Man selects only for his own good; Nature only for that of the being which she tends.” 
                                         - Charles Darwin

Sculpting Ecosystems

Wild organisms are exquisitely adapted to surviving and reproducing in their ancestral habitat. That means the vast majority of mutations will be harmful - and the same is true of human-designed changes.  Because natural selection typically eliminates our alterations, we have previously been unable to use biology to stably change ecosystems.

On the rare occasions that we can provide a fitness benefit - typically by engineering the environment - the responsible genes are hard to link to other traits and readily transfer into wild-type competitors.  Thus, if we want to engineer biology outside of the controlled environments of the laboratory, clinic, and factory, we must learn to establish, link, and maintain fitness advantages.

In microbes, we are working to privilege engineered strains over wild counterparts by conferring evolutionarily stable resistance to bacteriophages. Our goal is to use this technology to precisely replace target strains within complex microbial ecosystems.

In multicellular organisms, we are leveraging gene-level rather than organismal-level selection to counteract the penalty imposed by the fact that our changes impair the organism's ability to replicate. Gene drive systems benefit associated genome edits by distorting inheritance in their favor. We outlined how RNA-guided gene drives based on CRISPR could be used to drive many types of genomic alterations through wild populations over generations and demonstrated their efficacy in yeast.

Current projects in the gene drive area involve building and testing safeguards for the responsible development of these technologies, including the ability to reverse any unwanted genetic changes, to maintain costly traits in populations through cyclic changes, and to reliably predict how drive systems are likely to evolve in wild populations.

Perhaps importantly, we are developing local versions of gene drive systems to grant communities the ability to solve problems affecting their environments without forcing those changes on others. Daisy drive systems inevitably lose "daisy elements" over generations; when they run out, they cease to exhibit drive and become normal engineered genes. We're working to combine this feature with ways of blocking gene flow, thereby keeping engineered and wild populations separate and "pure", which in turn should enable communities to keep changes within their own political boundaries, and also to precisely restore the original wild-type genetics if needed.

Daisy drives are particularly important because we can't reliably predict the consequences of altering complex systems. The best approach is to make the smallest possible change that we think can solve the problem, but only at a local level. If it works well, we can scale up; if not, we reverse the change as best we can. Because they are inherently self-scaling, global drive systems violate these rules, and consequently should only be used to combat the most devastating diseases, such as malaria and schistosomasis, where we have few if any other options.