Introduction: Nearly all synthetic biomaterials present an elastic, mechanically static environment to cells, despite tissues being viscoelastic and displaying dynamic stiffening during development and disease progression (e.g., liver fibrosis). Here we investigated cellular mechanotransduction using elastic hyaluronic acid (HA) hydrogels that permitted in situ stiffening in the presence of cells as well as hydrogels that combined covalent and supramolecular crosslinking to form viscoelastic gels also amenable to dynamic mechanical tuning.
Results: HA was modified with methacrylate (Me), β-cyclodextrin (CD), and/or adamantane (Ad) groups. MeHA was photocrosslinked to create elastic, covalently crosslinked gels. CDMeHA was mixed with AdHA and then photocrosslinked to create viscoelastic gels containing reversible interactions between Ad and CD groups , as well as covalent crosslinks between the CDMeHA chains only. Hepatic stellate cells (HSCs, main liver cell in fibrosis) isolated from rats and cultured on elastic hydrogels responded to dynamic changes in substrate stiffness, although stiffening within the first week of culture resulted in similar myofibroblast phenotype after 14 days (> 15000 μm2 spread area, YAP nuclear/cytoplasmic intensity ratio > 1.5). To more accurately capture native tissue properties, we developed viscoelastic hydrogels [2,3] that combine covalent and physical interactions. Elastic and viscoelastic hydrogels with variable crosslinking dynamics but equivalent HA content and elastic moduli were fabricated (E ~ 1, 5, and 12 kPa for low, mid, and high relative crosslinking). Viscoelastic gels could be stiffened with UV exposure during secondary photocrosslinking from E ~ 1 to ~ 10 kPa. Loss moduli (G”) were consistently an order of magnitude higher for viscoelastic over elastic gels with equivalent storage moduli (G’). Viscoelastic gels also displayed stress relaxation of ~ 20% for an initial strain, which was not observed in elastic gels. While human mesenchymal stem cells (hMSCs) displayed similar spread areas between gels with higher crosslink densities (~ 4000-5000 μm2), hMSCs were significantly more spread on viscoelastic gels (1413 ± 135 μm2) compared to elastic gels (879 ± 64 μm2, p < 0.001) with low crosslinking (E ~ 1 kPa).
Conclusions: We developed a hydrogel platform based on control over HA crosslinking to fabricate cell culture substrates with tunable mechanics (stiffness, viscoelasticity). In situ stiffening promoted HSC spreading and acquisition of a myofibroblast-like phenotype on elastic substrates. We then introduced a fabrication method for viscoelastic gels that displayed stress relaxation and significantly higher loss moduli than equivalent elastic gels. For soft gels of equivalent moduli hMSCs displayed significantly higher spread area on viscoelastic gels. Ongoing work is evaluating the role of ligand mobility in viscoelastic systems on cellular mechanotransduction. More broadly, the hydrogels developed here are well-defined systems for understanding mechanotransduction and identifying targets for therapeutic intervention in disease.
References: 1) Rodell CB, Biomacromolecules, 2013, 14(11): 4125-34; 2) Cameron AR, Biomaterials, 2011, 32(26): 5979-93; 3) Chaudhuri O, Nat Commun, 2015, in press.