Plug and play genes

Researchers have designed a way to streamline the construction of synthetic gene networks, a paper published online this week in Nature Biotechnology reports. The technique could speed up the process of building such networks, the authors say. The study was "very solidly executed," said J. Christopher Anderson, a bioengineer at the University of California, Berkeley. Image: Liz_Henry/flickr Synthetic biology, the process of combining genes to create artificial networks, holds promise for numer

Written byTia Ghose

| 3 min read

Researchers have designed a way to streamline the construction of synthetic gene networks, a paper published online this week in Nature Biotechnology reports. The technique could speed up the process of building such networks, the authors say. The study was "very solidly executed," said J. Christopher Anderson, a bioengineer at the University of California, Berkeley.

Image: Liz_Henry/flickr

Synthetic biology, the process of combining genes to create artificial networks, holds promise for numerous applications, including sensing toxic chemicals, creating biofuels and generating new drugs. But creating artificial networks can be time-consuming and frustrating: after assembling their artificial network, researchers can spend months painstakingly--and blindly--swapping out different promoter regions, RNA regulators, or chemical inputs, just to get the circuit to do what they want. To see if they could speed up the construction process, James J. Collins, a bioengineer at Boston University, and his colleagues assembled a library of 20 versions of two gene promoters--which are repressed by the tetracycline repressor (TeTR) and lac1 repressor--each producing different levels of gene expression. To generate the library, they used the sequence of either TeTR or Lac1 promoters, but tweaked each one's effectiveness by nestling it between up to a million different runs of randomly chosen base pairs--which altered how polymerases and other chemical regulators bind to and interact with the promoter regions. Then, they tracked how much each of the 20 versions of the promoter altered downstream gene activity by measuring the expression of a fluorescent marker. The 20 different sequences they chose covered the range from minimum to maximum gene expression. Using a computer model with parameters set by the activity levels they calculated from the library, the group designed and built a simple gene circuit with an output gene regulated by both TetR and Lac1 promoters. The simulation was able to predict how fluorescent protein expression varied with levels of promoter-inhibiting chemicals. Next, the group used the simulation to model the behavior of a more complicated circuit called a genetic timer. The timer network hinges on two "mutually inhibitory" genes--each one wants to be on, but also wants to turn the other one off, Collins told The Scientist. Chemical inducers can then "flip" the circuit from expressing one of the genes or the other at a precise time, similar to a toggle switch. On its own, the computer simulation was not able to accurately predict the behavior of the genetic timer circuit, Collins said. So the team built one timer network using a promoter from each of their libraries, and tracked its behavior over time. Using the information gleaned from that one network, the group was able to calibrate their model and make accurate predictions for all the other possible network combinations. Finally, the team used the computer model to construct a genetic timer circuit in yeast. They could accurately time when the yeast would clump together, a process important for brewing. The team is hoping to broaden the approach to generate libraries of regulatory RNA or proteins, Collins said. Using "combinatorial libraries to engineer genetic circuits" is a fairly novel idea, but it's "where the field is going," Anderson said. However, he cautioned, while the strategy the team used may work well for switches, where you're "only measuring the end state of the system," it may not apply so easily to more complex networks. For genetic switches, measuring fluorescent output can be fast, even when screening through the output for "10 to the sixth" different sequences, he said. However, creating well-characterized libraries for systems that cycle through different states may be more challenging, because "you'd have to follow each of those million things over time to see whether they were cycling."
**__Related stories:__***linkurl:Switched on science;http://www.the-scientist.com/article/display/54580/
[1st May 2008]*linkurl:Synthetic memory in eukaryotes?;http://www.the-scientist.com/news/display/53598/
[19th September 2007]*linkurl:New molecular switch for genes;http://www.the-scientist.com/news/display/53431/
[26th July 2007]