Fibroblast growth factor-2 primes human mesenchymal stem cells for enhanced chondrogenesis

Abstract

Human mesenchymal stem cells (hMSCs) are multipotent cells capable of differentiating into a variety of mature cell types, including osteoblasts, adipocytes and chondrocytes. It has previously been shown that, when expanded in medium supplemented with fibroblast growth factor-2 (FGF-2), hMSCs show enhanced chondrogenesis (CG). Previous work concluded that the enhancement of CG could be attributed to the selection of a cell subpopulation with inherent chondrogenic potential. In this study, we show that FGF-2 pretreatment actually primed hMSCs to undergo enhanced CG by increasing basal Sox9 protein levels. Our results show that Sox9 protein levels were elevated within 30 minutes of exposure to FGF-2 and progressively increased with longer exposures. Further, we show using flow cytometry that FGF-2 increased Sox9 protein levels per cell in proliferating and non-proliferating hMSCs, strongly suggesting that FGF-2 primes hMSCs for subsequent CG by regulating Sox9. Indeed, when hMSCs were exposed to FGF-2 for 2 hours and subsequently differentiated into the chondrogenic lineage using pellet culture, phosphorylated-Sox9 (pSox9) protein levels became elevated and ultimately resulted in an enhancement of CG. However, small interfering RNA (siRNA)-mediated knockdown of Sox9 during hMSC expansion was unable to negate the prochondrogenic effects of FGF-2, suggesting that the FGF-2-mediated enhancement of hMSC CG is only partly regulated through Sox9. Our findings provide new insights into the mechanism by which FGF-2 regulates predifferentiation hMSCs to undergo enhanced CG.

Figure 1. FGF-2 enhances hMSC CG.

Human MSCs were expanded in varying concentrations of FBS with or without FGF-2 for 2 passages prior to CG and analyzed for cartilaginous properties upon differentiation in pellet culture. (A) End-point RT-PCR analysis showed that cartilage-specific gene expression was upregulated in the 2 FGF-2-expanded conditions (Lanes 2 and 4) at each time point compared to those without FGF-2 pretreatment (Lanes 1 and 3). Lane 1: 10% FBS; Lane 2: 5% FBS+FGF-2; Lane 3: 5% FBS; Lane 4: 1% FBS+FGF-2. (B) Significantly greater GAG accumulation was observed in the 2 FGF-2-expanded conditions at each time point compared to those without FGF-2 pretreatment. GAG content was normalized to DNA content. Values are mean ± SD. All conditions were significant with each other except the 2 FGF-2-expanded conditions at Day 21. p<0.05, n = 3.

Figure 2. FGF-2 reduces potency and modulates lineage-specific transcription factor expression of hMSCs.

Human MSCs were expanded for 2 passages in 5% FBS with (+) or without (−) FGF-2 supplementation and assayed for gene and protein expression prior to pellet culture. (A, B, C) Real-time RT-PCR analysis demonstrated that FGF-2-expanded hMSCs significantly downregulated the expression of embryonic stem cell markers (A), modulated the expression of key lineage-specific transcription factors (B), and upregulated the expression of L-Sox5 (C). All values were normalized to GAPDH expression and compared to the non-FGF-2-expanded control. Values are represented as mean ± SD. *p<0.05, **p<0.005, n = 3. (D) Western blot analysis showed that FGF-2-expanded hMSCs exhibited higher Sox9 protein levels than non-FGF-2-expanded hMSCs prior to CG, and after pellet formation, pSox9 protein levels were present at higher levels. GAPDH was used as a loading control.

Figure 3. FGF-2 enhances hMSC CG through both selection- and priming-based mechanisms.

Human MSCs were expanded for 2 passages in medium supplemented with FGF-2 for varying durations of time and assayed for cartilage-specific gene expression in pellet culture up to 21 days using real-time RT-PCR. Longer durations of FGF-2 exposure during hMSC expansion resulted in the significant upregulation of L-Sox5 and Col II gene expression, while exposure to FGF-2 for just 2 hours resulted in the significant upregulation of Sox9 and AGN by Day 21. All values were normalized to GAPDH expression and compared to the non-FGF-2-expanded control at Day 7. Values are represented as mean ± SD. *p<0.05, **p<0.005, n = 3.

Figure 4. FGF-2 primes hMSCs for CG by increasing basal Sox9 protein levels.

(A) Human MSCs were exposed to FGF-2 for 30 minutes or 2 hours and analyzed for Sox9 protein levels prior to pellet culture by Western blot analysis. Exposure to 5 and 50 ng/mL FGF-2 led to increased Sox9 protein levels in a dose-dependent manner. GAPDH was used as a loading control. (B) Human MSCs were exposed to FGF-2 for 2 hours and chondrogenically-induced in pellet culture. FGF-2-expanded hMSCs (+) showed elevated pSox9 protein levels by Western blot analysis. GAPDH was used as a loading control. (C) Human MSCs were treated with or without SB431542 for 15 minutes, exposed to combinations of 5 ng/mL FGF-2, 10 ng/mL TGF-β1 and 10 µM SB431542 for 2 hours, and analyzed for Sox9 mRNA and protein levels by end-point PCR and Western blot analysis, respectively. Exposure to FGF-2 led to increased Sox9 mRNA and protein levels, while TGF-β1 reduced Sox9 expression. Smad7 gene and pSmad3 protein levels were analyzed to confirm successful knockdown of TGF-β signaling. GAPDH was used as a loading control. (D) Human MSCs were expanded in FGF-2 for varying durations of time and analyzed for Sox9 protein using flow cytometry. Sox9 protein levels per cell were elevated at all time points after FGF-2 exposure compared to the control and gradually increased with longer exposures. Values are represented as fluorescence signal to noise ratios using geometric means.

Figure 5. FGF-2 elevates Sox9 protein levels in both proliferating and non-proliferating hMSCs.

Human MSCs were exposed to FGF-2 for 12 and 24 hours in the presence of BrdU, double-stained for Sox9 and BrdU, and analyzed using flow cytometry. (A) FGF-2 exposed hMSCs showed a higher percentage of BrdU⁺ cells than non-FGF-2-exposed hMSCs after 12 and 24 hours. (B) Sox9 protein levels were elevated in both BrdU⁺ and BrdU⁻ FGF-2-exposed hMSCs after 12 and 24 hours, with BrdU⁺ hMSCs showing more marked differences compared to non-FGF-2-exposed control hMSCs. Values are represented as fluorescence signal to noise ratios using geometric means.

Figure 6. FGF-2 enhances hMSC CG partially through a Sox9-mediated mechanism.

(A) RNAi experiments were performed to knockdown Sox9 expression. Human MSCs were transfected with siRNA targeted to Sox9 or non-targeting siRNA. 24 hours after transfection, hMSCs received FGF-2, and 72 hours after transfection, hMSCs were differentiated into the chondrogenic lineage in pellet culture. Sox9 knockdown was verified 48 and 72 hours after transfection by gene and protein analysis, respectively. (B) Western blot analysis showed that Sox9 levels were lower in both FGF-2- (+) and non-FGF-2-treated (−) hMSCs transfected with Sox9 siRNA compared to untransfected and non-targeting siRNA-transfected hMSCs. With each treatment, FGF-2 increased Sox9 levels. (C) After transfected hMSCs were differentiated into the chondrogenic lineage, real-time RT-PCR analysis showed that knockdown of Sox9 expression did not result in a concomitant decrease in Col II gene expression 7 and 21 days after pellet formation. However, FGF-2 pretreatment led to a significant increase in Col II gene expression in untransfected and Sox9 siRNA-transfected hMSCs.