Synthetic biology

A yeast for all reasons

Scientists have begun to overhaul a yeast's genome to make it more stable, engineerable and evolvable. Remarkably, the part-natural, part-synthetic yeast cells function and reproduce without obvious ill effects. See Letter p.471

The baker's yeast Saccharomyces cerevisiae is a model organism, and therefore one of the best-understood biological systems on the planet. Nevertheless, the Byzantine complexity of its inner workings still keeps bioengineers up at night, and continues to provide fodder for experimentation. If scientists could 'refactor' model organisms — that is, recode their genomes to be simpler and more amenable to human understanding and tinkering — then science and biotechnology based on those organisms could proceed at an accelerated pace. On page 471 of this issue, Dymond et al.1 present a major advance towards this end: the construction of a functional, partly synthetic version of the S. cerevisiae genome.

Rewriting genomes to meet the specifications of humans has been a stated goal of synthetic biology for some time2. But the labour and expense involved in actually making such radical changes to genomes, coupled with uncertainties about the chances of improving on what nature spent billions of years perfecting, have meant that only a few serious tilts at genome re-engineering have been made. Previous noteworthy examples include the refactoring of 12,000 base pairs (about 30%) of a virus genome3 and the removal of 'amber' stop codons — nucleotide sequences that signal the termination of translation — from the bacterium Escherichia coli4, a feat that should allow researchers to rewrite portions of the bacterium's genetic code at will. And, of course, the genome of a Mycoplasma bacterium has been synthesized de novo by workers at the J. Craig Venter Institute in Rockville, Maryland, and used to infuse a working cell5.

Dymond et al.1 have now raised the bar by starting work on eukaryotes (organisms such as fungi, plants and animals), which have much larger and more complex genomes than bacteria. More specifically, the authors have replaced sections of two chromosomes of S. cerevisiae — 90,000 bases at the end of chromosome IX, and 30,000 bases at the end of chromosome VI — with synthetic DNA. Their eventual goal is presumably to replace the entire genome of 12 million base pairs with a human-designed sequence.

To make their synthetic DNA, Dymond et al. entirely removed 20 regions from the naturally occurring yeast chromosomes. Most of these regions were repetitive sequences (which can cause DNA segments to be deleted or even cause chromosomes to mis-segregate), or sequences that were non-functional or redundant. The authors also recoded all genes longer than 500 bases to contain 'watermarks' — sequences that allow the synthetic DNA to be easily differentiated from natural sequences using standard laboratory methods, but that do not change the sequences of proteins encoded by genes. As in the previously reported work4 in E. coli, Dymond and colleagues1 modified amber stop codons in the DNA of S. cerevisiae so that they could be recoded in the future, for example to encode unnatural amino acids for insertion into yeast proteins.

Astoundingly, the authors found that yeast cells containing the modified genome suffered no growth defects and displayed minimal differences in gene expression in comparison with the wild-type strain. The entire sequence of the artificially added DNA was faithfully reproduced by living cells, which is either a testament to the robustness of human engineering or a sign that God's fingerprints are fainter than creationists would have you believe.

In addition to the changes mentioned earlier, Dymond and co-workers introduced sequence elements known as loxPsym sites after every non-essential gene in their synthetic DNA, and at several other positions. In the presence of an enzyme called Cre recombinase, these loxPsym sites stochastically recombined with each other, either deleting or inverting the intervening sequence of DNA. The authors were thus able to generate a vast library of yeast genomes, containing all manner of random architectures, at will. Such libraries could be screened or evolved to find new yeast strains that are better suited to living in a given environment. Moreover, because yeast is used to produce alcohol, proteins and high-value organic compounds, new strains generated in this way might prove to be useful for industry, in the same way that a simplified E. coli strain has proved to be an excellent platform for producing large quantities of proteins6.

The obvious extension of Dymond and colleagues' work is to rebuild the entire yeast genome. However, given that the currently completed synthetic sequences represent only about 1% of the whole genome, rebuilding the remainder is a daunting task. One issue is that, even though the aggregated cost of the materials, apparatus and consumables used in DNA synthesis has been steadily decreasing, the construction of entire genomes remains inordinately expensive7.

Even more problematic is the cost of labour. A comparison of recent endeavours in genome synthesis and modification (Table 1) reveals that DNA synthesis at the scale of the yeast genome will require either armies of scientists — such as the wonderful group of undergraduate students currently working on similar projects8 with Dymond and co-workers — or new methodologies. The authors' landmark work1 confirms that automated DNA synthesis and assembly techniques are becoming necessary, and that the total synthesis of genomes is likely to supersede piecemeal approaches to genome modification. Given a little push here and there from technological advances, the age of designer genomes is nigh.

Table 1 Labour required for genome synthesis

References

  1. 1

    Dymond, J. S. et al. Nature 477, 471–476 (2011).

    CAS  Article  ADS  Google Scholar 

  2. 2

    Endy, D. Nature 438, 449–453 (2005).

    CAS  Article  ADS  Google Scholar 

  3. 3

    Chan, L. Y., Kosuri, S. & Endy, D. Mol. Syst. Biol. 1, 2005.0018 (2005).

    Article  Google Scholar 

  4. 4

    Isaacs, F. J. et al. Science 333, 348–353 (2011).

    CAS  Article  ADS  Google Scholar 

  5. 5

    Gibson, D. G. et al. Science 329, 52–56 (2010).

    CAS  Article  ADS  Google Scholar 

  6. 6

    Sharma, S. S., Blattner, F. R. & Harcum, S. W. Metab. Eng. 9, 133–141 (2007).

    CAS  Article  Google Scholar 

  7. 7

    Carr, P. A. & Church, G. M. Nature Biotechnol. 27, 1151–1162 (2009).

    CAS  Article  Google Scholar 

  8. 8

    Dymond, J. S. et al. Genetics 181, 13–21 (2009).

    CAS  Article  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding authors

Correspondence to Peter J. Enyeart or Andrew D. Ellington.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Enyeart, P., Ellington, A. A yeast for all reasons. Nature 477, 413–414 (2011). https://doi.org/10.1038/477413a

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.