Feature: The meaning of (artificial) life

It was once widely believed that life differed fundamentally from non-life; the defining characteristic of life was the existence of some insubstantial vital force – an élan vital – for which no physical process could account.

The first blow to this theory was dealt by German chemist, Friedrich Wöhler, in 1828, when he synthesised urea in the lab, demonstrating that even ‘vital’ organic compounds were formed from the same lifeless building blocks as common inorganic compounds. Yet, urea alone isn’t alive.

It took a further 182 years for the notion of vitalism to finally be put to rest with a paper produced by the team at the J. Craig Venter Institute (JCVI), published in Science on May 20. As stated by Craig Venter during his press conference, the paper details the creation of the “first self-replicating species that we’ve had on the planet whose parent is a computer.”

Venter and his team, in the culmination of 15 years of toil, had synthesised an entire genome from “four bottles of chemicals”, assembled it, and transplanted it into a host cell of a different species, which then went on to replicate using the synthetic DNA as its blueprint. Noticeably absent from that blueprint was élan vital.

But how significant is this development? Despite the hyperbolic headlines that sprang up in the mainstream press announcing that we had, at last, created life by our own hand, the actual achievement was somewhat more modest. Venter’s team didn’t achieve abiogenesis.

They also didn’t build an organism entirely from scratch – the synthetic chromosome took occupancy in the cellular abode of an existing bacterium. And the genome they used was largely designed by nature, if assembled in a lab.

But what Venter’s team did achieve was a technical triumph in synthetic biology, says Dr Kirby Siemering, of the Australian Genome Research Facility. “It’s a major milestone in terms of being a proof of principle demonstration of the technology.”

And this technological breakthrough will have potentially far reaching ramifications for the burgeoning discipline of synthetic biology, with Venter, among others, hoping it will open the door to designer organisms with a manifold of uses, such as production of biofuels, vaccines, biologics and organisms for bioremediation.

Three steps to artificial life

The actual process to create the synthetic version of Mycoplasma mycoides, dubbed M. mycoides JCVI-syn1.0. was a three step affair, beginning with an excruciatingly accurate sequence of the bacteria in question.

In fact, it took two sequences of the 1.08 million base pair (bp) genome to achieve sufficient confidence in the accuracy, one a standard lab strain, the other a cloned version produced by the JCVI. The two sequences differed only at 95 sites, which is suggestive of the differences between the organisms rather than errors in sequencing, says Siemering.

---PB---

Venter’s team then added a few watermarks to clearly identify JCVI-syn1.0 from any of its wild brethren. They even developed their own ‘language’ for encoding alphanumeric characters and punctuation in DNA, which was a necessary measure, said Venter.

“Mathematicians have been hiding and writing messages in the genetic code for a long time, but it’s clear they were mathematicians and not biologists because, if you write long messages with the code that the mathematicians developed, it would more than likely lead to new proteins being synthesised with unknown functions.”

The watermarks, encoded in four ~1150 bp sections, include the names of 46 contributors to the project, along with a Web address (an artefact that may bamboozle future scientists should they decode the watermark many, many years from now), along with three quotations. The latter were added “because, with the first genome, we were criticised for not trying to say something more profound than just signing the work,” said Venter.

The next step was actually assembling the artificial genome, which they accomplished by breaking it down into 1078 cassettes of 1080 bps each, with an 80 bp overlap between adjacent cassettes. These were then synthesised by Blue Heron, in Washington in the US.

Each cassette was rigorously verified by Blue Heron, although there were some errors to be corrected once the cassettes were received due to an early draft sequence being sent to Blue Heron which differed slightly from the final desired sequence. Many of these differences – mainly small insertions or deletions that weren’t predicted to be located in genes – were left uncorrected, with a handful of biologically significant differences corrected using polymerase chain reaction (PCR), in vitro recombination, aided by a QuikChange II mutagenesis kit.

The cassettes were then assembled in yeast, and transferred to E. coli, using methods developed by the JCVI and outlined in a 2007 paper that appeared in Science. The individual 1 kbp cassettes were assembled first into 10 kbp segments, then a second stage fused them in to 100 kbp segments, and after a several stage refinement and verification process, into the final 1 Mbp chromosome.

However, it was at this point that the team ran into the first of its potentially show-stopping problems. The chromosome assembled in yeast looked for all intents and purposes like the real thing, yet they were unable to ‘boot up’ any cells using it. It took the team two years to find out why.

---PB--- “It turns out the DNA in the bacterial cell was actually methylated,” said Venter. “And the methylation protects it from the restriction enzyme, from digesting the DNA.” Yet the DNA assembled in yeast was unmethylated. As a result, the team had to develop a way to methylate the genome prior to transplantation. For good measure, they also knocked out the restriction enzyme from the host cell, Mycoplasma capricolum.

Things were looking promising for the team’s first successful transplantation of M. mycoides JCVI-syn1.0 into the M. capricolum host cell, but then another problem emerged that ground the project to a halt: the synthetic chromosome still wouldn’t boot up.

Eventually, after months spent combing through the assembled sequences of the artificial genome, the team discovered an error in a single base pair. This one-in-a-million error – a frameshift in the gene dnaA, which is essential for chromosomal reproduction – had rendered their artificial genome ‘lifeless’.

“There’s parts of the genome where it cannot tolerate even a single error,” said Venter. “And then there’s parts of the genome where we can put in large blocks of DNA, as we did with the watermarks, and it can tolerate all kinds of errors. So it took about three months to find that error and repair it.”

The final step was transplanting the artificial M. mycoides JCVI-syn1.0 into the M. capricolum cell and seeing if it would replicate. Even after everything that had proceeded, this was the step that was the most astounding, says Professor Ian Paulsen, from Macquarie University.

“I was talking to the team at JCVI a few years ago and they were describing to me what they were planning to do. I remember thinking ‘that’s crazy, that’s never going to work’. I’m still astonished you can put a synthetic genome into a bacterium and get it to displace its own genome.”

Once that step had been successfully completed, it was only a matter of time before the first motes of blue colonies of M. mycoides JCVI-syn1.0 appeared. The team then verified that the new growths were of the synthetic variety by identifying the watermarks and by sequencing the new colony.

Interestingly, they found a few difference between the growing bacteria and the original synthetic genome, including some duplications and an insertion that the synthetic bacteria picked up from E. coli. Even though these changes interfered with two genes, they didn’t impact the viability of M. mycoides JCVI-syn1.0.

More importantly, the didn’t find any traces of M. capricolum, indicating the synthetic genome had entirely displaced the original. They even found that M. mycoides JCVI-syn1.0, while it appears similar to its natural M. mycoides cousin, it grows slightly faster.

---PB---

It’s life, Jim. . .

The synthesis and transplantation of M. mycoides JCVI-syn1.0 is the culmination of 15 years work by the JCVI, and while it represents the end point of the lab’s initial goal of creating a viable synthetic genome, it’s just the starting point for the broader synthetic biology programme.

It took a number of incremental developments over the past decade and a half, along with profound increases in the capability of the technology, to produce the artificial genome, but now the real adventure begins in shaping artificial genomes to perform tailored functions on demand. It’s this that is Craig Venter’s true goal.

The hope is that synthetic biology will offer greater power and flexibility to custom make organisms than does current genetic modification, says Siemering. “The major advantage is that you can optimise the organism completely,” he says.

“With simple gene splicing of existing organisms, the process you’re intending to use it for is not its natural function in nature, so it still has a lot of redundant or non-optimal processes that it carries out. The benefit of synthetic biology is having something that is optimised for the purpose it’s intended for, such as having the maximal conversion rates for biofuels.”

Venter’s aim is build up a toolkit for synthetic biology, starting with the minimal set of genes required for life. This harkens back to his Minimal Genome Project, which used a relative of M. mycoides, M. genitalium, which has a genome consisting of only half a million base pairs and 482 genes. (In fact, M. genitalium was originally planned to be the basis of the first synthetic genome, but the JCVI team found that it replicated too slowly, so they switched to the faster growing M. mycoides, even though it had a genome twice the size of M. genitalium.

However, advances in sequencing and synthesis technology made the larger genome a minor inconvenience compared to the advantages of having a much faster growing bug.) This minimal organism can then be used as the foundation for tailor made life that can be applied to various tasks, such as biofuel production.

Venter already has his own private spin-off company, Synthetic Genomics, which plans to use synthetic biology technology to develop products in environmental genomics, microbiology, biochemistry, bioinformatics, plant genomics, genome engineering, synthetic biology and climate change.

As the authors of the Science paper state: “we expect that the cost of DNA synthesis will follow what has happened with DNA sequencing and continue to exponentially decrease. Lower synthesis costs combined with automation will enable broad applications for synthetic genomics.”

---PB---

According to Siemering, this also raises the notion of gene prospecting, where various ecosystems are scoured for new genes that can be appropriated for biotech applications. “Australia has a very rich biodiversity,” he says. “We we’ll have a lot of gene content that will be useful for synthetic biology.”

And this future of custom made organisms touted by Venter may be only a few years away, says Paulsen. “It never pays to bet against Craig. It’s not something we’ll see in the next few months. But it’s certainly feasible in the next couple of years. Or it might turn out that there are further technical problems to overcome, and it’ll be five years or more.”

Ultimately, the creation of M. mycoides JCVI-syn1.0 was less a Frankenstein moment, and more an Apollo Moon landing moment: one small step in a long series of incremental advances, but a giant symbolic leap for synthetic biology. The creation of life using the parts of one organism and a synthetic genome designed in a computer and assembled from raw chemicals in a lab is one thing, but designer organisms built entirely from scratch – that’s more like flying to Alpha Centauri.