Scratchbuilt genomes
[Here’s the pre-edited version of my latest story for Nature’s online news. I discuss this work also in the BBC World Service’s Science in Action programme this week.]

By announcing the first chemical synthesis of a complete bacterial genome [1], scientists in the US have shown that the stage is now set for the creation of the first artificial organisms – something that looks likely to be achieved within the next year.

The genome of the pathogenic bacterium Mycoplasma genitalium, made in the laboratory by Hamilton Smith and his colleagues at the J. Craig Venter Institute in Rockville, Maryland, represents an increase by more than a factor of ten in the longest stretch of genetic material ever created by chemical means.

The complete genome of M. genitalium contains 582,970 of the fundamental building blocks of DNA, called nucleotide bases. Each of these was stitched in place by commercial DNA-synthesis companies according to the Venter Institute’s specifications, to make 101 separate segments of the genome. The scientists then used biotechnological methods to combine these fragments into a single genome within cells of E. coli bacteria and yeast.

M. genitalium has the smallest genome of any organism that can grow and replicate independently. (Viruses have smaller genomes, some of which have been synthesized before, but they cannot replicate on their own.) Its DNA contains the instructions for making just 485 proteins, which orchestrate the cells’ functions.

This genetic concision makes M. genitalium a candidate for the basis of a ‘minimal organism’, which would be stripped down further to contain the bare minimum of genes needed to survive. The Venter Institute team, which includes the institute’s founder, genomics pioneer Craig Venter, believe that around 100 of the bacterium’s genes could be jettisoned – but they don’t know which 100 these are.

The way to test that would be to make versions of the M. genitalium genome that lack some genes, and see whether it still provides a viable ‘operating system’ for the organism. Such an approach would also require a method for replacing a cell’s existing genome with a new, redesigned one. But Venter and his colleagues have already achieved such a ‘gene transplant’, which they reported last year between two bacteria closely related to M. genitalium [2].

Their current synthesis of the entire M. genitalium genome thus provides the other part of the puzzle. Chemical synthesis of DNA involves sequentially adding one of the four nucleotide bases to a growing chain in a specified sequence. The Venter Institute team farmed out this task to the companies Blue Heron Technology, DNA2.0 and GENEART.

But it is beyond the capabilities of the current techniques to join up all half a million or so bases in a single, continuous process. That was why the researchers ordered 101 fragments or ‘cassettes’, each of about 5000-7000 bases and with overlapping sequences that enabled them to be stuck together by enzymes.

To distinguish the synthetic DNA from the genomes of ‘wild’ M. genitalium, Smith and colleagues included ‘watermark’ sequences: stretches of DNA carrying a kind of barcode that designates its artificiality. These watermarks must be inserted at sites in the genome known to be able to tolerate such additions without their genetic function being impaired.

The researchers made one further change to the natural genome: they altered one gene in a way that was known to render M. genitalium unable to stick to mammalian cells. This ensured that cells carrying the artificial genome could not act as pathogens.

The cassettes were stitched together into strands that each contained a quarter of the total genome using DNA-linking enzymes within E. coli cells. But, for reasons that the researchers don’t yet understand, the final assembly of these quarter-genomes into a single circular strand didn’t run smoothly in the bacteria. So the team transferred them to cells of brewers’ yeast, in which the last steps of the assembly were carried out.

Smith and colleagues then extracted these synthetic genomes from the yeast cells, and used enzymes to chew up the yeast’s own DNA. They read out the sequences of the remaining DNA to check that these matched those of wild M. genitalium (apart from the deliberate modifications such as watermarks).

The ultimate evidence that the synthetic genomes are authentic copies, however, will be to show that cells can be ‘booted up’ when loaded with this genetic material. “This is the next step and we are working on it”, says Smith.

Advances in DNA synthesis might ultimately make this laborious stitching of fragments unnecessary, but Dorene Farnham, director of sales and marketing at Blue Heron in Bothell, Washington, stresses that that’s not a foregone conclusion. “The difficulty is not about length”, she says. “There are many other factors that go into getting these synthetic genes to survive in cells.”

Venter’s team hopes that a stripped-down version of the M. genitalium genome might serve as a general-purpose chassis to which might be added all sorts of useful designer functions, for example including genes that turn the bacteria into biological factories for making carbon-based ‘green’ fuels or hydrogen when fed with nutrients.

The next step towards that goal is to build potential minimal genomes from scratch, transplant them into Mycoplasma, and see if they will keep the cells alive. “We plan to start removing putative ‘non-essential’ genes and test whether we get viable transplants”, says Smith.

References
1. Gibson, D. G. et al. Science Express doi:10.1126/science.1151721 (2008).
2. Lartigue, C. et al. Science 317, 632 (2007).