Skip to main content
NCBI home page
Stem Cells logoLink to Stem Cells
. 2023 Apr 21;41(6):617–627. doi: 10.1093/stmcls/sxad032

Extracellular Vesicles From Primed Adipose-Derived Stem Cells Enhance Achilles Tendon Repair by Reducing Inflammation and Promoting Intrinsic Healing

Hua Shen 1,✉, Ryan A Lane 2
PMCID: PMC10267691  PMID: 37085269

Abstract

Achilles tendon rupture is a common sports-related injury. Even with advanced clinical treatments, many patients suffer from long-term pain and functional deficits. These unsatisfactory outcomes result primarily from an imbalanced injury response with excessive inflammation and inadequate tendon regeneration. Prior studies showed that extracellular vesicles from inflammation-primed adipose-derived stem cells (iEVs) can attenuate early tendon inflammatory response to injury. It remains to be determined if iEVs can both reduce inflammation and promote regeneration in the later phases of tendon healing and the underlying mechanism. Therefore, this study investigated the mechanistic roles of iEVs in regulating tendon injury response using a mouse Achilles tendon injury and repair model in vivo and iEV-macrophage and iEV-tendon cell coculture models in vitro. Results showed that iEVs promoted tendon anti-inflammatory gene expression and reduced mononuclear cell accumulation to the injury site in the remodeling phase of healing. iEVs also increased collagen deposition in the injury center and promoted tendon structural recovery. Accordingly, mice treated with iEVs showed less peritendinous scar formation, much lower incidence of postoperative tendon gap or rupture, and faster functional recovery compared to untreated mice. Further in vitro studies revealed that iEVs both inhibited macrophage M1 polarization and increased tendon cell proliferation and collagen production. The iEV effects were partially mediated by miR-147-3p, which blocked the toll-like receptor 4/NF-κB signaling pathway that activated the M1 phenotype of macrophages. The combined results demonstrate that iEVs are a promising therapeutic agent that can enhance tendon repair by attenuating inflammation and promoting intrinsic healing.

Keywords: extracellular vesicles, tendon injury and repair, adipose-derived stem cells, exosomes, tendon healing, inflammation

Graphical abstract

Graphical Abstract.

Graphical Abstract

Open in a new tab

Significance Statement.

Using a clinically relevant tendon injury and repair model, this study revealed that extracellular vesicles generated by inflammation-primed adipose-derived stem cells (iEVs) can target macrophages and tendon cells directly and enhance tendon structural and functional recovery by limiting inflammation and promoting intrinsic healing. Results further identified miR-147-3p as one of the active components of iEVs that modulate macrophage inflammatory response by inhibiting toll-like receptor 4/NF-κB signaling pathway. These promising results have opened the way for the clinical application of iEVs in the treatment of tendon injuries and other related disorders.

Introduction

Tendon injury, accounting for nearly half of all sports-related injuries, is one of the most common and challenging orthopedic conditions.1-5 Being the largest and strongest tendon in the body, the Achilles tendon is one of the most injured tendons due to sports, exercise, or athletic activities.4,6 The tendon attaches the calf muscle to the calcaneus bone. By transmitting muscle force to the calcaneus, the Achilles tendon enables plantar flexion of the foot, which is required for varied types of locomotion, such as walking, running, and jumping. Current clinical interventions after Achilles tendon rupture include surgical or nonsurgical treatments and rehabilitation.7 Despite state-of-the-art surgical and rehabilitation techniques, many patients suffer from long-term pain and reduced function following Achilles tendon rupture. A clinical study in a general population reported that major functional deficits persist 2 years after acute Achilles tendon rupture.8 Others reported that calf muscle performance deficits remain 7 years after an Achilles tendon rupture.9 Similarly, 30%-40% of professional athletes were unable to return to play after Achilles tendon ruptures; those who did return played in fewer games, had less play time, and performed at a lower level than before the injury.10,11 Thus, there is a great need for new treatments to effectively improve Achilles tendon healing after injury.

The unsatisfactory outcomes after tendon injury are associated with excessive inflammation mainly driven by infiltrating macrophages and inadequate regeneration mediated by resident tendon cells, including tenocytes, tendon stem/progenitor cells, and epitenon cells.5,12-17 Activated macrophages primarily exhibit 2 functional phenotypes: a default pro-inflammatory M1 phenotype and an alternative anti-inflammatory and pro-regenerative M2 phenotype.18 Although inflammation is required to initiate the healing process and clear damaged tissues, excessive/sustained inflammation causing tendon cell death, matrix degradation, and peritendinous scar formation impedes tendon structural and strength recovery and excursion.12,13,19-21 Moreover, the collagen-producing tendon cells lack the capacity to regenerate functional tissue after ­injury.5,14-17 Therefore, our treatment strategy has focused on both limiting macrophage inflammatory response and stimulating tendon cell activity and function to promote scarless intrinsic healing for minimal postoperative complications and accelerated functional recovery.

We recently found that extracellular vesicles generated by inflammation-primed adipose-derived stem cells (iEVs) can reduce inflammation in the early phase of tendon healing.22 It remains to be determined if iEVs may also promote tendon structural and functional recovery while limiting inflammation in the later phases of tendon healing and the underlying mechanisms. Extracellular vesicles (EVs) are nanosized vesicles carrying varied types of active molecules including microRNAs (miRNAs) and mRNAs that mediate EV functions.23,24 While nearly all types of cells can generate EVs, EVs from different types of cells and cells at different functional states carry different cargo molecules and thereby exbibit distinct therapeutic potential.22,25-28 We hypothesized that iEVs from inflammation-primed adipose-derived stem cells (iASCs) can both reduce inflammation and stimulate intrinsic tendon healing by shuttling active molecules that regulate macrophage and tendon cell functions. To test the hypothesis, this study investigated the dose effect of iEVs on tendon inflammatory and healing responses up to 28 days after Achilles tendon injury and repair in a preclinical mouse model. The mechanism of iEV action, including active cargo molecules, was explored in vitro using iEV-macrophage and iEV-tendon cell coculture models. Our results showed that iEVs are a promising therapeutic agent for tendon repair.

Materials and Methods

Animals

Adult NF-κB-GFP-luciferase transgenic reporter mice (NGL; Jackson Laboratory; 4-5 months old) of both sexes were used for isolating adipose-derived stem cells (ASCs), preparing bone marrow-derived macrophages, and conducting all in vivo studies. ScxGFP tendon reporter mice (2-3 months old of both sexes) were used for tendon cell isolation.29 All experimental procedures were conducted in accordance with the Public Health Service Policy on Humane Care and Use of Laboratory Animals and approved and overseen by the Washington University Institutional Animal Care and Use Committee.

Cell Isolation and Culture

ASCs expressing the mesenchymal stem cell markers CD29, CD44, and CD90 were isolated from subcutaneous fat of adult NGL mice and cultured in Minimum Essential Medium Alpha (αMEM; Mediatech Inc.) containing 10% fetal bovine serum (FBS; Life Technologies).22,30 iASCs were prepared by pretreating ASCs with interferon γ (IFNγ, 100 ng/mL; R&D Systems) in the same medium for 24 hours. iASCs exhibited an average population doubling time of 2.5 ± 0.4 days (N = 3 isolations) and a viability of 96.3% ± 3.6% similar to those of ASCs (2.4 ± 0.4 days, P = .716 vs. iASCs; 95.7% ± 5.3%, P = .864 vs. iASCs). iASCs showed no apparent differences with ASCs in stem cell marker expression (Supplementary Fig. S1). Bone marrow-derived macrophages were cultured in αMEM containing 10% conditioned medium from L929 cell culture (L929-CM), 100 unit/mL penicillin, 100 μg/mL streptomycin (Life Technologies), and 10% FBS.22 Tendon cells were isolated from tail tendons and cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Mediatech, Inc.) supplemented with 10% FBS.30 Over 90% of isolated tendon cells expressed ScxGFP and about 0.3% ScxGFP+ cells coexpressed CD146.

Preparation of EVs

Naïve EVs and iEVs were isolated from untreated naïve ASCs and iASCs, respectively. Both types of cells were cultured in αMEM containing 2% EV-free FBS for 48 hours. The resulting conditioned medium was cleared via centrifugation at 500 × g for 10 minutes and 10 000 × g for 30 minutes at 4 °C. Vesicles in the cleared medium were collected by ultracentrifugation at 100 000 × g for 70 minutes at 4 °C and resuspended in DPBS as previously described.22

Characterization of EVs

Vesicle morphology was examined via transmission electron microscopy (TEM).22 Vesicle size and concentration were determined by Nanoparticle Tracking Analysis (NTA; NanoSight NS300, Malvern Instruments Ltd.). CD9+/CD63+/CD81+ iEVs were separated from bulk iEVs with Exo-Flow streptavidin magnetic microbeads (System Biosciences) coupled with biotinylated anti-mouse CD63 (Miltenyi, clone REA563), CD9 (Miltenyi, clone MZ3), and CD81 (Miltenyi, clone EAT2) antibodies. Half of iEVs captured by the beads were stained with Exo-FITC Universal Exosome Stain (System Biosciences) and analyzed via flow cytometry, using microbeads prepared in parallel without capturing antibodies as negative controls. The remaining half of captured iEVs were eluted from microbeads with Exosome Elution Buffer (System Biosciences) and subjected to NTA analysis to determine vesicle size and percentage in total iEV population.

Macrophage Inflammatory Response

Macrophages expressing the NGL NF-κB-luciferase reporter were either pretreated with iEVs at a dose corresponding to an iEV donor and recipient cell ratio of 20:122 or transfected with a mirVana miRNA mimic mmu-miR-147-3p (Thermo Fisher Scientific) with Lipofectamine RNAiMAX Transfection Reagent (Life Technologies) in αMEM containing 2% EV-free FBS and 10% EV-free L929-CM for 48 hours. Cells pretreated with equal volume of DPBS or control mimics in the same medium were used as iEV or miRNA mimic negative controls. The pretreated cells were stimulated with lipopolysaccharides (LPS, 100 ng/mL; Sigma-Aldrich) and IFNγ (50 ng/mL) in αMEM containing 10% FBS and 10% L929-CM for 24 hours. The resulting conditioned medium was collected and assessed for inflammatory cytokine and chemokine concentrations using a ProcartaPlex multiplex immunoassay kit (Thermo Fisher Scientific). The cells were either lysed and assessed for NF-κB activity with an Illumination Firefly Luciferase Enhanced Assay Kit (Gold Biotechnology) or dislodged and stained with fluorophore-conjugated antibodies recognizing F4/80 (Miltenyi, clone REA126) and iNOS (Miltenyi, clone REA982). The stained cells were analyzed by flow cytometry after cell debris, dead cells, and doublets had been excluded based on scatter signals.

Tendon Cell Activity and Function

Tendon cells were treated with iEVs at a dose corresponding to an iEV donor and recipient cell ratio of 20:1 or miR-147-3p mimics as described earlier in DMEM containing 2% EV-free FBS for 48 hours. Conditioned medium from the culture was collected to determine type I collagen production using a Mouse Type I Collagen Detection Kit (Chondrex). The cells were lysed to assess cell proliferation with a CyQUANT Cell Proliferation Assay (Thermo Fisher Scientific) according to the manufacturer’s instructions. Type I collagen production was normalized by tendon cell counts. All results are presented as fold changes compared to the average of corresponding control.

In Vivo Study Design

NGL mice (15 males and 17 females) from 4 litters were subjected to right Achilles tendon 2/3 transection and suture repair. Mice from 2 of the 4 litters (N = 17) were preassigned for a video-based gait analysis to assess tendon functional recovery after injury.31 Mice from the other 2 litters (N = 15) were preassigned for live bioluminescence imaging (BLI) to determine repair site NF-κB activity.22 The repaired animals were divided into 3 litter- and gender-matched groups (N = 10-11/group) and treated with 0 (CTRL), 1E+09 (+iEVL), and 5E+09 (+iEVH) iEVs loaded on a collagen sheet. BLI (N = 5-6/group) was conducted 1 day before and 1, 4, 7, 13, and 28 days after tendon repair. Functional assessments (N = 5-6/group) were performed 1-2 days before and 13, 20, and 27 days after tendon repair. All repaired mice were euthanized 28 days after injury and repair. The integrity of repaired tendons was assessed under a dissecting microscope. Postoperative gap formation and rupture were defined as partial and complete loss of the continuity of the repaired tendons, respectively.22 Tendon inflammatory and healing responses were further assessed histologically (N = 3-4/group) and quantitatively by TaqMan PCR for changes in tendon gene expression (N = 7-8/group) using intact tendons from age and gender-matched healthy NGL mice as intact tendon controls (N = 8).

Mouse Achilles Tendon Injury, Repair, and Treatment

The methods for mouse right Achilles tendon 2/3 transection, suture repair, and iEV delivery were described previously.12,22 In brief, Achilles tendon transection was performed at the midpoint level between the calcaneal insertion and the musculotendinous junction under isoflurane anesthesia. The transected tendon was immediately repaired with a 2-strand modified Kessler technique.22 iEVs were pre-loaded to the surface of one side of a collagen sheet containing 2 mg/mL type I collagen (Corning Life Sciences) and applied around the surface of repair site with the iEV-loaded side facing repaired tendon.12,22 After recovery from anesthesia, repaired mice were returned to their home cages without physical restraint.

Tendon Histology

Mouse Achilles tendons attached to the calcaneus bone and calf muscle were fixed with 4% paraformaldehyde in phosphate buffered saline (PBS), decalcified in 14% ethylenediaminetetraacetic acid (pH 7.2; Sigma-Aldrich), and embedded in paraffin. Serial sagittal sections (5 μm thick) were prepared and stained with a Russell Movat Pentachrome Stain Kit (StatLab). The pentachrome stain reveals cell nuclei and elastic fibers in dark purple to black color, collagen in yellow, and intensely acidophilic collagen fibers in red on a yellow background.22,32 Stained sections were scanned with a NanoZoomer Whole Slide Imaging System (2.0-HT, Hamamatsu). Images at the injury site were exported at 40× magnification for cellularity and collagen content assessments and 10× magnification for scar tissue analysis. The cellularity of repaired tendons was determined histologically by counting the number of fibroblast-like cells and mononuclear cells in the tendon region of exported images. The posterior and fat pad scar areas were defined as the area occupied by scar tissue posterior to repaired tendon and the area occupied by scar tissue at the fat pad region anterior to repaired tendon, respectively. The scar tissue area and collagen area were determined with the area analysis and color range selection tools of Adobe Photoshop CC 2015.5 as previously described.22 Every sample was assessed blindly on 2 different sections containing the Achilles tendon and its attachments at the level approximately halfway from the tendon surface. The results were averaged and normalized by tissue size.

Ankle Joint Angle Measurement

Because the Achilles tendon is required for plantar flexion of the ankle joint, the angle of ankle joint was used as a measurement of tendon functional recovery.31 The angle was determined when mice were at a full standing position (ie, all paws are in contact with treadmill belt) while running on a rodent treadmill (EXER 3/6, Columbus Instruments) at a speed of 6~8 m/minute. All mice were pretrained. A 30-second video that recorded the sagittal view of mouse during running was taken with a digital camera on the indicated days before and after tendon injury and repair. All frames with mouse at the full standing position (11 ± 2 frames per mouse per time point) in each video were analyzed with ImageJ 1.52a. for the ankle joint angle formed from the fibular head and the fifth metatarsal head to the calcaneal tuberosity.

RNA Isolation and Quantitative RT-PCR

Total RNAs from cultured cells and Achilles tendons were isolated with TRIzol reagent and RNeasy Mini Spin Column (Qiagen Sciences) and reversely transcribed into cDNAs using a SuperScript IV VILO Master Mix (Life Technologies).22,30 The relative abundances of genes of interest were determined by TaqMan PCR using primers and probes purchased from Applied Biosystems TaqMan Gene Expression Assays. Ipo8 was used as an endogenous reference gene. Changes in tendon gene expression were determined by the comparative Ct method and expressed as relative mRNA abundance.

EV miRNA TaqMan PCR

EV RNA was isolated with a SeraMir Exosome RNA Column Purification Kit (System Biosciences). RNA integrity was determined by an Agilent 2100 bioanalyzer. cDNA of EV miRNA was synthesized from 10 ng total RNA using a TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems) and primers specific for mmu-miR-147-3p (alias, hsa-miR-147b; Applied Biosystems). The relative abundance of EV miRNA was determined by TaqMan PCR using TaqMan Universal PCR Master Mix II and TaqMan primers and probes from Applied Biosystems. hsa-miR-423-3p was used as an endogenous reference miRNA due to its abundancy and steadiness in ASC EVs.

Small RNA-Sequencing and Data Analysis

cDNA Library preparation, RNA-sequencing, data acquisition, quality control, and processing were performed by the Washington University Genome Technology Access Center. In brief, cDNA libraries were prepared from 5 μL of each RNA sample using a TruSeq Small RNA library Preparation kit (Illumina) and amplified for 11 cycles with oligos adding a unique index. The resulting samples were sized and quantified with a Bioanalyzer High Sensitivity DNA Analysis kit (Agilent). A region table was used to determine molar concentration between 145 and 160 bp. An equimolar amount was pooled, and the size was selected. Libraries were run on HiSeq 3000 (Illumina) using single reads extending 50 bp. Differentially expressed miRNAs were identified using generalized linear models and filtered for False Discovery Rate (FDR) adjusted P value ≤ .05. The sequencing data have been deposited in NCBI’s Gene Expression Omnibus and are accessible through the accession number GSE217484.33

Statistical Analysis

Unless described elsewhere, all data are shown as mean ± standard deviation and analyzed with Prism 9 (GraphPad Software, LLC). Two-tailed t tests and one-way analysis of variance (ANOVA) followed by Tukey’s or Dunn’s multiple comparisons tests (when appropriate) were used for 2- and multiple-group comparisons of in vitro data. Two-way repeated-measures ANOVA followed by FDR multiple comparisons tests with 2-stage step-up procedure of Benjamini, Krieger, and Yekutieli was used to assess the longitudinal impact of iEV treatment on tendon NF-κB activity. Two-way repeated-measures ANOVA followed by Tukey’s multiple comparisons tests was used to evaluate the impact of iEVs on ankle joint angle recovery. One-way ANOVA followed by Tukey’s or Dunn’s multiple comparisons tests (when appropriate) was used to evaluate the effect of iEVs on tendon gene expression and tendon histomorphometry. Outliers were identified using the ROUT method with an FDR of 1%.34 Significance was set at P < .05.

Results

Characterization of iEVs

iEVs are pancake-like and of varied sizes as revealed by TEM imaging (pointed by white arrows in Fig. 1A). Bulk iEVs exhibited a mean mode size of 146.4 ± 12.4 nm (N = 3 independent isolations; Fig. 1B) and contained approximately 70% CD9+/CD63+/CD81+ exosomes (a major subtype of EVs). As expected, CD9+/CD63+/CD81+ exosomes separated from bulk iEVs via a magnetic bead-assisted sorting system (Fig. 1C) were smaller (mean mode size: 105.4 ± 6.2 nm, N = 4 independent isolations) than bulk iEVs (P = .002).

Figure 1.

Figure 1.

Open in a new tab

Characterization of extracellular vesicles from inflammation-primed adipose-derived stem cells (iEVs). (A) A representative transmission electron microscopy image of iEVs (indicated with white arrows); scale bar = 100 nm. (B) Representative nanopartical tracking analysis of iEVs. (C) Representative flow cytometry analysis of iEVs sorted with microbeads coupled with (right peak) or without (left peak) biotinylated antibodies against indicated exosome markers.

The Dose Effect of iEVs on the Tendon Inflammatory Response to Acute Injury and Repair

The tendon inflammatory response was first assessed longitudinally in live NGL mice subjected to right Achilles tendon injury and repair. BLI of NF-κB activity at the repair site revealed a 2- to 3.5-fold increase in untreated control mice (CTRL) until at least 13 days after repair (Fig. 2A). iEV treatment at either dose effectively shortened the inflammatory response to less than 7 days. Higher than preinjury level NF-κB activity was only detected in iEV-treated mice (+iEVL and +iEVH) 1 and 4 days after repair. High-dose iEVs were more effective than low-dose iEVs, significantly reducing NF-κB activity 1, 4, 7, and 13 days after tendon injury and repair (Fig. 2A).

Figure 2.

Figure 2.

Open in a new tab

iEVs dose-dependently attenuate tendon inflammatory response to acute injury and repair. (A) Differential changes in injury site NF-κB activity in mice treated with 0, 1E+09, and 5E+09 iEVs (CRTL, +iEVL, and +iEVH) on the indicated days after Achilles tendon injury and repair. *P < .05 compared to the pre-injury level (–1 day) of corresponding groups. —P < .05 between indicated groups. (B and C) The different expression of inflammatory (B) and anti-inflammatory (C) genes in intact and repaired tendons subjected to the indicated treatments 28 days after injury and repair. —P < .05 between indicated groups.

As we previously assessed the effect of iEVs on early tendon inflammatory response,22 this study determined changes in tendon inflammatory gene expression 28 days after injury and iEV treatment. Results showed that the expression levels of many but not all inflammatory genes (ie, Il1b, Il6, Ifng, and Tnf but not Nos2, Fig. 2B) were higher than normal, and no significant differences were detected between iEV-treated and untreated tendons. Despite these increases, the relative expression levels of these genes in the remodeling phase were either very low (Il6 and Ifng, Fig. 2B) or substantially reduced to less than 5% (Il1b, Fig. 2B, P < .001) and 35% (Tnf, Fig. 2B, P <.001) of those detected previously in untreated control tendons 7 days after injury and repair.22 Notably, tendons treated with iEVs at either dose expressed ­higher-than-normal levels of Arg1, a marker for the anti-inflammatory M2 macrophages (Fig. 2C). High-dose iEVs also dramatically increased the expressions of anti-inflammatory genes Il13 that activates macrophage M2 phenotype and Il1rn, an IL1β-inhibitor (Fig. 2C), thus supporting a positive role of iEVs in promoting macrophage M2 polarization.

Histological assessment was conducted on pentachrome-stained tendon sections 28 days after injury and repair. Results revealed that tendon injury led to massive scar tissue formation along the posterior surface (red crosses in Fig. 3A) and around the fat pad region of untreated tendons (white asterisk in Fig. 3A) coupled with extensive mononuclear cell accumulation to the injury site (red arrowheads in Fig. 3D). In accordance with the effects detected above, iEVs at either dose effectively limited the scar tissue formation and mononuclear cell accumulation (Fig. 3A-3I).

Figure 3.

Figure 3.

Open in a new tab

iEVs dose-dependently regulate Achilles tendon remodeling 28 days after injury and repair. (A-F) Representative images of pentachrome-stained Achilles tendon sections from mice treated with 0, 1E+09, or 5E+09 iEVs (CRTL, +iEVL, or +iEVH). The pentachrome stain reveals cell nuclei and elastic fibers in dark purple to black color, collagen in yellow, and intensely acidophilic collagen fibers in red on a yellow background. Black braces and white asterisks in A-C mark the peripheral borders of repaired Achilles tendons and the fat pad region. ↓, transection site; +, posterior scar. Red and white arrowheads in D-F point to mononuclear cells and fibroblast-like cells, respectively. Scale bar in A is equal to 250 μm and applies to A-C. Scale bar in D is equal to 25 μm and applies to D-F. (G-K) Semiquantitative evaluation of the effects of iEVs on peritendinous scar formation (G and H), tendon cellularity (I and J), and collagen deposition (K). —P < .05 between indicated groups.

The Dose Effect of iEVs on Tendon Remodeling After Injury and Repair

Changes in tenogenic and tendon matrix gene expression were assessed in Achilles tendons 28 days after injury and repair. Results showed extensive increases in expression of all genes evaluated in untreated control tendons compared to intact tendons (Fig. 4). Interestingly, application of iEVs at both doses resulted in a smaller increase in Tnmd expression than untreated control. High-dose, but not low-dose, iEVs also suppressed the increase in Scx and Col1a1 expression in the remodeling phase of healing (Fig. 4).

Figure 4.

Figure 4.

Open in a new tab

iEVs modulate tendon gene expression in the remodeling phase of tendon healing. (A and B) The differential expression of genes involved in tendon cell growth and differentiation (A) and matrix remodeling (B) in intact and repaired tendons treated with 0, 1E+09, or 5E+09 iEVs (CRTL, +iEVL, or +iEVH) 28 days after Achilles tendon injury and repair. —P < .05 between indicated groups.

Consistently, histological assessment revealed robust fibroblast-like cell accumulation to the injury center of repaired tendons from all groups (white arrowheads in Fig. 3D-3F). While there were no significant differences in fibroblast-like cell density among the 3 repair groups (Fig. 3J), cells in iEV-treated tendons were better aligned with the long axis of Achilles tendon than those in untreated tendons (white arrowheads in Fig. 3D-3F). With controlled Col1a1 expression, high-dose iEV both significantly increased collagen accretion in the injury center and effectively limited peritendinous scar formation after tendon injury and repair (Fig. 3A-3H and 3K), demonstrating a positive role of iEVs in regulating tendon injury response.

The Dose Effect of iEVs on Tendon Repair Outcome

Mouse functional recovery after Achilles tendon repair was assessed in live mice by a video-based gait analysis for ankle joint angle of repaired limbs. The angle of all mice assessed was slightly over 90° prior to injury and reduced to around 80° 13 days after injury and repair (Fig. 5). In untreated mice, the defect was slowly reduced at a rate of approximately 2° per week but remained significantly lower than preinjury level by 27 days after repair. Treatment with iEVs at both doses significantly increased the recovery rate by 30% and 43%, respectively. As a result, by 27 days after repair, untreated mice showed nearly 50% functional deficit compared to preinjury level. In contrast, iEVL- and iEVH-treated mice experienced only approximately 20% deficits.

Figure 5.

Figure 5.

Open in a new tab

Changes in ankle joint angle in mice treated with 0, 1E+09, or 5E+09 iEVs (CRTL, +iEVL, or +iEVH) on the indicated days after Achilles tendon injury and repair. *P < .05 compared to respective preinjury level. —P < .05 between indicated groups.

Postmortem assessments of tendon gap and rupture formation were performed 28 days after injury and repair. In accordance with improved tendon structural and functional recovery, only one gap formation was noted in a total of 11 repairs treated with low-dose iEVs. Likewise, one incidence of gap formation was detected in a total of 10 repairs treated with high-dose iEVs. By contrast, 2 gap formations and 3 ruptures were found in a total of 11 untreated repairs.

The In Vitro Effect of iEVs on Macrophage Inflammatory Response

The combined results from this and prior studies support that iEVs attenuate tendon inflammatory response by inhibiting macrophage NF-κB signaling and promoting an M1-to-M2 phenotypic transition.22 To explore this further, we stimulated bone marrow-derived F4/80+iNOS− macrophages with TLR4 agonists LPS and IFNγ. As expected, the stimulation activated NF-κB (Fig. 6A), leading to an F4/80+iNOS+ M1 phenotype (Supplementary Fig. S2) with mass releases of inflammatory cytokines IL-1β, TNFα, and IL-6 and chemokines CXCL1 and CXCL10 but not M2 phenotype stimuli IL-4 and IL-13 (Fig. 6B). Pretreatment of macrophages with iEVs inhibited NF-κB activation by LPS and IFNγ (Fig. 6C) and dramatically reduced IL-1 and IL-6 release by macrophages (Fig. 6D).

Figure 6.

Figure 6.

Open in a new tab

iEVs modulate macrophage inflammatory response by inhibiting toll-like receptor 4 (TLR4)/NF-κB signaling pathway with regulatory miRNAs. (A) Macrophage NF-κB activity induced by TLR4 agonists LPS and IFNγ. (B) Macrophage cytokine and chemokine release stimulated by LPS and IFNγ. (C and D) The effect of iEVs on macrophage NF-κB activity (C) and inflammatory cytokine release (D) induced by LPS and IFNγ. (E) Volcano plot of differentially carried miRNAs by iEVs and EVs from inflammation-primed and naïve ASCs (iASC and ASC), respectively. Red and black dots, adjusted P-value ≤.05 and >.05. (F) The relative abundances of miR-147b in iEVs, EVs, and their parent cells. (G and H) The effect of miR-147b mimic overexpression on macrophage NF-κB activity (G) and inflammatory cytokine release (H) induced by LPS and IFNγ. Control, control mimics. —P < .05 between indicated groups for A-D and F-H.

To identify active molecules mediating the ­anti-inflammatory effect of iEVs, we performed a small RNA-sequencing analysis. Given iEVs have been found to be more effective than naïve EVs in reducing tendon inflammatory response,22 we compared iEVs and their parent iASCs with naïve EVs and ASCs. Results revealed that both naïve and primed EVs contained more miRNAs than parent cells (Supplementary Fig. S3).33 Further differential analysis of iEV and naïve EV miRNAs identified miR-147 as the most distinctively enriched miRNAs in iEVs (P = .00072; Fig. 6E). The finding was validated by Taqman PCR. As shown in Fig. 6F, iEVs carried the most miR-147-3p (alias, miR-147b) compared to naïve EVs and their parent cells. Overexpression of miR-147b but not negative control mimics in macrophages effectively inhibited macrophage NF-κB activation (Fig. 6G), iNOS expression (Supplementary Fig. S2), and IL-1β and IL-6 releases (Fig. 6H) induced by LPS and IFNγ.

The In Vitro Effect of iEVs on Tendon Cells

Besides macrophages, iEVs have been found to target ScxGFP+ tendon cells in vivo.22 To better understand the effect of iEVs on tendon cells, we cocultured tendon cells and iEVs. Results confirmed that iEVs were taken up by tendon cells (Fig. 7A) and promoted tendon cell proliferation compared to untreated cells (Fig. 7B). Significantly, the effect was coupled with a 45% increase in type I collagen production by tendon cells (Fig. 7C). Overexpression of miR-147b mimics in tendon cells did not recapitulate the effect of iEVs on tendon cell proliferation (Fig. 7D, P = .925), indicating other active molecules carried by iEVs mediated the effect.

Figure 7.

Figure 7.

Open in a new tab

iEVs promote tendon cell proliferation and type I collagen production. (A) A representative fluorescence image of tendon cells stained with 4ʹ,6-diamidino-2-phenylindole (DAPI) in blue and iEVs stained with PKH26 in red. (B) Differential changes in tendon cell population 48 hours after vehicle control or iEV treatment. (C) Relative type I collagen production per tendon cell 48 hours after vehicle control or iEV treatment. (D) Changes in tendon cell population 48 hours after treatment with negative control miRNA mimics or miR-147b mimics. —P < .05 between indicated groups for B and C.

Discussion

A major clinical challenge for biological improvement of tendon repair is to achieve the contrasting goals of reducing inflammation without compromising tendon structure and strength recovery and stimulating tendon tissue regeneration while avoiding peritendinous scar formation.5,21,35,36 Prior work has established the role of iEVs in attenuating inflammation in the early phase of tendon healing.22 It remained to be determined if iEVs might promote tendon structural and functional recovery while limiting scar tissue formation in the later phases of tendon healing. Here, we examined the long-term dose effect of iEVs in a clinically relevant mouse Achilles tendon injury and repair model. Results first confirmed that iEVs dose-dependently reduced repair site NF-κB activity in the early inflammatory phases of tendon healing. Results further showed that iEVs increased anti-inflammatory gene expression and reduced inflammatory cell accumulation in the later remodeling phases of healing. Significantly, iEVs both promoted collagen deposition and tendon structural recovery and limited peritendinous scar formation, leading to less postoperative complications, and accelerated functional recovery. The subsequent in vitro studies further revealed that iEVs can reduce inflammation via miR-147b that targets TLR4/NF-κB signaling pathway and inhibits macrophage M1 polarization. iEVs can also promote tendon healing by increasing tendon cell proliferation and collagen production. Collectively, our results demonstrate that iEVs are a promising therapeutic agent for enhanced tendon repair by both reducing inflammation and stimulating intrinsic healing response.

Available evidence strongly supports that the ­anti-inflammatory function of iEVs at least partially results from their ability to inhibit macrophage TLR4/NF-κB signaling by delivering anti-inflammatory miRNAs including miR-147b. Specifically, iEVs were previously found to reduce Tlr4 expression and NF-κB activation after Achilles tendon injury.22 This study further showed that iEVs blocked macrophage NF-κB activation and IL-1β and IL-6 production triggered by the TLR4 agonists LPS and IFNγ. The subsequent small RNA-sequencing study revealed that iEVs are distinctively enriched with miR-147b. Like iEVs, overexpressing miR-147b mimics inhibited macrophage NF-κB activation and M1 polarization, leading to reduced IL-1β and IL-6 production induced by LPS and IFNγ. Consistently, miR-147 has been reported to reduce inflammatory cytokine expression in peritoneal macrophages stimulated with ligands to TLR4 in vitro37 and attenuate aortic inflammation as an active component of other stem cell EVs in vivo.38 Further study is needed to identify the target molecule of miR-147b in the TLR4/NF-κB signaling pathway.

iEVs also targeted tendon cells and increased tendon cell proliferation and type I collagen production. Overexpression of miR-147b mimics in tendon cells showed no apparent effects on tendon cell proliferation, indicating other iEV molecules contributed to the effect. Besides miRNAs, other small non-coding RNAs and mRNAs carried by iEVs may contribute to the iEV effects on tendon cells. More research into the active components of iEVs will help to fully elucidate the mechanism of iEV action in enhancing tendon repair.

Our results showed dose-dependent effects of iEVs on tendon repair. As expected, high-dose iEVs were more effective than low-dose iEVs in reducing repair site NF-κB activity in the early inflammatory phase of tendon healing and in promoting anti-inflammatory genes Il1rn and Il13 expression and collagen deposition in the later remodeling phase. Interestingly, iEVH-treated tendons expressed less Col1a1 and Scx than controls in the remodeling phase, which contrasts with accelerated recovery following iEV treatment. As tendon cell propagation and matrix production are expected to peak in the earlier proliferative phase, the reduced tendon gene expression possibly reflects an accelerated healing response to iEV treatment. This idea is supported by our prior study, showing iEVs increased collagen deposition in the injury center by 3-fold as early as 7 days after repair.22 Moreover, excessive inflammation is known to cause disorganized collagen deposition and fibrotic scar formation, which impair tendon function.35,36 The reduced Col1a1 expression in the remodeling phase of healing is likely also a result of the ­anti-inflammatory effect of iEVs. Consistently, iEVs were found to reduce mononuclear cell accumulation in repaired tendons and inhibit peritendinous scar formation.

Because Achilles tendon integrity is required for plantar flexion of the foot and locomotor movements such as running, the ankle joint angle of the injured limb during running was used as an index for Achilles tendon functional recovery.31 As expected, the angle was significantly reduced following Achilles tendon transection. With less peritendinous scar formation and accelerated intrinsic tendon matrix regeneration, mice treated with either dose of iEVs exhibited higher recovery rates than untreated mice. While high-dose iEVs more effectively reduced inflammation and promoted tendon matrix regeneration and remodeling than low-dose iEVs, no apparent dose effect of iEVs on the recovery rate was detected. The use of a moderate running protocol (ie, 6~8 m/minute without slope) likely limits the sensitivity of this measurement. Future assessment of tendon functional recovery will be conducted during uphill running at a higher speed (eg, 12-16 m/min).

A limitation of this study is the lack of biomechanical analysis of repaired tendons. Although the ankle joint analysis demonstrated the ability of iEVs to alleviate mouse functional deficits after Achilles tendon injury and repair, further evaluation of the iEV effect on tendon biomechanical property could provide additional information for clinical application of iEVs in treating tendon injury. A second limitation is the lack of a control group for EVs from naïve ASCs. Our prior studies showed that iEVs are more effective than naïve EVs in reducing macrophage and tendon inflammatory responses using the same in vitro and in vivo models.22 Moreover, iEVs, but not naïve EVs, were found to stimulate collagen deposition in the injury center as early as 7 days after repair.22 Therefore, this study focused on the long-term dose effect and the mechanism of action of iEVs. Our new findings regarding the selective enrichment of miR-147b in iEVs and its role in regulating macrophage polarization further support that iEVs are a superior biologic for enhanced tendon repair.

Conclusion

Our results showed that iEVs can effectively enhance Achilles tendon healing by both targeting macrophages to attenuate inflammation and tendon cells to promote intrinsic tendon healing and functional recovery. Results also identified miR-147-3p as one of active components mediating the anti-inflammatory function of iEVs. These findings open the way for the clinical application of iEVs in tendon repair and provide the basis for future engineering of component-defined and disease-specific iEVs to treat tendon injuries and other related disorders.

Supplementary Material

sxad032_suppl_Supplemental_Materials
Click here for additional data file. (544.3KB, pdf)

Contributor Information

Hua Shen, Department of Orthopaedic Surgery, Washington University School of Medicine, St. Louis, MO, USA.

Ryan A Lane, Department of Orthopaedic Surgery, Washington University School of Medicine, St. Louis, MO, USA.

Acknowledgment

The authors thank Dr. Ratna B. Ray for her help with developing this project and reading the manuscript. The authors also thank Dr. Dimitrios Skouteris for his assistance with mouse surgeries. The graphic abstract was created with BioRender.com.

Funding

This research was funded by the NIH/NIAMS R21 AR075274, the Foundation for Barnes-Jewish Hospital, and the Washington University Institute of Clinical and Translational Sciences (partially supported by the NIH/NCATS UL1 TR002345). This publication was made possible in part by the NIH/NIAMS P30 AR074992, which supported the histological service provided by the Washington University Musculoskeletal Research Center.

Conflict of Interest

Hua Shen is an inventor with a pending patent that covers the iEVs and collagen sheet used in this study. Ryan A. Lane indicated no financial relationships.

Author Contributions

Conception and design, funding acquisition, collection and assembly of data, data analysis and interpretation, manuscript writing, and final approval of manuscript: H.S. Collection and analysis of data, and final approval of manuscript: R.A.L

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  • 1. Butler DL, Juncosa-Melvin M, Boivin GP, et al. Functional tissue engineering for tendon repair: a multidisciplinary strategy using mesenchymal stem cells, bioscaffolds, and mechanical stimulation. J Orthop Res. 2008;26(1):1-9. 10.1002/jor.20456 [DOI] [PubMed] [Google Scholar]
  • 2. Pennisi E. Tending tender tendons. Science. 2002;295(5557):1011. 10.1126/science.295.5557.1011 [DOI] [PubMed] [Google Scholar]
  • 3. de Jong JP, Nguyen JT, Sonnema AJ, et al. The incidence of acute traumatic tendon injuries in the hand and wrist: a 10-year population-based study. Clin Orthop Surg. 2014;6(2):196-202. 10.4055/cios.2014.6.2.196 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Järvinen TA, Kannus P, Maffulli N, Khan KM.. Achilles tendon disorders: etiology and epidemiology. Foot Ankle Clin. 2005;10(2):255-266. 10.1016/j.fcl.2005.01.013 [DOI] [PubMed] [Google Scholar]
  • 5. Linderman SW, Gelberman RH, Thomopoulos S, Shen H.. Cell and biologic-based treatment of flexor tendon injuries. Oper Tech Orthop. 2016;26(3):206-215. 10.1053/j.oto.2016.06.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Hess GW. Achilles tendon rupture: a review of etiology, population, anatomy, risk factors, and injury prevention. Foot Ankle Spec. 2010;3(1):29-32. 10.1177/1938640009355191 [DOI] [PubMed] [Google Scholar]
  • 7. Erickson BJ, Cvetanovich GL, Nwachukwu BU, et al. Trends in the management of achilles tendon ruptures in the United States Medicare Population, 2005-2011. Orthop J Sports Med. 2014;2(9):2325967114549948. 10.1177/2325967114549948 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Olsson N, Nilsson-Helander K, Karlsson J, et al. Major functional deficits persist 2 years after acute Achilles tendon rupture. Knee Surg Sports Traumatol Arthrosc. 2011;19(8):1385-1393. 10.1007/s00167-011-1511-3 [DOI] [PubMed] [Google Scholar]
  • 9. Brorsson A, Grävare Silbernagel K, Olsson N, Nilsson Helander K.. Calf muscle performance deficits remain 7 years after an achilles tendon rupture. Am J Sports Med. 2018;46(2):470-477. 10.1177/0363546517737055 [DOI] [PubMed] [Google Scholar]
  • 10. Amin NH, Old AB, Tabb LP, et al. Performance outcomes after repair of complete Achilles tendon ruptures in national basketball association players. Am J Sports Med. 2013;41(8):1864-1868. 10.1177/0363546513490659 [DOI] [PubMed] [Google Scholar]
  • 11. Trofa DP, Miller JC, Jang ES, et al. Professional athletes’ return to play and performance after operative repair of an achilles tendon rupture. Am J Sports Med. 2017;45(12):2864-2871. 10.1177/0363546517713001 [DOI] [PubMed] [Google Scholar]
  • 12. Shen H, Kormpakis I, Havlioglu N, et al. The effect of mesenchymal stromal cell sheets on the inflammatory stage of flexor tendon healing. Stem Cell Res Ther. 2016;7(1):144. 10.1186/s13287-016-0406-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Sunwoo JY, Eliasberg CD, Carballo CB, Rodeo SA.. The role of the macrophage in tendinopathy and tendon healing. J Orthop Res. 2020;38(8):1666-1675. 10.1002/jor.24667 [DOI] [PubMed] [Google Scholar]
  • 14. Sakabe T, Sakai K, Maeda T, et al. Transcription factor scleraxis vitally contributes to progenitor lineage direction in wound healing of adult tendon in mice. J Biol Chem. 2018;293(16):5766-5780. 10.1074/jbc.RA118.001987 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Best KT, Loiselle AE.. Scleraxis lineage cells contribute to organized bridging tissue during tendon healing and identify a subpopulation of resident tendon cells. FASEB J. 2019;33(7):8578-8587. 10.1096/fj.201900130RR [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Harvey T, Flamenco S, Fan CM.. A Tppp3+Pdgfra+ tendon stem cell population contributes to regeneration and reveals a shared role for PDGF signalling in regeneration and fibrosis. Nat Cell Biol. 2019;21(12):1490-1503. 10.1038/s41556-019-0417-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Best KT, Korcari A, Mora KE, et al. Scleraxis-lineage cell depletion improves tendon healing and disrupts adult tendon homeostasis. Elife. 2021;10:e62203. 10.7554/eLife.62203 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Murray PJ. Macrophage polarization. Annu Rev Physiol. 2017;79: 541-566. 10.1146/annurev-physiol-022516-034339 [DOI] [PubMed] [Google Scholar]
  • 19. Gelberman RH, Linderman SW, Jayaram R, et al. Combined administration of ASCs and BMP-12 promotes an M2 macrophage phenotype and enhances tendon healing. Clin Orthop Relat Res. 2017;475(9):2318-2331. 10.1007/s11999-017-5369-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Tsuzaki M, Guyton G, Garrett W, et al. IL-1 beta induces COX2, MMP-1, -3 and -13, ADAMTS-4, IL-1 beta and IL-6 in human tendon cells. J Orthop Res. 2003;21(2):256-264. 10.1016/S0736-0266(02)00141-9 [DOI] [PubMed] [Google Scholar]
  • 21. Arvind V, Huang AH.. Reparative and maladaptive inflammation in tendon healing. Front Bioeng Biotechnol. 2021;9:719047. 10.3389/fbioe.2021.719047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Shen H, Yoneda S, Abu-Amer Y, Guilak F, Gelberman RH.. Stem cell-derived extracellular vesicles attenuate the early inflammatory response after tendon injury and repair. J Orthop Res. 2020;38(1):117-127. 10.1002/jor.24406 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Valadi H, Ekström K, Bossios A, et al. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol. 2007;9(6):654-659. 10.1038/ncb1596 [DOI] [PubMed] [Google Scholar]
  • 24. Phinney DG, Di Giuseppe M, Njah J, et al. Mesenchymal stem cells use extracellular vesicles to outsource mitophagy and shuttle microRNAs. Nat Commun. 2015;6:8472. 10.1038/ncomms9472 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Phillips W, Willms E, Hill AF.. Understanding extracellular vesicle and nanoparticle heterogeneity: novel methods and considerations. Proteomics. 2021;21(13-14):e2000118. 10.1002/pmic.202000118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Chamberlain CS, Kink JA, Wildenauer LA, et al. Exosome-educated macrophages and exosomes differentially improve ligament healing. Stem Cells. 2021;39(1):55-61. 10.1002/stem.3291 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Song K, Jiang T, Pan P, Yao Y, Jiang Q.. Exosomes from tendon derived stem cells promote tendon repair through ­miR-144-3p-regulated tenocyte proliferation and migration. Stem Cell Res Ther. 2022;13(1):80. 10.1186/s13287-022-02723-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Feng W, Jin Q, Ming-Yu Y, et al. MiR-6924-5p-rich exosomes derived from genetically modified Scleraxis-overexpressing PDGFRα(+) BMMSCs as novel nanotherapeutics for treating ­osteolysis during tendon-bone healing and improving healing strength. Biomaterials. 2021;279:121242. 10.1016/j.biomaterials.2021.121242 [DOI] [PubMed] [Google Scholar]
  • 29. Pryce BA, Brent AE, Murchison ND, Tabin CJ, Schweitzer R.. Generation of transgenic tendon reporters, ScxGFP and ScxAP, using regulatory elements of the scleraxis gene. Dev Dyn. 2007;236(6):1677-1682. 10.1002/dvdy.21179 [DOI] [PubMed] [Google Scholar]
  • 30. Shen H, Gelberman RH, Silva MJ, Sakiyama-Elbert SE, Thomopoulos S.. BMP12 induces tenogenic differentiation of adipose-derived stromal cells. PLoS One. 2013;8(10):e77613e77613. 10.1371/journal.pone.0077613 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Liang JI, Chen MY, Hsieh TH, et al. Video-based gait analysis for functional evaluation of healing achilles tendon in rats. Ann Biomed Eng. 2012;40(12):2532-2540. 10.1007/s10439-012-0619-z [DOI] [PubMed] [Google Scholar]
  • 32. Movat HZ. Demonstration of all connective tissue elements in a single section; pentachrome stains. AMA Arch Pathol. 1955;60(3):289-295. [PubMed] [Google Scholar]
  • 33. Shen H. Differential small RNA-Seq analysis for active EV cargo identification. 2022. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE217484
  • 34. Motulsky HJ, Brown RE.. Detecting outliers when fitting data with nonlinear regression – a new method based on robust nonlinear regression and the false discovery rate. BMC Bioinf. 2006;7:123. 10.1186/1471-2105-7-123 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Nichols AEC, Best KT, Loiselle AE.. The cellular basis of fibrotic tendon healing: challenges and opportunities. Transl Res. 2019;209(July):156-168. 10.1016/j.trsl.2019.02.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Thomopoulos S, Kim HM, Das R, et al. The effects of exogenous basic fibroblast growth factor on intrasynovial flexor tendon healing in a canine model. J Bone Joint Surg Am. 2010;92(13):2285-2293. 10.2106/JBJS.I.01601 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Liu G, Friggeri A, Yang Y, et al. miR-147, a microRNA that is induced upon toll-like receptor stimulation, regulates murine macrophage inflammatory responses. Proc Natl Acad Sci USA. 2009;106(37):15819-15824. 10.1073/pnas.0901216106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Spinosa M, Lu G, Su G, et al. Human mesenchymal stromal cell-derived extracellular vesicles attenuate aortic aneurysm formation and macrophage activation via microRNA-147. FASEB J. 2018;32(11):fj201701138RR. 10.1096/fj.201701138RR [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

sxad032_suppl_Supplemental_Materials
Click here for additional data file. (544.3KB, pdf)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Stem Cells are provided here courtesy of Oxford University Press

RESOURCES