Synthetic XNA molecules can evolve and store genetic information, just like DNA

By Ed Yong | April 19, 2012 2:00 pm

Out of all the possible molecules in the world, just two form the basis of life’s grand variety: DNA and RNA. They alone can store and pass on genetic information. Within their repetitive twists, these polymers encode the stuff of every whale, ant, flower, tree and bacterium.

But even though DNA and RNA play these roles exclusively, they’re not the only molecules that can. Vitor Pinheiro from the MRC Laboratory of Molecular Biology has developed six alternative polymers called XNAs that can also store genetic information and evolve through natural selection. None of them are found in nature. They are part of a dawning era of “synthetic genetics”, which expands the chemistry of life in new uncharted directions.


DNA looks like a twisting ladder. Its sides are chains of a sugar called deoxyribose (the D in DNA). Each sugar is attached to one of four ‘bases’ – these form the rungs of the ladder, and are signified by the letters A, C, G and T. RNA is similar, with three important exceptions. It’s typically only half a ladder – a single helix, rather than DNA’s famous double. Its ‘T’ rung is a ‘U’. And its sugar is ribose rather than deoxyribose (hence, the R in RNA).

Both of these molecules are called nucleic acids. So are Pinheiro’s XNAs, but they make their ladders using different sugars. If arabinose stands in for deoxyribose, you get ANA instead of DNA. If cyclohexane plays the part, you get CeNA. If the role goes to threose, you get TNA, and so on. These differences aside, all the XNAs use the same bases. Any of them could pair up with a complementary strand of DNA or RNA.

“They are very interesting with respect to the origin of life,” says Jack Szostak, a Harvard biologist who studies life’s beginnings and was not involved in the study. “In principle, many different polymers could serve the roles of RNA and DNA in living organisms.  Why then does modern biology use only RNA and DNA?”

Most biologists now think that RNA preceded DNA as life’s chief information molecule. Phil Holliger, who led the new study, says that the “inescapable conclusion” is that its dominance was the result of a “frozen accident at the origin of life”. RNA may have gained supremacy because of random factors rather than some inherent quality, just as VHS and Blu-Rays eventually won out over Betmax and HD-DVDs.

The alternative is that some nucleic acids may be better at copying themselves, or speeding up other chemical reactions. “Phil’s work will certainly make it possible to compare the functional abilities of a wide range of synthetic nucleic acids,” says Szostak.


Pinheiro created his XNAs by tweaking a natural enzyme called DNA polymerase, which copies DNA. It ‘reads’ a piece of DNA, grabs nearby bases, and assembles a matching strand. If you set the polymerase loose upon its own gene, you can get the enzyme to make more copies of itself.

Here’s the clever bit. DNA polymerase is normally very fussy about the bases it grabs. It only selects ones with a deoxyribose sugar so that it assembles DNA, rather than any other nucleic acid. But Pinheiro evolved the enzyme so that it prefers to use the building blocks of his XNAs instead.

He started with a varied pool of polymerases, all slightly different, and all mixed with their own corresponding gene. He then supplied them with the XNA building blocks. Within these pools of varied enzymes, some were better at building nucleic acids with the weird sugar backbones. By selecting for these unusually efficient polymerases, Pinheiro quickly evolved enzymes that could assemble XNA strands from DNA ones.

He also created an enzyme that could do the reverse – convert XNA into DNA. Of course, no natural enzyme can even begin to do this, so the evolution trick didn’t work in this direction. Instead, Pinheiro used a more brute-force approach: he took a different polymerase, randomly mutated it, and looked for versions that could do the XNA-to-DNA conversion. Eventually, he moulded one.

Pinheiro ended up with enzymes that could copy information between XNA and DNA, with an accuracy of 95 per cent or more. With more work, it should be possible to cut DNA out of the loop altogether, so that XNAs can be directly built from XNAs. If this is possible, Szostak adds, “In the longer run, it may be possible to design and build new forms of life that are based on one or more of these non-natural genetic polymers.”

There are already hints of this. The team has so far managed to copy FANA and FANA, CeNA from CeNA, and even HNA from CeNA. However, all these steps were far less efficient than working through DNA. But Holliger says that there would be few benefits to abandoning the middle-man, because “it’s convenient to go through DNA.” That’s because all of our genetic technology is geared to our standard nucleic acids. If we moved towards XNA-only experiments, we would also have to tweak our sequencing tools and cloning techniques to match.


The XNAs have important differences to their more familiar cousins. “We haven’t gone very far from standard DNA chemistry, but the XNAs already have strikingly different properties,” says Holliger.

For a start, they are extremely tough. In the natural world, DNA and RNA are beset by many dangers. Acids with break some of their rungs, and many enzymes will easily cut their backbones. But XNAs have no such problem. Their unnatural nature makes them invulnerable to enzymes, extreme pH values, and other harsh conditions. “They’re really tough as nails. We’ve thrown the whole New England Biolabs catalogue at them,” says Holliger, referring to a massive directory of chemical reagents.

These properties mean that the XNAs are well-suited for certain applications. For decades, scientists have created short strands of DNA or RNA called aptamers, which are designed to stick to specific targets. They could act as sensors that reveal the presence of a specific molecule, or deliver drugs to diseased cells by latching onto telltale proteins. Their uses are legion, but they are fragile tools.

If aptamers were built from XNAs (and more on this later), they would be tougher. However, they would still retain their key feature: they could evolve to recognise different targets. Just like RNA, many of the XNAs fold up into complicated three-dimensional structures. Alex Taylor used this property to create HNA aptamers (H is for anhydrohexitol) that recognise a protein and an RNA shape, by repeatedly selecting for the ones that form the closest fit.


This new study is just one of many attempts to expand the palette of molecules that carry genetic information. Every part of the ladder is up for alterations, from the bases to the sugars. For example, Steve Benner from the Foundation for Applied Molecular Evolution, has created a polymer that adds two new bases – Z and P – to the existing quartet of A, G, C and T. “It allows higher information density,” he says.

“This is just the beginning,” says Holliger. “We’re going to try and get to ever more divergent chemistry.”

Of course, any talk of unnatural chemicals, especially hard-to-destroy ones, is bound to spark a discussion about risks. Indeed, in a related editorial, Gerald Joyce from the Scripps Institute says that while biologists are starting to “frolic on the worlds of alternative genetics”, they “must not tread into areas that have the potential to harm our biology.”

George Church from Harvard University says that before assessing the technology’s benefits and risks, we’ll have to see examples of what it can achieve that “cannot be achieved using DNA or RNA”. He says, “The risk could go up by increasing the survivability of XNA relative to [DNA or RNA]. That might lead to a full replacement of natural nucleic acids with these artificial ones.

But Benner that the XNAs would actually “mitigate the risks associated with biotechnology”. Many of the worries about genetic modification revolve around altered genes going feral, and running amok through wild populations. But XNAs shouldn’t be able to do that. In a world ruled by DNA and RNA, XNAs would be invisible. They would sit behind a genetic firewall, unable to exchange genetic information with living things. Such molecules, far from being an unnatural danger, might actually be the ultimate biosafety tool.

Reference: Pinheiro, Taylor, Cozens, Abramov, Renders, Zhang, Chaput, Wengel, Peak-Chew, McLaughlin, Herdewijn & Holliger. 2012. Synthetic Genetic Polymers Capable of Heredity and Evolution. Science

More on synthetic biology:


Comments (15)

  1. Hello,
    Great article. My mind is racing. excited to see the aptamer discussion. I’d like to read the primary article but the link to Pinheiro et al appears to be dead.

  2. lovely reading your articles.

  3. Steve

    Steven Benner is not at the University of California in Santa Cruz. He is Director of the Foundation for Applied Molecular Evolution and the Westheimer Institute for Science and Technology in Gainesville FL.

    [Oops! Sorry. Corrected – Ed]

  4. Josiah

    Would that be, “Xtreme Nucleic Acid?”

  5. ST

    This is very exciting and tremendous work.
    But I gotta say stuff like this “Synthesis and reverse transcription establish heredity (defined as the ability to encode and pass on genetic information) in all six XNAs.” is really annoying. When it comes to genetic material we have a pretty clear definition what it means to be heritable and the authors redefined it to make it fit their work. If synthesis and reverse transcription means heritable storage of genetic information, then a simple paper print out of a genetic sequence has the same heredity properties of XNA.

  6. By sheer coincidence I’d just finished re-reading (after 39 years [sic]) Watson’s “DOUBLE HELIX” today. Leaving aside the blatant sexism towards Franklin, I found myself just as riveted as when I first read this book all those years ago – even though I knew the ending this time 😉

    I’d hardly put the book down when Adam Rutherford’s tweet came through drawing my attention to this article. Beautifully written and fascinating stuff; and very much the icing on the cake of my reading earlier in the day.

    Even from the point of view of pure science, the exercise of constructing and manipulating synthetic nucleic acids is awe inspiring. Consideration of the possible applications of this technology makes it all the more exciting.

    Can’t remember the last time I enjoyed reading a science article as much as enjoyed this one.

  7. megan

    Wow, it makes the concept of Synthetic lifeforms so real and more probable. Like the replicants in Blade Runner. Sent to mine uninhabitable planets and extremes of space due to their chemistry biology. But then programmed to cell die early for safety and prevention of over running DNA/RNA archaic originals.

  8. Can’t wait to tell some intelligent design fans that Vitor Pinheiro did a better job then their proposed designer :)

  9. Was it just a fluke that life on earth used RNA/DNA? Would these XNA alternates be just as likely bases for life on other planets? Am I asking you to speculate with far too little data?

  10. floodmouse

    If the XNA variants are really bulletproof against most forms of chemical damage, wouldn’t that prevent them from going through fast evolution like DNA-based life? It seems like there wouldn’t be so many mutations to select from.

  11. @Mike – Thanks! Appreciated.

    @Roedy – That’s the idea that Holliger is suggesting, but the point is that we don’t know. But we can now test RNA/DNA against these others to see if they have any advantages.

    @Floodmouse – Mutation isn’t just produced by chemical damage. There’s also basic copying error. That contributes a lot too.

  12. Odin

    Hope they come up with a triple helix next.

  13. T.g.Padmanabhan

    Victor Pinheiro will be worshiped as the God by new artificial life forms based on XNAs !!!
    chances for giving the artificial life forms immortality by tweaking the XNAs are bright!!
    Long live the artificial God- Victor Pinheiro!!
    Thanks to Discover Magazine!

  14. sourav

    it”s a miracle in science with a ray of hope for creating artificial life.

  15. Anthony Hauser

    Um…the “CeNA” is claimed by the article to be Cyclohexane, but alkanes (hence the ANE at the end) don’t have pi-bonds. It would instead be a cyclohexene with a pi-bond. Although then the chair-conformation structure present wouldn’t be that good because it’s unstable in that presentation.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.

See More

Collapse bottom bar