Building microbes that could act as factories for making fuels or pharmaceuticals requires incorporating novel DNA sequences into the cells. But assembling all of the genetic pieces needed to engineer these new microbial strains can be slow.
Building a novel organism can take a lot of trial and error, says Steve C. C. Shih, a postdoctoral fellow in the laboratory of Anup K. Singh at Sandia National Laboratories, in Livermore, Calif., and the Joint BioEnergy Institute. First, scientists must identify a gene that could give an organism a desired function. They must then put the gene into a ring of DNA called a plasmid that carries it into the cell and facilitates gene expression. Finally, they test the new microbe for the desired traits. This process often doesn’t work the first time, so researchers have to go through this cycle many times, Shih says. One bottleneck is building plasmids, which requires substantial time in the lab.
To speed the production of synthetic organisms, Shih and his colleagues developed a droplet-based microfluidic device that automates plasmid construction and delivery to cells. The device contains interconnecting pathways made from electrode “bricks.” To move a droplet of liquid along these pathways, Shih controls the electrodes, flipping the switch on one brick to create an electrostatic force that pulls a droplet along the pathway toward the next. Brick by brick, the researchers can move a droplet, or series of droplets, through the device’s electrode pathways, assembling plasmids by drawing from different pools of DNA pieces, mixing them, incubating them, and then transfecting them into cells.
In this proof-of-principle experiment, Shih’s laboratory used the device to build 16 different plasmids—all possible pairs of four different genes and four promoters, controllable on-off switches for gene expression. The team loaded the device’s reservoirs with solutions of each gene, each promoter, the plasmid backbone, and the necessary reagents.
To make the 16 different plasmids, the researchers drew droplets from the appropriate reservoirs and sent them into a mixing area on the device. They pulled the resulting droplets into a capillary, where all 16 droplets were kept separate but incubated together at the right temperature to complete the plasmid assembly. After incubation, the droplets moved, one by one, into a part of the device that contained either bacteria or yeast cells. An electrical charge then made the cell walls permeable to the plasmids, allowing them to enter the cells and create new organisms ready for testing. For both bacteria and yeast, subsequent gene-sequencing tests showed that 95% or more of the 16 plasmids’ sequences were correct, Shih says, a promising result.
Douglas Densmore of Boston University says this device is an “important step” for speeding the development of genetic building block libraries for synthetic biology. He’d like to see a device that can make 100 or even 1,000 plasmids in a single run. Shih is working to increase output but is limited by how many electrodes will reasonably fit on the two-dimensional device. He’s developing a 3-D microfluidic device, which would make it “very easy to increase the electrode density,” he says.