On the path toward sophisticated cellular computation, synthetic biologists are constantly seeking better ways to program logic into cells. One way to do this is using DNA segments known as inversion recombination elements. In essence, these inversion elements act like binary switches that can write ones and zeroes directly into DNA. In the July 30, 2008, issue of PLOS, SynBERC researchers Timothy Ham, Sung Kuk Lee, Jay Keasling and Adam Arkin demonstrate how engineers can combine two or more such elements together to design complex logical systems in DNA.

“By writing binary states directly to the DNA, we avoid the limitations posed by other kinds of biological memory and logic,” says Timothy Ham. The hard-coded DNA switch does not require protein expression to “remember” its state, even upon cell death. “By carefully positioning the recombinase recognition sites in the sequence, we can increase the number of possible binary states logarithmically with each new switch.” So while one switch has just two possible states, two switches can have six possible states, and three switches can have a total of up to 15 states.

This kind of biological circuitry is qualitatively different from past innovations: It goes beyond simple Boolean logic (if this, then this) to more a more sophisticated sequential logic required for complex computation. (The underlying chemical networks are even theoretically capable of Turing-machine like computations, although no serious scientist is holding their breath for a tete-a-tete with E. coli.) In the mid-term, synthetic biologists do expect to scale-up this logic in cells to construct systems that require a cell’s ability to recognize and remember what it has experienced. The most prominent examples include cells that incorporate safety controls that limit their deployment, such as engineered microorganisms to treat disease, clean up water and soils, or even serve as emergency blood substitutes. Such applications require a level of sensing, actuating and regulatory logic that less sophisticated chemical production pathways can’t support. Ham says, “Scaling to larger applications, with more states and deeper sequential logics, is certainly central to synthetic biology.”

The double inversion switch has already served as the basis of one SynBERC -associated project that was done in Chris Voigt’s lab by Eli Groban, Liz Clarke, and Ryan Clark. The work involved the well-studied bacterial chemotaxis system, which allows cells to move towards nutrients and away from toxins. Bacteria contain multiple external sensors that sense an array of different small molecules. These sensors all feed into the same central pathway that controls cellular movement. Since many different signals all converge into one signaling pathway, the cell integrates signals from multiple receptors without having the ability to distinguish between different input signals, other than deciphering a difference between food and toxins.

The three researchers programmed E. coli to distinguish between two stimuli and move towards one or the other depending on the absence or presence of an outside signaling molecule. This required the engineering of two orthogonal signaling systems and the replacement of the native chemotaxis signaling machinery with the DNA for the novel system. Finally, they inserted an externally controlled DNA switch, and the double inversion switch of Ham et al was just right. The switch was programmed to recognize a small molecule signal and direct cell movement towards one stimulus or another. This DNA encoded switch also contained a memory function, allowing the progeny of each bacterium to remember which direction to move. Groban states, “This project represents the embodiment of the SynBERC vision of synthetic biology: a series of parts were characterized and then combined in a living organism to confer to a bacterium a novel behavior – a behavior that is entirely under human remote control.”