothelial cells and promotes angiogenesis. Circ Res. stem cells differentiate into endothelial cells in vitro and improve postnatal 2019. neovascularization in vivo. Biochem Biophys Res Commun. 2005;332:370–9. 28. Mandviwala T, Khalid U, Deswal A. Obesity and cardiovascular disease: a risk 7. Miranville A, Heeschen C, Sengenès C, Curat CA, Busse R, Bouloumié A. factor or a risk marker? Curr Atheroscler Rep. 2016;18 Available from: https:// Improvement of postnatal neovascularization by human adipose tissue- doi.org/10.1007/s11883-016-0575-4. derived stem cells. Circulation. 2004;110:349–55. 29. Poirier P, Eckel RH. Obesity and cardiovascular disease. Curr. Atheroscler. 8. Dimmeler S, Leri A. Aging and disease as modifiers of efficacy of cell Rep. 2002;4(6):448–53. therapy. Circ. Res. 2008;102(11):1319–30. 30. Guglielmi V, Sbraccia P. Epicardial adipose tissue: at the heart of the obesity 9. Ferrer-Lorente R, Bejar MT, Tous M, Vilahur G, Badimon L. Systems biology complications. Acta Diabetol. 2017;54:805–12. approach to identify alterations in the stem cell reservoir of subcutaneous 31. Matloch Z, Kotulák T, Haluzík M. The role of epicardial adipose tissue in adipose tissue in a rat model of diabetes: effects on differentiation potential heart disease. Physiol Res. 2016;65:23–32. and function. Diabetologia. 2014;57:246–56. 32. Fitzgibbons TP, Czech MP. Epicardial and perivascular adipose tissues and 10. Oñate B, Vilahur G, Camino-López S, Díez-Caballero A, Ballesta-López C, their influence on cardiovascular disease: basic mechanisms and clinical Ybarra J, et al. Stem cells isolated from adipose tissue of obese patients associations. J Am Heart Assoc. 2014;3:1–15. show changes in their transcriptomic profile that indicate loss in 33. Lima-Martínez MM, Blandenier C, Iacobellis G. Epicardial adipose tissue: stemcellness and increased commitment to an adipocyte-like phenotype. more than a simple fat deposit? Endocrinol y Nutr. 2013;60:320–8. BMC Genomics. 2013;14:626–37. 34. Silverman KJ, Lund DP, Zetter BR, Lainey LL, Shahood JA, Freiman DG, et al. 11. Oñate B, Vilahur G, Ferrer-Lorente R, Ybarra J, Díez-Caballero A, Ballesta- Angiogenic activity of adipose tissue. Biochem Biophys Res Commun. 1988; López C, et al. The subcutaneous adipose tissue reservoir of functionally 153:347–52 Available from: http://www.sciencedirect.com/science/article/pii/ active stem cells is reduced in obese patients. FASEB J. 2012;26:4327–36. S0006291X88812294. 12. Talman AH, Psaltis PJ, Cameron JD, Meredith IT, Seneviratne SK, Wong DTL. 35. Melincovici CS, Boşca AB, Şuşman S, Mărginean M, Mihu C, Istrate M, et al. Epicardial adipose tissue: far more than a fat depot. Cardiovasc Diagn Ther. Vascular endothelial growth factor (VEGF) – key factor in normal and 2014;4:416–29. pathological angiogenesis. Rom. J. Morphol. Embryol. 2018;59(2):455–67. Lambert et al. Stem Cell Research & Therapy (2019) 10:361 Page 12 of 12 36. Shibuya M. Vascular endothelial growth factor and its receptor system: physiological functions in angiogenesis and pathological roles in various diseases. J. Biochem. 2013;153(1):13–9. 37. Ruge T, Carlsson AC, Larsson A, Ärnlöv J. Endostatin: a promising biomarker in the cardiovascular continuum? Biomark. Med. 2017;11(10):905–16. 38. Kim YM, Hwang S, Kim YM, Pyun BJ, Kim TY, Lee ST, et al. Endostatin blocks vascular endothelial growth factor-mediated signaling via direct interaction with KDR/Flk-1. J Biol Chem. 2002;277:27872–9. 39. Cross MJ, Claesson-Welsh L. FGF and VEGF function in angiogenesis: signalling pathways, biological responses and therapeutic inhibition. Trends Pharmacol. Sci. 2001;22(4):201–7. 40. Arderiu G, Peña E, Aledo R, Badimon L. Tissue factor-Akt signaling triggers microvessel formation. J Thromb Haemost. 2012;10:1895–905. 41. Huang H, Bhat A, Woodnutt G, Lappe R. Targeting the ANGPT-TIE2 pathway in malignancy. Nat Rev Cancer. Nature Publishing Group; 2010;10:575–585. Available from: https://doi.org/10.1038/nrc2894 42. Kroon ME, Koolwijk P, Van Der Vecht B, Van Hinsbergh VWM. Hypoxia in combination with FGF-2 induces tube formation by human microvascular endothelial cells in a fibrin matrix: involvement of at least two signal transduction pathways. J Cell Sci. 2001;114:825–33. 43. Jung J-S, Volk C, Marga C, Navarrete Santos A, Jung M, Rujescu D, et al. Adipose-derived stem/stromal cells recapitulate aging biomarkers and show reduced stem cell plasticity affecting their adipogenic differentiation capacity. Cell Reprogram. 2019;21:187–99. 44. Semenza GL. Vasculogenesis, angiogenesis, and arteriogenesis: mechanisms of blood vessel formation and remodeling. J Cell Biochem. 2007;102:840–7. 45. Bejar MT, Ferrer-Lorente R, Peña E, Badimon L. Inhibition of Notch rescues the angiogenic potential impaired by cardiovascular risk factors in epicardial adipose stem cells. FASEB J. 2016;30:2849–59. 46. Hamdi H, Planat-Benard V, Bel A, Puymirat E, Geha R, Pidial L, et al. Epicardial adipose stem cell sheets results in greater post-infarction survival than intramyocardial injections. Cardiovasc Res. 2011;91:483–91. 47. Taimeh Z, Loughran J, Birks EJ, Bolli R. Vascular endothelial growth factor in heart failure. Nat. Rev. Cardiol. 2013;10(9):519–30. 48. Ranjit PM, Anuradha C, Vishnupriya S, Girijasankar G, Girish K, Chowdary YA. Endogenous angiogenesis inhibitor endostatin: an overview. Asian J. Pharm. Clin. Res. 2012;5(3):1–8. 49. Siekmann AF, Lawson ND. Notch signalling and the regulation of angiogenesis. Cell Adhes Migr. 2007;1:104–6. 50. Risau W. Mechanisms of angiogenesis. Nature. 1997;386(6626):671–4. 51. Carmeliet P. Angiogenesis in health and disease. Nat Med. 2003;9:653–60. 52. Badimon L, Borrell M. Microvasculature recovery by angiogenesis after myocardial infarction. Curr Pharm Des. 2018;24:2967–73. 53. Agrahari V, Agrahari V, Burnouf PA, Chew CH, Burnouf T. Extracellular microvesicles as new industrial therapeutic frontiers. Trends Biotechnol. 2019; 37(7):707–29. 54. Bian X, Ma K, Zhang C, Fu X. Therapeutic angiogenesis using stem cell- derived extracellular vesicles: an emerging approach for treatment of ischemic diseases. Stem Cell Res Ther. 2019;10:158 Available from: https:// stemcellres.biomedcentral.com/articles/10.1186/s13287-019-1276-z. 55. Badimon L, Suades R, Arderiu G, Peña E, Chiva-Blanch G, Padró T. Microvesicles in atherosclerosis and angiogenesis: from bench to bedside and reverse. Front Cardiovasc Med. 2017;4:1–20. 56. Kang T, Jones TM, Naddell C, Bacanamwo M, Calvert JW, Thompson WE, et al. Adipose-derived stem cells induce angiogenesis via microvesicle transport of miRNA-31. Stem Cells Transl Med. 2016;5:440–50. 57. Ren S, Chen J, Duscher D, Liu Y, Guo G, Kang Y, et al. Microvesicles from human adipose stem cells promote wound healing by optimizing cellular functions via AKT and ERK signaling pathways 06 Biological Sciences 0601 Biochemistry and Cell Biology. Stem Cell Res Ther. 2019;10:43–57. 58. Wang S, Aurora AB, Johnson BA, Qi X, McAnally J, Hill JA, et al. The endothelial-specific microRNA miR-126 governs vascular integrity and angiogenesis. Dev Cell. 2008;15:261–71. 59. Arderiu G, Peña E, Aledo R, Espinosa S, Badimon L. Ets-1 transcription is required in tissue factor driven microvessel formation and stabilization. Angiogenesis. 2012;15:657–69. 60. Dittmer J. The biology of the Ets1 proto-oncogene. Mol Cancer. 2003:29–50. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.