Synthia Gets a Shotgun - Goodbye genetic engineering?

English

What do ocean-going yachts, space-traveling bacteria and synthetic life have in common? J. Craig Venter, of course.  The self-styled genome tycoon has been busy pushing the boundaries on what may appear at first glance to be unrelated enterprises. Nothing could be further from the truth. A suite of recently uncovered patent applications lodged by Venter and his colleagues reveal not only an attempt to grab ownership over much of synthetic biology (see news release) but also a breathtakingly bold business plan for producing millions of new synthetic organisms per day. At the heart of this are plans for a new, automated process enabling rapid assembly of complete synthetic genomes - plans that, if realised, could render current genetic engineering techniques quaint and obsolete. Venter calls it "homologous in vitro recombination" or "combinatorial genomics." ETC suggests it might be properly dubbed "shotgun synthesis" and it has the potential to blast apart current biotech practice.

These days genetic engineering, the standard practice of transferring a piece of DNA from one organism to another, is a routine, plodding and rather old-fashioned lab technique. There are several steps involved - ranging from identifying suitable DNA, cutting it out of its host organism, transforming it into a circular strand of DNA (a plasmid) and then somehow lodging it in the genome of the organism you intend to alter using a virus or a gene gun method. For some years synthetic biology has promised to speed up the front end of that process (the acquiring DNA bit) since it's now possible to specify on the Internet exactly what DNA you want. Within a fortnight of clicking the order button, DNA synthesis foundries such as GENEART or Blue Heron can cheaply provide a custom-made plasmid ready for genetic engineering.

At present most synthetic biologists engineer their synthetic DNA into trusty lab microbes such as E. coli or yeast in order to see if their designer gene sequences "work." Consider a simplistic computing analogy - the E. coli genome is the equivalent of an operating system such as Windows or OSX and the circular engineered strand of synthetic DNA is a programme that the researcher hopes will carry out some task such as making a protein or altering behavior of the organism. Of course biology is much messier than computing but that over-simplified metaphor is nonetheless a basic conceit of synthetic biology. Venter's quest to create a novel organism with a minimal genome ( which we call "Synthia") is an attempt to create a stripped-down operating system that's leaner and meaner than E. coli and on which synthetic DNA programmes could be implemented. Venter's recently published portfolio of patent applications seeks to own this operating system and the method of adding extra synthetic DNA to it.

Surprisingly, however, Venter's approach does away with genetic engineering altogether. Instead of constructing a plasmid and blasting it into the existing genome of a living organism, Venter and his colleagues will place the entire synthetic genome into a bacterial cell and make that cell "boot up." Significantly, they have broken down the entire minimal genome into 101 short fragments of DNA or "gene cassettes." Any one of these cassettes can be removed and replaced by a synthetically altered cassette or, indeed, extra synthetic cassettes could be added (so let's say there might be 102 or 103 fragments rather than the original 101). The patent applications then describe a method by which all of those fragments are mixed in a reaction chamber and assemble themselves in one go into a complete genome. The inspiration for this miraculous assembly is a favourite bug of Venter's - an almost un-killable microbe called Deinococcus radiodurans that is thought to be able to travel in outer space because it can survive radiation doses that are three to five thousand times the lethal dose for humans. Even though radiation bombardment shatters the genome of Deinococcus radiodurans into tiny fragments, the bug uses "repair proteins" (specialised enzymes) to piece its genome back together again. Like magic, it can be fully functioning again within 24 hours. Craig Venter has called D. radiodurans "the ultimate genome assembly machine". E. coli, in fact, has similar repair proteins. By using these in a purified form along with a mix of other enzymes Venter's patent applications suggests that it's possible to swiftly assemble the fragments of the Synthia genome, including extra gene cassettes, into a complete genome. Theoretically, it's a new, fast method of engineering new DNA into the genome.

For those familiar with Venter's past work this "fragment-then-piece-back-together-again" approach might sound oddly familiar. It's the principle behind his 'shotgun sequencing' method, which his private team at Celera used to decode the human genome faster than the U.S. government project. In shotgun sequencing the genome of an organism is blasted into small fragments that are rapidly sequenced in parallel and then reassembled inside a computer into one complete digital sequence.  In this new method, Venter is once again fragmenting the genome and re-assembling his fragments, this time in vitro - not in a computer - and with some extra DNA fragments deliberately thrown into the mix. The inventor of shotgun sequencing appears to be developing "shotgun synthesis" and his ambitions for this new technique are far from modest.

Not content with creating a faster method of genetic engineering, Venter is looking to emulate the robotic methods used in drug discovery to further speed up the creation of new life forms. In the patent applications his team describe a process that Venter calls "combinatorial genomics." This precisely matches a process described in a Wired magazine interview several years ago:

"If you want to find the role of 100,000 genes, Venter says, the trick is to find a way of doing 100,000 experiments at once. All you would need that's not already available is a synthetic genome, a sort of all-purpose template onto which you could attach any gene you wished, like inserting a blade onto a handle. You could then test the resulting concoction to see if it performed a specific vital task, such as metabolizing sugar or transporting energy. Using existing robotic technologies, you could do thousands of such experiments at once, in much the same way that a combinatorial chemist tests thousands of chemical compounds simultaneously to see if they have the desired effect on a target molecule. Most will not. But the ones that do can be investigated further. 'I call it combinatorial genomics,' Venter tells me. 'It's one of my better ideas if it works. In fact, it's one of my better ideas if it doesn't work.'"

In fact, the recently published patent applications claim that this process should enable the automatic production of not just thousands but "millions of different genomes" - a claim that Venter also made recently in this TED talk where he outlined his combinatorial genomics vision. Specifically, the patent applications describe a robotic system that rapidly assembles and tests synthetic genomes in a fast throughput model either by "installing" them into cells or by making them express themselves in a "cell-free environment... comprising the necessary transcriptional and translational machinery to express genes" - probably something similar to these microfluidic chips developed by David Kong of MIT which mix cellular contents in tiny silicon chambers.

If Craig Venter does indeed develop the capacity to create millions of new synthetic organisms per day he will also need a ready supply of thousands of new unexamined genes to test in his Synthia operating system. Here, too, his team is already way out in front. His Sorcerer II Expedition spent 2 years trawling the world's oceans collecting organisms in seawater samples to be rapidly sequenced in the new field of metagenomics. Metagenomics takes the study of genomes to the level of entire ecosystems. Ignoring the messy boundaries of individual species, metagenomic sequencing faithfully records all the genes found at a particular location without trying to ascribe them to this, that, or another organism. In effect the whole genetic material of an ecosystem is presented as a soup. Geneticists can then comb through the gene sequences in this undifferentiated soup to identify genes with similar structures and properties.

Earlier this year Venter's team announced they had so far identified 6.12 million new proteins uncovered from 7.7 million genetic sequences from the first phase of the Sorcerer II's voyage. Within this wider dataset are hundreds of thousands of similar genes - photoreceptors, for example, that might be used to engineer synthetic organisms that convert sunlight to hydrogen.  Attempting to engineer those genes one at a time into an existing microbe would be a daunting challenge with conventional genetic engineering techniques, but with the "shotgun synthesis" and combinatorial genomics approach, such a challenge is - at least theoretically - more do-able. Venter hopes it may yield hundreds, maybe thousands of industrially useful proteins. Venter's metagenomics prospecting doesn't stop at sea, his Institute is also sampling airborne bacteria in downtown Manhattan and Venter half-jokes that he hopes to create  "Whole Earth Gene Catalog" which sounds not unlike proposals Venter made to Google to make all the genes in the world Googlable. Indeed Venter's ambitions don't stop at Earth - rumours are that Venter has also talked of sampling bacteria from the edge of the atmosphere in the hope of finding DNA from outer space. In our report, Extreme Genetic Engineering, we joked that the ability to digitise and beam back life forms could create a new form of "star trek biopiracy." Little did we realise how literally Craig Venter would boldly go there...