Bacterial cells used their expanded genetic code to make a modified version of a green fluorescent protein. Credit: Images by William B. Kiosses

Life has spent the past few billion years working with a narrow vocabulary. Now researchers have broken those rules, adding extra letters to biology's limited lexicon.

Chemist Floyd Romesberg of the Scripps Research Institute in La Jolla, California, and his colleagues manipulated Escherichia coli bacterial cells to incorporate two types of foreign chemical bases, or letters, into their DNA. The cells then used that information to insert unnatural amino acids into a fluorescent protein1.

Organisms naturally encode heritable information using just four bases: adenine (A), thymine (T), cytosine (C) and guanine (G). These form pairs that hold together DNA’s double helix, and different three-letter sequences code for each of the 20 amino acids that make up the proteins in living cells. The new work is the first to show that unnatural bases can be used to make proteins within a living cell.

The achievement, Romesberg says, shows that synthetic biology — a field focused on imbuing organisms with new traits — can accomplish its goals by reinventing the most basic facets of life. “There is no biological system so fundamental and more intimately related to what we are than information storage and retrieval,” he says. “What we’ve done is design a new part that functions right alongside the existing parts and can do everything they do.”

Alphabet extensions

Several teams are attempting to expand the genetic code. The four natural DNA bases can be arranged in 64 different three-letter combinations, called codons, that specify amino acids. But redundancy in this code — for instance, CGC, CGA, CGG and CGT all stand for the amino acid arginine — means that nearly all proteins needed for life are made of just 20 amino acids.

Researchers including geneticist George Church of Harvard Medical School in Boston, Massachusetts, are working on repurposing redundant codons to specify new amino acids. Romesberg’s group is exploring a different strategy: adding an entirely new base pair into DNA. That would vastly increase the number of possible codons, in theory giving cells the ability to exploit more than 100 extra amino acids.

Although Church still believes that his own approach is more practical for most applications, he describes the new work as a “milestone in exploring the fundamental building blocks of life”.

Researchers first imagined an expanded genetic alphabet in the early 1960s. The first big success came in 1989, when a team led by chemist Steven Benner, then at the Swiss Federal Institute of Technology in Zurich, forged DNA molecules containing modified forms of cytosine and guanine2. These “funny” DNA letters, as Benner has called them, could replicate and make RNA and proteins in test-tube reactions.

Funnier DNA

Over the past two decades, Romesberg’s team has made hundreds of even funnier DNA molecules. Unlike conventional base pairs in DNA and those made by Benner’s team — which are bound together by shared hydrogen atoms — these foreign bases stick together because of their insolubility in water, largely mimicking how grease droplets clump in water.

To function in living cells, though, the foreign base pairs need to sit alongside natural bases without disturbing the shape of DNA or disrupting essential tasks, such as the processes that faithfully copy DNA and transcribe it into messenger RNA — an intermediary molecule between DNA and proteins. In 2014, Romesberg’s lab reported a breakthrough: a strain of E. coli with a loop of DNA containing a single, unnatural base pair3. The ‘alien DNA’ was made of chemicals called dNaM and d5SICS (dubbed X and Y, respectively). But the cells divided sluggishly, and tended to lose their foreign DNA over time.

In a paper published earlier this year4, Romesberg’s team created a healthier, semi-synthetic E. coli that didn’t so readily reject its foreign DNA (in this version, d5SICS was replaced with a similarly shaped chemical called dTPT3). Yet this strain, as did the one reported in 2014, lacked the ability to use its new codons.

Credit: Graphic: Adapted from Synthorx Inc. 2015. Structure: Pedelacq, J. D. et al. Nat. Biotechnol. 24, 79–88 (2006)/RCSB PDB

In the latest research1, reported in Nature on 29 November, the team created healthy cells that can finally wield their foreign DNA. In separate experiments, the cells incorporated two unnatural amino acids (called PrK and pAzF) into a protein that emits a soft, green glow. Both the foreign bases and amino acids were fed to the cells, and any organism that somehow escaped the lab would not be able to produce them. To allow the cells to use these new components, the researchers created modified versions of molecules called tRNAs, which function to read codons and ferry the appropriate amino acids to the cells' protein factories — ribosomes.

The new amino acids did not change the shape or function of the green fluorescent protein. But “now that we can store and retrieve information”, says Romesberg, “let’s do something with it.” In unpublished work, his team has inserted a foreign base pair into a key site in the gene implicated in antibiotic resistance. Bacteria that shed their foreign DNA become sensitive to penicillin-related drugs.

Candy store

Romesberg has started a biotechnology company, called Synthorx, also in La Jolla, that is attempting to incorporate unnatural amino acids into protein-based drugs such as IL-2, a protein that regulates numbers of white blood cells. The approach could be used to design drugs that are taken up by cells more easily, for example, or that are less toxic or break down more quickly. Proteins could also be designed to have properties that conventional amino acids lack, such as the ability to strongly attract electrons. “It’s like being a kid in a candy store,” says Romesberg. But in this case, "the kid spent 20 years fantasizing about getting into that candy store. All of sudden I’m thinking what kind of candy can I get.”

Teams led by Benner and Ichiro Hirao, a biological chemist at the Institute of Bioengineering and Nanotechnology in Singapore, have already developed test-tube systems for using foreign DNA to encode unnatural amino acids. But Hirao sees advantages to moving into living cells. Proteins containing unnatural amino acids could be made at larger scale and more cheaply using bacterial cells, he says. Bringing the technology to eukaryotic cells would allow for the development of new antibody drugs, too.

However, Benner, who is now based at the Foundation for Applied Molecular Evolution near Gainesville, Florida, suggests that because Romesberg’s system relies on relatively weak hydrophobic forces to hold foreign base pairs together, its potential for industrial applications might be limited. Cells may tolerate the rare foreign base, Benner says, but “one simply cannot build an entire genetic system from them”.

Romesberg and his colleagues are now working on expanding their genetic alphabet further. So far, the team has identified 12 more codons containing X and Y that are functional, says Romesberg, but “there’s a lot yet to do”.

figure c

Images by William B. Kiosses

Bacterial cells used their expanded genetic code to make a modified version of a green fluorescent protein.