Fig. 1
The microfluidic T-maze. a A microfluidic T-maze device. The scale bar shown in the figure is 2 mm. b By using source and buffer channels running parallel to the maze to generate a steady chemoattractant gradient within the hydrogel base, the device creates concentration gradients that have the same magnitude through consecutive junctions (here shown for log-sensing). c Schematic of a mixed population of cells swimming through the maze. Better chemotaxers (yellow) become increasingly concentrated at sequential junctions. d Schematic of the device and e cross-section (corresponding to the red dashed line in (d)). Flow of the chemoattractant through the source channel generates a gradient in the hydrogel, and hence also along each T-junction section (vertically oriented channel segments in (d)) of the maze channel. A droplet containing cells is placed at the inlet shown in (d), and cells swim from left to right through the maze, making multiple chemotactic decisions through the consecutive junctions
