Oncostatin M-Enriched Small Extracellular Vesicles Derived from Mesenchymal Stem Cells Prevent Isoproterenol-Induced Fibrosis and Enhance Angiogenesis

Abstract

Myocardial fibrosis is a pathological hallmark of cardiac dysfunction. Oncostatin M (OSM) is a pleiotropic cytokine that can promote fibrosis in different organs after sustained exposure. However, OSM released by macrophages during cardiac fibrosis suppresses cardiac fibroblast activation by modulating transforming growth factor beta 1 (TGF-β1) expression and extracellular matrix deposition. Small extracellular vesicles (SEVs) from mesenchymal stromal cells (MSCs) are being investigated to treat myocardial infarction, using different strategies to bolster their therapeutic ability. Here, we generated TERT-immortalized human MSC cell lines (MSC-T) engineered to overexpress two forms of cleavage-resistant OSM fused to CD81TM (OSM-SEVs), which allows the display of the cytokine at the surface of secreted SEVs. The therapeutic potential of OSM-SEVs was assessed in vitro using human cardiac ventricular fibroblasts (HCF-Vs) activated by TGF-β1. Compared with control SEVs, OSM-loaded SEVs reduced proliferation in HCF-V and blunted telo-collagen expression. When injected intraperitoneally into mice treated with isoproterenol, OSM-loaded SEVs reduced fibrosis, prevented cardiac hypertrophy, and increased angiogenesis. Overall, we demonstrate that the enrichment of functional OSM on the surface of MSC-T-SEVs increases their potency in terms of anti-fibrotic and pro-angiogenic properties, which opens new perspectives for this novel biological product in cell-free-based therapies.

Keywords: Oncostatin M; extracellular vesicles; fibrosis; isoproterenol; mesenchymal stem cells.

Figure 1

Generation of OSM-loaded MSC-T-SEVs through OSM sequence modification and fusion to CD81TM. (a) Structure scheme of native and recombinant proteins with introduced variants in the OSM sequence. The PP domain sequence was removed in mature OSM (matOSM), and a nucleotide mutation was introduced to change an R for a G in CS2 to block excision of matOSM from SEV membranes. The entire OSM sequence was maintained in mutant OSM (mutOSM), but the same R-to-G mutation in CS2 was introduced. Both matOSM and mutOSM sequences are anchored to CD81TM to favor their presence on SEV membranes. (b) General experimental procedure to obtain engineered MSC-T and SEV samples. Step 1: Cell culture preparation for lentiviral transfection. Lentiviruses containing CD81, matOSM-CD81, or mutOSM-CD81 were used to generate engineered MSC-T. Step 2: MSC-T expansion and incubation with EVs-free medium containing doxycycline for 48 h for supernatant collection and protein isolation from cell lysates. Step 3: SEV isolation from cell supernatants by ultracentrifugation and filtration. Fusion of two proteins is represented with a dash (-). Scissors represent where OSM cleavage occurs, while red crosses represent where the excision does not occur. OSM, Oncostatin M; SP, signal peptide; CS, cleavage site; PP, propeptide; R, arginine; G, glycine: MSC-T, immortalized mesenchymal stromal cells; EVs, extracellular vesicles; SEVs, small extracellular vesicles. Image was created using BioRender.

Figure 2

SEVs derived from genetically modified MSC-T cells carry the recombinant proteins mat/mutOSM-CD81 on their surface but maintain the same physical characteristics in terms of morphology and size distribution. (a) OSM detection in protein samples of cell lysates or SEVs from MSC-T (Ctrl) (lane 1), MSC-T-CD81 (CD81) (lane 2), MSC-T-matureOSM-CD81 (matOSM-CD81) (lane 3), and MSC-T-mutantOSM-CD81 (mutOSM-CD81) (lane 4). OSM was identified at both native molecular weight (22–24 kDa) and its expected molecular weight after fusion to CD81 (45–47 kDa). SEV protein extracts, 10 µg/lane; cell lysate protein extracts, 30 µg/lane. (b) Representative images of SEVs obtained by transmission electron microscopy (scale bar = 100 nm). (c,d) Nanoparticle tracking analysis, showing average particles size (nm) and particles size distribution histograms. (e) ExoView analysis showing percentage of OSM-, CD81-, or CD9-positive SEVs in samples after capture with a CD63 antibody. (f) Both the proportion of SEVs double-positive for CD63 and other proteins (OSM, CD81 and CD9) and triple-positive for these markers was analyzed in all SEVs experimental groups and plotted as pie charts. Each experiment was performed three times, and data were analyzed by ANOVA and Tukey’s post hoc test, using values from control cells or SEVs as the reference and represented as mean ± SEM (ns: non-significant).

Figure 3

Protein levels of the OSM receptors LIF-R/GP130 and IL-31RA are increased after HCF-V stimulation in vitro. (a) Representative Western blots for GP130, IL-31RA, and IL-31RB and β-tubulin (Tub) expression in HCF-V cells were measured under basal conditions (Ctrl), after starvation (Starved) and after starvation and stimulation with a pro-fibrotic cocktail containing L-ascorbic acid 2-phosphate, dextran sulphate, and recombinant TGFβ-1. Samples were collected both after starvation and after 0.5, 1, 2, 4, 24, and 48 h of starvation and fibrosis stimulation. Untreated cells in basal conditions were used as control. (b) GP130 protein expression quantification. (c) IL-31RB protein expression quantification. (d) IL-31RA protein expression quantification. Protein content was measured by densitometry in ImageJ 1.53t, using Tub as a loading control for each condition. Results were obtained from three independent experiments and were analyzed with ANOVA and Tukey’s post hoc test, using values from ctrl cells as a reference and represented as mean ± SEM (* p < 0.05; *** p < 0.001).

Figure 4

MutOSM-CD81-enriched SEVs have an enhanced ability to reduce HCF-V proliferation and telo-Collagen1α1 expression and secretion to the extracellular space. (a) Quantification of HCF-V proliferation rate under basal conditions (Ctrl), after starvation (starved), and after treatment with SEVs. The ratio between number of proliferative cells (Ki-67-positive nuclei, in green) and total number of cells (DAPI-positive nuclei, in blue) is shown. (b) Representative images of Ki-67 and DAPI nuclei staining for each experimental condition. (c,d) Quantification of telo-Collagen1α1 protein, which is shown in green. Mean fluorescence intensity (MFI) normalized to number of cells and extracellular telo-Collagen1α1 area are shown. (e) Representative images of telo-Collagen1α1 detection (shown in green) in samples under all experimental conditions. DAPI was used for nuclei staining (shown in blue). Scale bar = 200 µm. Both starved and stimulated cells were treated with equivalent volumes of vehicle (PBS). All results were obtained from three independent experiments and were analyzed with ANOVA and Tukey’s post hoc test. All experimental conditions were compared with both starved cells and starved cells plus ctrl-SEVs for statistical analysis in proliferation experiments. Values for stimulated cells and stimulated cells treated with ctrl-SEVs were used as a reference to compare with the other conditions for the in vitro fibrosis model and are represented as mean ± SEM (ns, non-significant differences; * p < 0.05, ** p < 0.01, and *** p < 0.001).

Figure 5

MutOSM-CD81 SEVs reduce collagen deposition in an isoproterenol (ISO) in vivo model of myocardial infarction type 2. (a) Schematic of the experimental procedure. Three experimental groups were included: ISO (ctrl), ISO + CD81-SEVs, and ISO + mutOSM-CD81-SEVs. Fibrosis was induced by a daily ISO subcutaneous (SC) injection (dose: 150 mg/kg) during 5 consecutive days. Two intraperitoneal (IP) doses of SEVs (CD81 or mutOSM-CD81) were injected (day 1 and day 7). An equivalent volume of vehicle (PBS) was injected in the ISO-only experimental group. (b) Representative images of Picrosirius red staining. Scale bar: 200 µm. (c) Quantification of fibrotic area in Picrosirius red-stained heart sections. The percentage of collagen-covered area related to total heart area is represented for each animal. (d) Representative images of periostin (POSTN) and α-smooth muscle actin (αSMA) double staining. Scale bar: 50 µm. (e) Quantification of perivascular fibrosis by POSTN staining in pixels and mean fluorescence intensity. Five animals were included in each group. ImageJ was used for image analysis and colorimetric and fluorescence quantification. Data were analyzed with ANOVA and Tukey’s post hoc test, using values from ctrl group (ISO) as reference conditions and represented as mean ± SEM. Data from CD81-SEVs and mutOSM-CD81-SEVs were also compared (ns, non-significant; * p < 0.05, ** p < 0.01, and *** p < 0.001).

Figure 6

MutOSM-CD81 SEVs reduce cardiac hypertrophy and increase angiogenesis in an isoproterenol (ISO) in vivo model of myocardial infarction type 2. (a) Representative images of wheat germ agglutinin (WGA) staining (shown in red) and nuclei staining (DAPI, shown in blue) for each experimental group. Scale bar: 50 µm. (b) Measurement of cardiomyocyte area, using WGA staining in heart sections of animals. (c) Immunodetection of F4/F80 (pan-macrophage marker; red) and CD274 (PD-L1; pro-inflammatory Mφ1 marker; green) or CD206 (pro-regenerative Mφ2 marker; green) in heart samples 21 days after ISO treatment. Scale bar: 50 μm. (d) Quantification of double-positive cells per mm². Five sections of 0.14 mm² per mouse were analyzed. (e) Measurement of microvasculature, using an antibody to CD31. Circular stained structures with an area up to 23.5 µm² were counted. (f) Representative images of CD31 staining (shown in red) and nuclei staining (DAPI, shown in blue) for each experimental group. Scale bar: 50 µm. Five animals were included in each group. ImageJ was used for image analysis and colorimetric and fluorescence quantification. Data were analyzed with ANOVA and Tukey’s post hoc test, using values from the control group (ISO) as reference conditions and represented as mean ± SEM. Data from CD81-SEVs and mutOSM-CD81-SEVs were also compared (ns, non-significant statistic differences; * p < 0.05, ** p < 0.01, and *** p < 0.001).