A form of life that can stably store genetic information using a six-letter, three-base-pair alphabet?

For around 16 years, Floyd Romesberg’s group has been exploring un-natural alternatives (UBPs) to the Watson-Crick base pairs (C-G and A-T) that form part of the genetic code in DNA. Recently they have had remarkable success with one such base pair, called X and Y (for the press) and dNaMTP and d5SICSTP (in scholarly articles).[1],[2] This extends the genetic coding from the standard 20 amino acids to the possibility of up to 172 amino acids. Already, organisms engineered to contain X-Y pairs in their DNA have been shown to express entirely new (and un-natural) proteins.

There is also some measure of controversy. Why? Well, you might spot why with the structures of the bases as shown below.

I first note that d5SICS only has one exemplar in the Cambridge structural database (CSD), with the deoxyribose ring replaced by something quite different. The dNaM sub-structure is rather more abundant (360), although none have a deoxyribose ring attached. So we cannot really tell how these molecules might interact when adjacent (they are after all described as a base pair). But it is unlikely to be via hydrogen bonds, since d5SICS has only C-H groups, and dNaM has no acidic hydrogens either. Hence this base pair is described as being hydrophobic! I might suggest that some small molecule analogues of the two systems above are rapidly made and their crystal structures determined so that we might have at least some data about their interactions (or absence thereof).

If you were set the task of designing some un-natural base pairs to splice into DNA, I doubt you would start with the premise of dropping the complementary base pairing induced by two or three pairs of hydrogen bonds. Of course the integrity of the double helix is retained because of the C-G/A-T base pairs accompanying the hydrophobic d5SICS-dNaM ones. The controversy is about exactly how many such hydrophobic base pairs can in fact be included before the DNA structure becomes unstable to life. 

When I first came across attempts to engineer new forms of DNA (and possibly life), it was directed at replacing the pentose sugar by a hexose,[3] a project that ultimately failed because the resulting DNA was too flexible. Now we have the enthralling prospect of the discovery of many new alternatives to the standard base pairs, with biochemical consequences I cannot even begin to imagine! 


  1. A.W. Feldman, M.P. Ledbetter, Y. Zhang, and F.E. Romesberg, "Reply to Hettinger: Hydrophobic unnatural base pairs and the expansion of the genetic alphabet", Proceedings of the National Academy of Sciences, vol. 114, 2017. http://dx.doi.org/10.1073/pnas.1708259114
  2. D.A. Malyshev, K. Dhami, H.T. Quach, T. Lavergne, P. Ordoukhanian, A. Torkamani, and F.E. Romesberg, "Efficient and sequence-independent replication of DNA containing a third base pair establishes a functional six-letter genetic alphabet", Proceedings of the National Academy of Sciences, vol. 109, pp. 12005-12010, 2012. http://dx.doi.org/10.1073/pnas.1205176109
  3. M. Egli, P.S. Pallan, R. Pattanayek, C.J. Wilds, P. Lubini, G. Minasov, M. Dobler, C.J. Leumann, and A. Eschenmoser, "Crystal Structure of Homo-DNA and Nature's Choice of Pentose over Hexose in the Genetic System", Journal of the American Chemical Society, vol. 128, pp. 10847-10856, 2006. http://dx.doi.org/10.1021/ja062548x

Tags: , , , , , , , , , , , ,

3 Responses to “
A form of life that can stably store genetic information using a six-letter, three-base-pair alphabet?

  1. Can’t d5SICS and dNaM form chalcogen bond between O and S? This bond can be strong enough to keep the pair together. It must be tested though.

  2. Henry Rzepa says:

    As I noted above, there are very few pointers available from measured crystal structures. The closest such “chalcogen” interaction I can find there is an intramolecular 2.94Å, DOI: 10.5517/ccyg3p5 with the sum of the vdW radii being 3.32Å. The intermolecular contacts in the few other examples are ~3.2Å. So it does not seem these interactions are very strong!

  3. Qadir Timerghazin says:

    On the other hand, if they managed to make it work with non-hydrogen bonded “hydrophobic” base pair, that bodes well for expanding the genetic code with proper hydrogen-bonded non-natural base pairs.

    What is unclear to me (although I haven’t read the actual paper yet) how do they achieve *complimentarity* between d5SICS and dNaM? If there are no hydrogen bonds, dNaM-dNaM and d5SICS-d5SICS pairing should also be possible, unless there are some steric factors at play…

Leave a Reply