Biophysical Cues Modulate the Anti-Fibrotic Response of Rabbit Corneal Stromal Cells | AIChE

Biophysical Cues Modulate the Anti-Fibrotic Response of Rabbit Corneal Stromal Cells

Authors 

Murphy, C., School of Medicine, University of California Davis
During corneal wound healing, changes to the cellular microenvironment involve alterations in the expression of soluble signaling molecules [e.g. transforming growth factor (TGFβ)] as well as modulation of the biophysical attributes of the corneal stroma. These stimuli mediate the transformation of quiescent Keratocytes to activated Fibroblasts and further to contractile Myofibroblasts (KFM transition). Although, myofibroblast cells are critical for effective wound repair, excessive numbers and/or prolonged residence of these cells in the cornea can lead to stromal haze and fibrosis. Heat Shock Protein 90 (HSP90), a chaperone protein, stabilizes the TGFß1 pathway that is critical in KFM transformation. We hypothesized that inhibiting HSP90 would prevent the fibrotic phenotype of rabbit corneal stromal cells in vitro. This hypothesis further implies that HSP90 inhibition may minimize stromal haze in vivo. That biophysical stimuli such as topography and stiffness can mediate phenotypic alterations of corneal stromal cells is well established. However, the effect of biophysical parameters on governing the effect of drugs is lesser known. Here, primary rabbit corneal stromal fibroblasts (RCF) were cultured on both topographically patterned substrates [planar control and alternating ridges and grooves of 1400 nm pitch stamped into polyurethane (NOA81); pitch = groove + ridge width] as well as substrates having different stiffness [Tissue Culture Plastic (TCP; ~1GPa), as well polyacrylamide gels of 5kPa and 25 kPa moduli]. These cells were treated in the presence and/or absence of 10 ng/mL TGFß1, and 500 nM 17-(Allylamino)-17-demethoxygeldanamycin (17-AAG; an Hsp90 inhibitor) for 72 hours. Protein and gene expression were quantified by Western blotting and qPCR, respectively. Immunocytochemistry was also used to visualize a myofibroblast marker, alpha smooth muscle actin (αSMA), of treated RCF cells. On planar NOA81 surfaces and on stiff TCP surfaces, TGFß1 treatment promoted the myofibroblast phenotype while 17-AAG promoted conversion of activated fibroblasts to the keratocyte phenotype, in the presence or absence of TGFß1. Interestingly, 25 kPa and 1400 nm pitch patterned substrates inhibited development of the myofibroblast phenotype in the presence or absence of TGFß1. However, for 17-AAG-treated cells, both 5 kPa and 1400 nm pitch substrates promoted development of the myofibroblast phenotype. Our data suggest that HSP90 inhibition may improve corneal wound healing outcomes but could have adverse consequences in normal unwounded corneas.