Abstract
Dendritic cell (DC) migration is required for efficient presentation of antigen to T cells and the initiation of an adaptive immune response. In spite of its importance, many aspects of DC migration have not been characterized. DCs encounter a variety of environments with different stiffness and geometry, but the effect of these parameters on DC migration has not yet been determined. We addressed this question by comparing DC motility on standard migration surfaces (polydimethylsiloxane (PDMS)-coated coverslips) and micropost array detectors (mPADs). These two surfaces differ in both stiffness and geometry. We found that DC migration was affected by substrate type, with significant increases in speed and significant decreases in persistence time on mPADs made of PDMS as compared to spin-coated PDMS coverslips. To determine whether the geometry or compliance of the post arrays was responsible for these changes in DC migration, we quantified DC motility on mPADs of identical geometry but different stiffness. Migration was indistinguishable on these mPADs, suggesting that DCs are responsive to geometry of ligand presentation and not stiffness. Further, by micropatterning ligands on flat PDMS surfaces in similar geometries to the mPAD arrays, we determined that DCs respond to the geometry of printed ligand. Finally, we used a variety of small molecule inhibitors to identify pathways involved in geometry sensing. We saw a significant role for myosin contractility and α5β1 integrin engagement. We also noted significant reorganization of the actin cytoskeleton into dynamic actin rings when DCs were motile on posts. From these experiments, we conclude that DCs are insensitive to substrate compliance in the range tested but respond to changes in geometry via a mechanism that involves integrin function, myosin contractility, and remodeling of the actin cytoskeleton. As a possible explanation, we postulate a consistent role for filopodial extension and contraction as the driver of DC motility.