(339a) Fibroblasts Affect Each Other’s Directional Decision Making Process during Chemotaxis in Microfluidic Tissue Mimicking Mazes

Fibroblast chemotaxis within complex micro-architectures, such as tissue engineering scaffolds, is critical to successful biomanufacturing of complex organs (e.g., the heart, liver, and kidney). Current chemotaxis assays of fibroblasts (i.e. transwell membrane, gel migration, and microfluidic-assisted one-dimensional gradient generator) either lack all the dynamic features of the cell migration such as decision-making, speed, and persistence or they are too simple to represent actual microenvironments experienced by the cells inside tissue scaffolds. To better understand/control migration of fibroblasts within those scaffolds, we performed a chemotaxis-based assay in 2D microfluidic mazes which are meant to mimic complex tissue pore geometries by providing cells with navigation choices such as long paths, short paths, and dead ends in the presence of a complex chemotactic gradient.

The mazes consist of 24 µm-wide channels which connect a cell seeding compartment to a media reservoir containing chemoattractant of different concentrations. By inducing migration of NIH 3T3 fibroblasts - a cell type common to tissue engineering, by a chemoattractant PDGF-BB gradient formed inside of the mazes, we were able to track and characterize various important cell migration parameters, such as directional decision-making, speed and persistence. Due to the small size of the channel, cells entered the maze in sequence, one-by-one. In the presence of 15 to 50 ng/ml of PDGF-BB, the fibroblasts could migrate steadily across the maze to the source of chemoattractant with an average migration speed of 1 - 3 µm/min and without reversing the direction. This is in contrast to the absence of chemoattractant, which resulted in no cells navigating into the maze. Moreover, the fibroblasts tended to reduced speed when they were about to reach the chemoattractant source.

It was found that, given the choice, the cells preferred not to follow each other and made opposite directional decisions when choosing paths through the maze: if the leading cell selected the short path, the following cell chose the long one and vice versa. Moreover, this trend occurred at high probability regardless of the chemoattractant gradient magnitude in the maze. Secondly, the chemoattractant levels were found to influence the directional decision-making process of the cells: while the concentration of 15 ng/ml brought about a bias towards the short path, a chemoattractant concentration of 35 ng/ml or higher resulted in lack of preference towards any single direction. This is likely because at low chemoattractant concentration, directional choices is dominated by a steeper gradient, while at higher chemoattractant concentration, the difference in the gradients between the long and the short path becomes indistinguishable to the migratory cells.

Finally, even though migrating cells displayed frequent change in shape and length, they were more likely to be aligned along the migration direction in a head-tail morphology. This helps to predict the cell directionality, thus allowing one to control cell migration at high temporal resolution. The results of this study will be useful for future scaffold design of complex tissues and organs.