Scientists Write their Own Genetic Code

Genome sequencing is all the rage, as technologies have enabled scientists to crack DNA codes faster and faster. But for those who would like to apply what has been learned from all that “reading” of genomes to actually writing their own, there are technological hurdles to overcome. Jingdong Tian, a biomedical engineering faculty member in the Pratt School and the Duke Institute for Genome Sciences & Policy (IGSP), is developing an automated gene synthesizer aimed to do just that.

“We are trying to build an automated microfluidic machine so everyone can do DNA synthesis right on their lab bench,” Tian told GenomeLIFE two years ago. “You would simply type in a sequence and the instrument would synthesize it.”

As described in a recent issue of Nature Biotechnology, Tian is well on the way to reaching his goal. His team has designed a 1-by-3 inch chip that can produce custom-made segments of DNA in two days that would typically require many large pieces of equipment, significant human labor and two weeks to produce.

The new technology also really cuts costs. As an example of how time-consuming and expensive current technology is, Tian cited the recent cloning of the entire genome of a single bacterium that took more than four years to complete, with a price tag of more than $40 million.

“Using current technology, it takes about 50 cents to a dollar to create each base pair of DNA; using the new chip reduces costs to less than half of one cent per base pair,” Tian says.

Step by step

Gene synthesis involves the synthesis, purification and assembly of “oligos,” short snippets of DNA, usually less than 50 base pairs each. The new chip performs all three of these activities at once to produce stretches up to 1,000 base pairs long (an average gene is about 3,000 base pairs long).

The chip itself has row upon row of tiny indentations, or wells. The biochemical equivalent of an inkjet printer shoots the desired DNA bases into each well where they assemble. Enzymes then release the resulting DNA strands from the walls of those wells.

“The chip basically combines the three steps into one, which can be completed in less than two days, and without all the labor currently needed,” Tian said. “Also, since the wells are so small, significantly smaller amounts of expensive chemicals are needed to run the reactions.”

The final step is the equivalent of a DNA spell check; errors in the code, usually missing base pairs or single letter typos, are automatically detected and fixed.

Reality check

Because researchers can produce so many oligos so quickly using the new technology, they can screen many versions with subtle differences to see which one works best for their desired application, Tian said.

There is already interest from various Duke labs, including that of his IGSP and Pratt colleague Lingchong You, to put the new technology to use. You’s lab designs synthetic gene circuits that program bacteria to carry out new and useful functions, much like software on a computer.

“Low-cost DNA synthesis can make us think about alternative strategies for engineering gene circuits or optimizing signaling pathways,” You says. “If I were smart enough, every circuit I implement should work as designed. But that’s far from the truth. Many circuits we build don’t work well or don’t work at all. High-throughput DNA synthesis offers an opportunity to synthesize a large pool of variants from a common design, to speed up the process of circuit optimizing.”

Tian says the group that set a record when it synthesized that bacterial genome, led by Craig Venter, has also been in touch. Still, there is plenty of work left to do. Their on-chip technology is sufficient for routine gene synthesis, but Tian would ultimately like to build a machine that can automate whole-genome synthesis from beginning to end. “We are about halfway to where we ultimately want to go,” he says.

Citation: “Parallel on-chip gene synthesis and application to optimization of protein expression”, Jiayuan Quan, Ishtiaq Saaem, Nicholas Tang, Siying Ma, Nicolas Negre, Hui Gong, Kevin P White and Jingdong Tian. Nature Biotechology 29 (2011).