(489f) A Large-Scale, Real-Time Array to Assess Dynamic Changes in Intracellular Signaling in Response to Biomaterial-Mediated Mechanical and Adhesive Stimuli

The rigidity and density of adhesion sites of cell culture substrates strongly influences cell phenotype and gene expression. Interactions between integrin receptors and the substrate mediate these phenomena at the cell membrane, but previous studies have focused on a limited number of intracellular signals that relay these physical cues to the nucleus. To investigate the signaling pathways that determine the intrinsic response of cell to its surroundings, we have applied a novel systems biology approach for large scale, dynamic quantification of transcription factor (TF) activity, which represents the output of complex intracellular signaling networks in response to various extrinsic factors including substrate mechanics and ligand presentation.

Specifically, we have designed polyethylene glycol (PEG)-based hydrogels in which the adhesive ligand density (e.g., RGD) and mechanical properties can be independently tuned across three orders of magnitude (0.15-10 kPa). Human foreskin fibroblasts (HFFs, ATCC CRL-2522) were cultured on these hydrogels where modulus and RGD concentration were independently varied. Prior to seeding on PEG hydrogels, HFFs were transduced by lentiviral reporter constructs in which luciferase expression is driven by consensus sequences known to bind certain TFs. Luciferase levels were then measured non-invasively by bioluminescence imaging of HFFs cultured in a 384-well plate format, providing a platform with rapid data acquisition to quantify TF activity. Measurements were taken at 3, 6, 9, 12 and 27 h. after HFF seeding. The integrated photon flux values for each well were normalized by basal levels of luciferase production, as measured in cells transduced by a reporter with the TATA box promoter driving luciferase production but lacking a TF binding element in the enhancer region. The dynamic activities of 53 TF reporter constructs were monitored over a 27 hr period after HFFs were seeded.

When substrate mechanics were varied, 22 reporters showed significantly different activities. These reporters represent the activity transcription factors previously identified to play a role in mechanotransduction, cell adhesion and migration (e.g., STATs, APs, GATAs) in addition to several factors that have not been previously linked to these events (e.g., hormonal regulators, Notch, p53). Furthermore, the effects of mechanic stiffness were TF dependent, observing activation and inhibitory tendencies on the patterns of dynamic activity, i.e., some factors increased only in HFFs cultured on the softest substrates while other increased only in cells on the hardest substrates.

When RGD density was varied, we have identified 8 reporters with significantly different activities dependent on ligand concentration. Of these, only 3 TF reporters (Gl1, MyB1 and TCF/LEF) displayed dependence on both substrate modulus and ligand density. While this is an indication of a cross-talk between the signaling pathways altered by environmental stiffness and ligand density, yet they required additional, non-overlapping signaling processes to transmit their full phenotypic effects. Gl1 and MyB1 are involved in cell cycle regulation, but have not previously been reported to play a role in cell-matrix interactions. The activities of both were significantly higher when cultured on the stiffest substrate or on substrates presenting the greatest density of RGD ligands. TCF and LEF are co-transcriptional activators that interact with ß-catenin and have been directly linked to mechanotransduction via cadherins at cell-cell junctions. In our studies, TCF/LEF activity was highest when HFFs were cultured on the stiffest substrates, but was lowest on substrates with the greatest RGD ligand densities.

These studies provide a more holistic view of the signaling pathways that determine the cellular response to mechanical and adhesive stimuli and can provide a mechanistic connection between biomaterial properties and cell phenotype. Our analysis suggests that cell phenotype can be differentially regulated by these two types of extracellular cues. These results provide valuable insight into the relative contributions of mechanical and chemical cues and the cross-talk between intracellular signaling pathways in response to these cues. Furthermore, these results are a significant step forward in our understanding of the underlying mechanisms that dictate the relationship between biomaterial design and cell phenotype and they might help to improve the design of biomaterials for regenerative medicine applications.