Synthetic biologists lack DNA 'parts' that are known to work reliably in different cells or environments. Credit: SAM OGDEN / SCIENCE PHOTO LIBRARY

EMERYVILLE, CALIFORNIA

Six months since it launched, the world's first factory for making professional-quality biological DNA 'parts' is beginning to stock its shelves.

More than 60 people — academic researchers, industry partners and interested members of the community — joined the staff at the International Open Facility Advancing Biotechnology (BIOFAB), at a meeting on 19 and 20 July to discuss the facility's progress so far and its aims over the next few years.

BIOFAB aims to supply synthetic biologists with a collection of genetic parts that they can use in their experiments. Biological parts — actually sequences of DNA — should have known and predictable functions, so they can be inserted into cells to boost the production of a particular protein, for example, or make it sensitive to a specific toxin. BIOFAB has begun to add some early attempts at parts sequences to its registry. But there is still some way to go before the fruits of BIOFAB's labours can be useful to researchers, says Drew Endy, the facility's director and a synthetic biologist at Stanford University in Palo Alto.

Synthetic biologists aim to redesign bacterial cells for useful ends, such as producing biofuels or creating new types of medicines. The idea is that DNA can be broken down into components that can be swapped in and out of cells like parts of a clock radio. But in reality, the same sequences often behave differently in different cells.

Set up in 2003 at the Massachusetts Institute of Technology in Cambridge, the Registry of Standard Biological Parts was the first collection of genetic components to be made available. Its plays host to parts mostly made by students who have taken part in the International Genetically Engineered Machine competition, an annual synthetic biology contest that has been running since 2003. But only some 2,300 out of the registry's current collection of 13,400 parts do what they are supposed to do.

Mix and match

BIOFAB aims to up the game by providing free biological parts that researchers can reliably use in their work, say Endy and codirector Adam Arkin, a bioengineer at the University of California, Berkeley. Housed at Lawrence Berkeley National Laboratory's Joint BioEnergy Institute in Emeryville, BIOFAB is so far a small-scale operation. Four or five staff scientists share a bench in a vast fourth-floor laboratory, where they design and test DNA constructs. Endy, Arkin and a few others bring the total number of staff to 10.

We have to start by quantifying the limitations of what can be done. ,

At the meeting, BIOFAB researchers presented their first efforts to lay the groundwork for designing and characterizing biological parts. The pilot project tested a dozen or so of the most commonly used gene promoters (regions of DNA that facilitate gene transcription) and segments of DNA that encode ribosome-binding sites (sequences of messenger RNA that control protein translation) to determine whether they behave consistently in different cellular contexts.

"Do we know enough to build strong, medium and weak synthetic promoters?" asked Vivek Mutalik, a BIOFAB scientist. Not really, he told the audience. To begin to answer that question, the researchers modelled different combinations of promoters and ribosome-binding sites, finding that they could predict about 70% of the variance in the combinations.

"We have to start by quantifying the limitations of what can be done," says Endy.

The next step is to use their analysis to design and build promoters and ribosome-binding sites with improved performance. "If we work our butts off, by the end of this year we should have closed that loop," says Endy, adding that this would be something researchers could use. "They would be better by a measurable amount, it would be quantified, and it's free."

Piece by piece

Synthetic biologists have struggled to standardize comparisons of how different parts are working. BIOFAB uses a technique devised by one of Endy's former students, Jason Kelly, in which researchers determine the relative activity of their promoter based on a widely used reference promoter1. It is not a perfect system, says Endy, but it is a start. It is unclear, however, whether these reference tools will work under industrial conditions, says Endy. Another continuing BIOFAB project is to create a bookshelf of reference tools tested under conditions used in commercial metabolic engineering.

BIOFAB isn't alone in working on such problems on a community-wide scale. Richard Kitney, a systems biologist from Imperial College London, is launching a similar parts-production facility in the United Kingdom that will work closely with BIOFAB. "We want to be able to say if we can replicate parameters across the Atlantic," says Kitney.

BIOFAB is also receiving a push from a recently announced synthetic competition, spearheaded by Richard Murray, a biological engineer at the California Institute of Technology in Pasadena. Murray and others are still refining the details, but the mandate will be to solve concrete problems — such as characterization and measurement. In addition to energizing the community, many of those solutions could provide protocols for BIOFAB.

Ultimately, the goals of BIOFAB — and of synthetic biology — must overcome some basic limitations in the field. "Is there a part today that we could say works?" asked a researcher at the meeting. "I don't think there is a single biological part today that would work in any environment you give it."