Insulin-Like Growth Factor 2 Secreted from Mesenchymal Stem Cells with High Glutathione Levels Alleviates Osteoarthritis via Paracrine Rejuvenation of Senescent Chondrocytes

Abstract

Senescent chondrocytes, which are increased in osteoarthritic (OA) cartilage, promote cartilage defects and the senescent knee microenvironment by inducing senescence to surrounding normal chondrocytes by secreting senescence-associated secretory proteins. Many studies have used mesenchymal stem cells (MSCs) to treat OA, but MSC treatment remains challenging for clinical application owing to MSC quality control, engraftment, and fibrocartilage regeneration. Here, rather than relying on the direct regeneration of MSCs, we present a novel strategy to suppress OA by MSC-mediated senescent chondrocyte targeting via the paracrine activity of MSCs, thereby improving the knee microenvironment. First, to enable quality control of umbilical cord MSCs, priming MSCs by supplementing human platelet lysate (hPL) greatly enhanced MSC functions by increasing cellular glutathione levels throughout serial passaging. Intra-articular injection of primed MSCs successfully suppressed OA progression and senescent chondrocyte accumulation without direct regeneration. Indirect coculture with primed MSCs using transwell ameliorated the senescence phenotypes in OA chondrocytes, suggesting paracrine rejuvenation. Based on secretome analysis, we identified insulin-like growth factor 2 (IGF2) as a key component that induces paracrine rejuvenation by primed MSCs. The rejuvenation effects of IGF2 act through autophagy activation through the up-regulation of autophagy-related gene expression and autophagic flux. To cross-validate the effects of secreted IGF2 in vivo, knockdown of IGF2 in primed MSCs substantially abolished its therapeutic efficacy in a rabbit OA model. Collectively, these findings demonstrate that hPL supplementation enables MSC quality control by increasing MSC glutathione levels. The therapeutic mechanism of primed MSCs was secreted IGF2, which induces paracrine rejuvenation of senescent OA chondrocytes by activating autophagy.

Fig. 1.

Generation of high functional MSCs using hPL. (A) Representative flow cytometric analysis of cellular GSH levels of naïve, primed MSC (P5 and P6), GSH contents (B), and GSH heterogeneity (C) of P6 primed MSCs determined by flow cytometry (n = 3 per group). (D) mRNA expression of stem cell markers in the naïve and primed MSCs measured by RT-qPCR (n = 3 per group). (E) Colony-forming assay in naïve and primed MSCs (n = 5 per group). (F) Evaluation of oxidant resistance of naïve and primed MSCs treated with indicated concentration of diamide and (G) the graph showing the percentage of decrease in GSH-high population upon diamide treatment in naïve and primed MSCs (n = 3 per group). (H) Left: Representative image of adipogenic (Oil Red O), osteogenic (Alizarin Red S), and chondrogenic (Alcian Blue) differentiation analysis of naïve and primed MSCs. Right: Quantification using absorbance at 510 nm for Oil Red O, 405 nm for Alizarin Red S, and 620 nm for Alcian Blue (n = 3 per group). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 2.

Intra-articular injection of primed MSCs suppresses OA progression without evidence of direct regeneration. (A) Representative histological analysis using Safranin-O and immunohistochemistry (IHC) (collagen type II, MMP13, collagen type X, DPP4, and human β2-microglobulin) in sham control and DMM-induced rabbits injected with PBS and naïve or primed MSCs. (B) Scoring of OA using OARSI grading system in sham control and DMM-induced OA rabbits that had undergone intra-articular injection of PBS and naïve or primed MSCs. (C to F) Quantification of collagen type II, MMP13, collagen type X, and DPP4 using scoring system. (G) Representative histological analysis using Safranin-O and DPP4 IHC staining in DMM-induced OA rabbits injected with PBS, primed MSCs, low-dose, mid-dose, or high-dose secretome, or triple high-dose secretome isolated from primed MSCs. (H and I) Scoring of OA using OARSI grading system and quantification of DPP4. **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 3.

Paracrine activity of primed MSC suppresses senescence phenotypes of OA chondrocytes. (A) Left: Representative immunofluorescence staining of p16 and p21 (red) in OA chondrocytes (OA), OA + naïve MSC media, OA + MSC priming media, and direct coculture of OA and naïve or primed MSCs. OA chondrocytes were distinguished with CellTracker Dil staining (green), and nuclei were counterstained with DAPI (blue). Right: Bar graphs showing quantification of p16⁺ OA chondrocytes (top) and quantification of p21⁺ OA chondrocytes (n = 3 per group) (bottom). (B) Schematic illustration of the experiments. (C) mRNA levels of p16 and p21 in OA, OA + naïve CM, and OA + primed CM measured by RT-qPCR. (D) Representative SA-β-Gal staining and quantification for OA, OA + naïve CM, and OA + primed CM (n = 3 per group). (E) Cell counts of OA, OA + naïve CM, and OA + primed CM for 14 d to evaluate proliferation (n = 3 per group). (F) mRNA expression of matrix degradation factors (ADAMTS5 and MMP13) in OA, OA + naïve CM, and OA + primed CM (n = 3 per group). (G) Schematic illustration of indirect coculture system using transwell. (H) mRNA levels of p16 and p21 in OA, OA + naïve MSCs, and OA + primed MSCs measured by RT-qPCR. (I) Representative SA-β-Gal staining and quantification for OA, OA + naïve MSCs, and OA + primed MSCs (n = 3 per group). (J) Cell counts of OA, OA + naïve MSCs, and OA + primed MSCs for 14 d to evaluate proliferation (n = 3 per group). (K) mRNA expression of matrix degradation factors (ADAMTS5 and MMP13) in OA, OA + naïve MSCs, and OA + primed MSCs (n = 3 per group). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 4.

IGF-2, highly expressed and secreted in primed MSCs, is a key molecule that rejuvenates senescent chondrocytes. (A) Secretome profile of naïve and primed MSCs and (B) bar graph showing top 30 secreted proteins increased in primed MSCs (n = 3 per group). (C) Schematic illustration of the experiments. (D) Percentage of BrdU⁺ chondrocytes treated with the 30 recombinant proteins (n = 3 per group). (E) mRNA expression (top) and protein secretion (bottom) of IGF2 in naïve and primed MSCs determined by RT-qPCR and ELISA, respectively (n = 3 per group). Representative immunofluorescence image showing active caspase-3 (F), Ki67, and DPP4 (G) in OA and OA chondrocytes treated with 200 ng/ml IGF2 and (H to J) quantification of active caspase-3, DPP4, and Ki67-positive chondrocytes (n = 3 per group). (K) Representative flow cytometry of p16-phycoerythrin (PE)- and Ki67-PE-stained DPP4-negative and DPP4-positive chondrocytes treated with or without 200 ng/ml IGF2 and quantification of p16- and Ki67-positive chondrocytes (n = 3 per group). (L) Schematic illustration of the experiments. (M) mRNA levels of p16 and p21 in OA, OA + primed MSCs, and OA + IGF2 knockdown primed MSCs measured by RT-qPCR (n = 3 per group). (N) (left) Representative Western blot image and (right) quantification of p16 and p21 in OA, OA + primed MSCs, and OA + IGF2 knockdown primed MSCs (n = 3 per group). (O) Proliferation measurement by cell counting in OA, OA + primed MSCs, and OA + IGF2 knockdown primed MSCs (n = 3 per group). **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 5.

IGF-2 rejuvenates senescent OA chondrocytes via autophagy activation. (A) mRNA expression of autophagy-related genes (ATG5, BECN1, ULK1, SIRT1, and LC3B) in non-OA, OA, and OA chondrocytes treated with 200 ng/ml IGF2 (n = 3 per group). (B) Representative Western blot image (left) and quantification of autophagy-related proteins (ATG5, BECN1, ULK1, SIRT1, and LC3B) (right) in non-OA, OA, and OA chondrocytes treated with 200 ng/ml IGF2 (n = 3 per group). (C) Representative immunostaining of LC3B (green) in OA and OA chondrocytes treated with 200 ng/ml IGF2 with or without HCQ (60 μM) and quantification of (D) autophagic flux and (E) autophagosome formation by measuring autophagic flux index and mean fluorescence intensity (MFI), respectively (n = 3 per group). (F) Proliferation measurement by cell counting in OA, OA + 200 ng/ml IGF2, and OA + 200 ng/ml IGF2 + 5 mM 3-MA (n = 3 per group). (G) mRNA expression of p16 and p21 in OA, OA + 200 ng/ml IGF2, and OA + 200 ng/ml IGF2 + 5 mM 3-MA (n = 3 per group). (H) Representative Western blot image (left) and quantification (right) of p16 and p21 in OA, OA + 200 ng/ml IGF2, and OA + 200 ng/ml IGF2 + 5 mM 3-MA (n = 3 per group). (I) Representative images of SA-β-Gal (left) and quantification (right) of SA-β-Gal cells in OA, OA + 200 ng/ml IGF2, and OA + 200 ng/ml IGF2 + 5 mM 3-MA (n = 3 per group). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 6.

Knockdown of IGF2 expression in primed MSCs using siRNA inhibits therapeutic efficacy in vivo. (A) Representative histological analysis using Safranin-O and IHC (collagen type II, MMP13, collagen type X, and DPP4) in sham control and DMM-induced rabbits injected with PBS, primed MSCs, or primed MSCs with IGF2 knockdown (IGF2KD MSCs). (B) Scoring of OA using OARSI grading system in sham control and DMM-induced OA rabbits that had undergone intra-articular injection of PBS, primed MSCs, or IGF2KD MSCs. (C to F) Quantification of collagen type II, MMP13, collagen type X, and DPP4 using scoring system. **P < 0.01; ***P < 0.001; ****P < 0.0001.