SCRaMbLEing the genome – evolution

Sc2.0 synthetic chromosomes contain recombination signals designed to allow removal of as many non-essential genes as possible, provided that these deletions do not interfere with rapid growth. This is referred to as genome minimization. To effect this and many other changes of the genome we designed a system called SCRaMbLE – Synthetic Chromosome Recombination and Modification by LoxP-mediated Evolution.

Each potentially dispensable ORF carries a loxP site in its 3′UTR. Using this strategy, all genes will initially retain their original locations on the chromosomes. As synthetic chromosomes (or eventually, the whole genome) is finished, Cre can be expressed in a inducible fashion. Two methods we typically use to control Cre are:

  1. A galactose inducible promoter driving expression of Cre;  synthetic strains can be exposed to galactose for a few minutes or hours, then put back into glucose
  2. A Cre-estrogen binding domain fusion, regulated by estradiol (estrogen).

Very subtle pulses may be required, as strains with very large numbers of loxP sites become extremely sensitive to Cre. Either Cre induction regimen outlined above could be repeated for days or weeks. One can simultaneously select for the fastest growing yeast. The survivors will, by definition, retain indispensable genes; we would be “letting the yeast tell us” what gene sets are consistent with fast growth. It is quite likely that there will also be rearrangements of gene order that emerge from this strategy that actually outperform wild type yeast in growth (or other phenotypes). The presence of the PCRTags in each gene, or deep sequencing methods, will allow us to rapidly and readily assess which genes were retained and which ones lost after such an outgrowth experiment, and will also be useful for detecting the breakpoints of rearrangements.

We acknowledge that there is significant risk in our approach, because we will (somewhat blindly) choose a “starting point” for our genome reengineering, as well as an order in which the genome will be reengineered. There are near-infinite permutations of the order in which these steps could be done and it may well be that success is possible by some but not other permutations. Precisely because of this danger of “painting ourselves into a corner”, by introducing genome-wide loxP sites we have engineered an evolutionary “escape hatch” into our strategy. As we synthesize the chromosomes, we tag all nonessential gene 3′ UTRs with loxP sites, as well as points where deletions are made such as tRNA genes and transposon loci, ultimately resulting in ~5000 sites genome-wide. If significant effects on fitness are observed that accumulate chronically (rather than suddenly on replacement of a particular 30kb segment), we can carry out in vitro evolution as outlined above, with or without using pulsatile expression of Cre in growers.