US scientists are making bacteria sort themselves by size.

Sorting cells is a key task in cell and systems biology. Although several cell-sorting methods exist, they subject the cells to chemical or mechanical stresses to separate them. George Whitesides' group at Harvard University, Cambridge, US, has taken a different approach to the problem. The team takes advantage of Escherichia coli cells' motility to sort them by size in a microfluidic device. 

In fluid, E. coli cells take random walks involving a period of swimming followed by tumbles that change their direction randomly. But without a chemical stimulus, the average displacement of a population of swimming cells is zero. To make the cells swim in one direction the team's device contains bacterial ratchets. This series of arrowhead shaped channels redirects cells swimming in the 'wrong' direction, so that eventually most cells are swimming in one 'right' direction.

A series of ratchets encourages bacteria to travel in one direction. Bacteria entering the ratchets from the left pass through to the right (left images), whilst those entering from the right are redirected back (right images)

Once the cells have passed through the ratchets, the next part of the device, the bacterial sorter, sorts the cells by size. This exploits the fact that shorter cells follow a curved microchannel better than longer cells, allowing short cells to be separated from the population.

Group member Elizabeth Hulme explains the biological use for the technique: 'For E. coli, length is directly related to age: short cells are young cells, long cells are old cells. In separating out the short cells, we were selecting for cells that had recently divided, and in doing so, were isolating a population that was more synchronised with respect to age - that is, the sorted cells were at a similar stage in the bacterial growth cycle.'

Damien Baigl from the Microfluidics, Chemical Organisation and Nanotechnology group at the École Normale Supérieure, Paris, France, calls the work 'conceptually beautiful'. He says that 'the methodology is efficient, robust and easy to implement,' adding that it is 'performed with minimised physical and chemical stresses on the cells. It promises to become a powerful tool to investigate challenging biological problems, such as phenotypic variability, in E. coli and other bacterial systems,' he says.

Hulme points out that this is just the beginning - the next challenge will be to develop other tools. 'What is needed is an expansion of the toolbox for manipulating and examining bacteria in lab-on-a-chip devices,' she says.

Edward Morgan