A New Microfluidic Platform for Studying Natural Killer Cell and Dendritic Cell Interactions

The importance of the bi-directional natural killer-dendritic cell crosstalk in coordinating anti-tumour and anti-microbial responses in vivo has been well established. However, physical parameters associated with natural killer-dendritic cell interactions have not been fully elucidated. We have previously used a simple "Y" shaped microfluidic device to study natural killer cell-migratory responses toward chemical gradients from a conditioned medium of dendritic cells. There are, however, limitations of the Y-shaped microfluidic devices that could not support higher throughput analyses and studies of cell-cell interactions. Here, we report two novel microfluidic devices (D³-Chip, T²-Chip) we applied in advanced studies of natural killer-cell migrations and their interactions with dendritic cells in vitro. The D³-Chip is an improved version of the previously published Y-shaped device that supports high-throughput analyses and docking of the cells of interest in the migration assay before they are exposed to a chemical gradient. The T²-Chip is created to support analyses of natural killer-dendritic cell cell-cell interactions without the requirement of promoting a natural killer cell to migrate long distances to find a loaded dendritic cell in the device. Using these two microfluidic platforms, we observe quantitative differences in the abilities of the immature and lipopolysaccharide-activated mature dendritic cells to interact with activated natural killer cells. The contact time between the activated natural killer cells and immature dendritic cells is significantly longer than that of the mature dendritic cells. There is a significantly higher frequency of an immature dendritic cell coming into contact with multiple natural killer cells and/or making multiple simultaneous contacts with multiple natural killer cells. To contrast, an activated natural killer cell has a significantly higher frequency of coming into contact with the mature dendritic cells than immature dendritic cells. Collectively, these differences in natural killer-dendritic cell interactions may underlie the differential maturation of immature dendritic cells by activated natural killer cells. Further applications of these microfluidic devices in studying natural killer-dendritic cell crosstalk under defined microenvironments shall enrich our understanding of the functional regulations of natural killer cells and dendritic cells in the natural killer-dendritic cell crosstalk.