Craig Venter Reflections: Building & Redesigning Genomes

The last bit of Craig Venter's career a followed, often a bit green with envy, was his work to build genomes.  This all revolved around the Mycoplasma genitalium they had knocked out a genome as a victory lap after completing the first bacterial genome.  This bug is held up as the bacterium with the smallest genome which can be grown in a defined medium.  To first reconstruct an entire genome, with watermarking sequences to prove it was done, required pushing every limit of the existing gene synthesis cloning technology.  An effort so intense, that collaborator Dan Gibson became a verb! (See that gene cluster we want?  We'll Gibson it into a BAC vector.")

Way back in graduate school I had wondered about the excitement around work in George Church's lab to build a 96-well oligo nucleotide synthesizer - we needed sequencing primers but how thrilling is that?  Then a postdoc pointed out gently how I wasn't being very imaginative - it was the first step to a genome synthesis system!

But while I sometimes daydreamed, that didn't lead anywhere.  At Codon Devices some of my coworkers set a then-record by synthesizing 100 kilobases - but Mycoplasma genome was five-fold that.  So Venter, Clyde Hutchinson, Ham Smith and a gaggle of both long time Venter collaborators as well as new recruits built all sorts of tools and techniques enabling construction of pieces in E.coli then Saccharomyces, culminating with the publication of a complete genome.  (Hmm, I wonder how much my kibitzing on the original publication is reflected in corrected sequences in the synthesized version?)

But they didn't stop there.  Next came minimizing the genome - identifying genes which could be left out.  And then later identifying pairs of genes that could not be deleted singly, only as a couple.  

There's a small community of researchers who have thrown every possible computational tool at the minimized "JCVI syn3.0" genome.  Frustratingly, there are still genes in that minimal set whose function remains a mystery - I suspect Venter was not happy with that.  And at 473 genes, it's still quite complex.  Much smaller genomes have been found in insect endosymbionts, but those don't grow in culture (at least, not yet).

Unlike his EST and shotgun sequencing efforts, building genomes has not become routine.  Gene synthesis is still far too expensive, and the cycle times are too long.  There's a public effort to rewrite Saccharomyces, a few synthetic E.coli genomes (one I will probably write up soon), a Caulobacter effort which I need to dive into  - and that seems to be it.  There's been talk of resynthesizing an entire human genome, which would be extremely valuable for biomedical research but would probably ignite all sorts of public outcry, but it never seems to get beyond talk.  

One of the most interesting, perhaps final contributions from Venter in this regard was a recent experiment that showed the recipient cell can be legally dead - unable to replicate - and still accept a donor genome.  So two pieces, neither of which fit most definitions of "alive', can be combined to make something that definitely fits all definitions of "alive".  That biology can't cleanly define its subject matter is a philosophical debate for the ages, but it yet another example of Craig Venter pushing the boundaries of biological science.