Dichotomous SMAD2/3 regulation and selective anti-hypertrophic activity of heparin during in vitro chondrogenesis of mesenchymal stromal cells

Background Endochondral instead of chondral differentiation hinders mesenchymal stromal cell (MSC) application for clinical cartilage regeneration. We previously showed that heparin-polyethylene glycol (PEG) hydrogels loaded with transforming growth factor beta (TGF-β) instructed stable chondral MSC development in vivo. We here assessed this approach in vitro, utilizing heparin-PEG hydrogels or soluble heparin supplementation of chondrogenic medium. Methods Human MSCs were cultured in heparin-PEG hydrogels (22.4 mg/mL crosslinked heparin, 120ng TGF-β1) or as hydrogel-free pellet cultures treated with soluble heparin (0, 10, 100, 700 μg/mL) in TGF-β1-containing (10 ng/mL) chondrogenic medium. Chondral and endochondral signaling (1-3 h, 4 weeks) and cartilage matrix formation (4 weeks) were analyzed using Western blot, histology, qPCR, ELISA, and enzyme activity. Results Unlike in vivo, human MSCs differentiated in heparin-PEG hydrogels into type X collagen and alkaline phosphatase-positive hypertrophic chondrocytes in vitro. Interestingly, treatment with soluble heparin (10-700 μg/mL) revealed reduced TGF-β-small mother against decapentaplegic (SMAD)3 but not SMAD2 activation at unaffected type II collagen and proteoglycan/DNA levels. We propose that the stimulation of the insulin-AKT pathway by heparin aided in maintaining SMAD2 activation which apparently plays a more prominent role than SMAD3 for MSC chondrogenesis. Heparin treatment inhibited the pro-hypertrophic WNT/β-catenin pathway in vitro but insufficiently silenced TGF-β-SMAD1/5/9 activation and unfortunately reduced anti-hypertrophic prostaglandin E2 (PGE2) levels. Ultimately, treatment with 10 μg/mL heparin reduced the upregulation of several hypertrophy markers ( MEF2C , IHH , IBSP mRNAs, alkaline phosphatase activity) below control levels, but type X collagen remained unresponsive. Thus, soluble heparin treatment was similarly selective and effective as previous anti-hypertrophic interventions (parathyroid-hormone related protein (PTHrP)-pulses, wingless-int (WNT)-inhibition), while offering technical simplicity, reduced cost, and solvent-free formulation. Conclusions Taken together, heparin-TGF-β showed a novel dichotomous SMAD2/3 inhibition at maintained chondrogenic power and context-dependent lineage-instructive properties: permitting endochondral differentiation in vitro but chondral development in vivo. Thus, environmental contributions are mandatory to allow heparin-PEG-guided chondral versus endochondral lineage commitment of MSCs in vivo, potentially involving SMAD1/5/9 suppressors and PGE2 sources.