to BioTechniques free email alert service to receive content updates.
Hot-wiring cells
Synthetic biologists are designing genetic circuits of increasing complexity. But how did the field get to this point, and where is it going? Nathan Blow examines the challenges, and potential applications, of engineering gene circuits
Nathan Blow, Ph.D.
Full Text (PDF)

But even in the face of these challenges, researchers have developed unique, clever circuit designs with components capable of logical computations, filtering, oscillation, noise propagation and memory from a variety of inputs.

Massive input

Yaakov (Kobi) Benenson, an assistant professor in the department of Biosystems Science and Engineering at the Swiss Federal Institute of Technology (ETH Zurich), initially became interested in the possible applications of gene circuits during his doctoral work in the lab of Ehud Shapiro at the Weizmann Institute of Science.

At the time, Benenson was focused on creating programmable biochemical-based computing systems.

“Even then, I always had medical applications in mind,” says Benenson. “Exploring the disease states of single cells, interrogating information from inside cells — this interested me.”

His work in the Shapiro lab had focused on in vitro biochemical computing systems. But to extend these studies to medical applications, Benenson realized that he needed to go inside the cell with his programmable computing systems — leading him directly into the world of synthetic biology and engineered gene circuits.

While gene circuits provide the opportunity to monitor the state of a cell (input coming from the cell, output that can be directly read like in the case of the first oscillating gene network), these engineered circuits also provide the opportunity for other interesting applications. For example, gene circuits could be engineered to activate expression of therapeutic transgene in response to a specific input. Similarly, bacteria can be engineered with specific metabolic gene circuits to respond to cues and produce novel outputs. Benenson saw this as the next step for his DNA computing research started in the Shapiro lab.

After graduating from Shapiro's lab, Benenson moved to Harvard University as a Bauer Fellow at the FAS Center for Systems Biology. It was here that Benenson, teaming up with Ron Weiss who at the time was at Princeton University, made his mark in circuit engineering with the demonstration that RNAi could act as a component in gene circuits in human kidney cells.

“The hallmark of decision making in cells is the requirement for a lot of inputs,” explains Benenson. “And how do we build large circuits to process multiple inputs? RNAi presents a scalable mechanism to accomplish this.”

To take advantage of RNAi as a circuit component, Benenson and Weiss first had to construct two or more mRNAs encoding the same proteins but with different non-coding regions, implementing an OR logic between the levels of these mRNAs. Next, sets of siRNA targets were added to the 3′ ends of the mRNAs. Endogenous inputs and their effects on these siRNAs were evaluated to confirm robust output from the mRNAs. From this starting point, the basic system could evaluate signals in a logical manner: in the cases where inputs block all siRNAs targeting the same mRNA, an mRNA will be transcribed, creating an AND logic. In addition, if the input activates an siRNA, the mRNA will be targeted by this siRNA and the output not transcribed (and vice versa), creating a NOT logic. The combination of OR, AND, and NOT gates can effectively support any logic computation.

“Another nice thing about RNAi-based circuits is the fact that it is easy to develop sensors for every miRNA in a cell,” notes Weiss who is quick to add that with nearly 1400 mammalian miRNAs currently identified, a lot of information can be gleaned from this approach. Weiss and Benenson demonstrated this nicely in a 2011 follow-up article published in Science where the pair described multi-input RNAi-based logic circuits that enable the cells themselves to describe who, or what, they are. In this circuit design, cell classification is accomplished when the circuit senses the expression levels of a set of endogenous miRNAs and then triggers a cellular response if those expression levels hit a certain threshold.

“The nice part is that this approach can be easily generalized and adapted to profiling different miRNAs and therefore different cell types,” explains Weiss, thus enabling the monitoring and identification of a potentially wide range of cells. In fact, Weiss says that with minor tweaks to the circuit, they are now able to even more robustly classify a growing range of cells using different endogenous miRNAs.

  1    2    3    4    5