Skip to main content
Premium Trial:

Request an Annual Quote

Scripps Team Creates High-Fidelity Unnatural Base Pair with Eye Toward Biologics


Work published recently by a Scripps Research Institute team has advanced the field of synthetic nucleic acids, demonstrating an unnatural base pair palatable to DNA replication machinery under standard PCR conditions, the first of its kind.

Previous work by Ichiro Hirao's group at Japan's Riken Institute created a pair of bases that showed high fidelity and efficiency, but using the pair in PCR required a specific commercial polymerase with exonuclease activities. In 2011, Hirao told PCR Insider this work would be used to develop aptamers and would be commercialized by spinout company TagCyx Biotechnologies. The group has since received a patent on the technology.

The new pair, described in a paper published in October in The Journal of The American Chemical Society was generated in the lab of Scripps researcher Floyd Romesberg. It is an improvement over previous work in that it is both efficiently replicated without sequence bias and efficiently transcribed into RNA, according to the researchers. And, although this is now the third unnatural base pair reported to replicate with PCR, both of the others require longer extension times and need polymerases with exonuclease activity.

According to Romesberg, their new pair, called dTPT3-dNaM, can replicate with Taq polymerase at normal extension times, making this "the first time that any unnatural base pair by any metric overlaps with a natural base pair," said Romesberg. He added, "When you amplify [our pair] with Taq you get a fidelity that, within blurry error bars, overlaps with a natural base pair with the lowest-fidelity polymerases. But…those are still natural polymerases that get the job done in nature. Arguably, they're the worst ones of that, but they are good enough for organisms to use," he said.

Complementary shape and hydrogen bonding are key ingredients for the two natural nucleotide pairs, A-T(U) and G-C. While other unnatural base pair builders have used the H-bond motif as a starting point, the heart of Romesberg's work is the hypothesis that hydrophobic and packing forces could be sufficient to mediate unnatural base pair replication.

Romesberg remarked that, at this time, "two of the three pairs that have been developed and that are advanced [to PCR] do not use hydrogen bonding at all, but use a completely different strategy than nature used, [which] is a real accomplishment of chemical or synthetic biology," he said.

Synthetic base pairs must simulate natural ones, and not kink up DNA structure or make DNA polymerase skip a beat. They must also not get in the way of the replication of natural bases, Romesberg said, the way "A-T doesn't get in the way of replicating G-C, and vice versa." This would allow researchers to "pepper in" unnatural base pairs into a genome, and store and retrieve them at will.

Romesberg asserted that having a third pair that replicates in the ranges of natural ones could "significantly increase the information density in the same length of DNA, because you have three bits instead of two." This could enhance using DNA for information storage, he said.

He also suggested that synthetic base pairs could expand coding for unnatural amino acids, giving access to properties not available using the 20 natural ones. "If you had a new base pair you could have an infinite number of new codons [and] in principle, you could synthesize proteins with as many different unnatural amino acids as you like," Romesberg said.

At the moment, Romesberg said his group is "focused on getting our base pairs into organisms." They hope to soon create a semi-synthetic single-celled organism with an expanded genetic alphabet.

Romesberg believes this might be particularly useful for developing biologic drugs. "One real upside of proteins is that you can evolve them and tailor them using the natural machinery within a cell," Romesberg said. For example, directed evolution could be done using phage display of an unnatural base-pair-encoded- and unnatural amino acid-containing protein. This would allow researchers to "evolve proteins in a matter of weeks by simulating natural Darwinian evolution," Romesberg said, sorting through a huge number of variants — on the order of 109 variants — for the best ones, instead of having to sort through one at a time. "Really what I'm interested in trying to do is combine the advantages of traditional small molecule therapeutics with biological therapeutics," Romesberg said.

When it comes to biologic drugs, this could be a big advance over current development techniques, and the market for biologics has shown no signs of flagging. A recent Nature Biotechnology report scrutinized growth in this sector, showing an annual increase of more than 18 percent in 2012, seven-fold higher than the sales in pharma overall.

In terms of commercialization, Romesberg stated that he has narrowed down his list of potential partners and is now evaluating possibilities, but declined to comment further at this time. He added, "what the world is looking for is some new hook in the protein therapeutics world… everyone thinks that this is the next big thing, and four out of the next $5 billion drugs are going to be proteins; everyone is looking for a new way to do something with them. So, if I were involved in commercialization, it would be along those lines, and I would not consider it until we had made significant progress toward doing it." Romesberg said that his lab currently has a new study on dTPT3-dNaM far into the review process, but could not comment further due to embargo issues.

Scripps, where this research was done, is somewhat of a hotbed of synthetic biology. Scripps' Gerald Joyce is developing a method using artificial RNA to quantify proteins. As Joyce recently told PCR Insider, these autocatalytic aptazymes are built out of non-biological, L-RNA molecules with the same self-replicating properties of their biological enantiomeric twins but with the bonus of being completely resistant to breakdown by ribonucleases. The institute's Peter Shultz is also tackling synthesis of unnatural nucleic acids, and recently won the Solvay prize for his efforts, including work to genomically re-code E. coli so that it can incorporate unnatural amino acids into proteins. Romesberg confirmed this local environment enhances his work, and added, "I'm biased, but I think Scripps is the most exciting concentration of chemical or synthetic biology in the world."