Monday 24 May 2010

Craig Venter's Synthetic Bacterium

Wow, it's been a while since I posted. I have been getting occasional prods to write something, and everything seems to have fit into place for it to happen now, so here we go.

Much has been said over the announcement last week of the creation of a "synthetic life form" by the J. Craig Venter Institute. Much of that has been woefully uninformed, which is probably the topic of another post. The news is impressive, to be sure, but it is not really that groundbreaking. The actual procedure is the bacterial equivalent to the procedure that produced Dolly the cloned sheep, technology that, while remarkable, is hardly news (although I would imagine that it might be more difficult to do with a bacterial cell than a mammalian one). The big advance in this project was the fact that the genome that was introduced into the cell was completely synthetic -- its sequence designed on a computer, and its chemistry produced in a laboratory. In fact these tasks are both routine in biology and biochemistry laboratories. The difference here is the application: specifically, the combining of the synthetic-DNA and cloning procedures. The fact that we have a functioning cell with a completely synthetic genome is a remarkable outcome, to be sure, but the principle has been an obvious one for a couple of decades now. If one were to compare this to a society-changing invention, it would be to Henry Ford's adoption of the assembly line to produce the same cars he had been making for several years, rather than the Wright Brothers' first powered flights.

Another aspect that gets sometimes mentioned but (I suspect) overlooked is the nature of this new organism's genome. True, it did not exist in any chemical state before its synthesis by human researchers, which is to say that it was not spliced together from fragments of preƫxisting genomes taken from extant viable organisms. However, that is true only in the chemical sense. The actual sequences for the genes used for this genome were taken from extant viable organisms, and if one were to do a phylogenetic analysis of any of those genes, the new organism would be a sister lineage for the source of whichever gene was used, not a completely separate branch from any other life form. In fact, as far as I understand it, there is nothing to distinguish the actual organism from a hypothetical one made from spliced bits of the same source genomes -- in other words, a clone in the more traditional (Dolly-the-sheep) sense. This work, important though it unquestionably is, is not the creation of a completely new life-form, but a proof-of-concept that such a creation is possible.

There are one minor and two major hurdles that need to be overcome before we actually do have the sort of unprecedentedly novel life-form that many in the press and the public think this is. The minor one is that the new organism that Venter and his team have created is based mostly on genes from an obligate parasite with highly specific environmental requirements: it is not what we would unreservedly call "free-living". Parasites can be tricky to grow in the lab, and parasites potentially missing parts of various biochemical pathways (which this one is likely to be) can be a nightmare to keep alive, even if they start off growing vigorously. Venter's new organism has a synthetic genome, yes, but it has been injected into another cell whose genome had been removed, leaving all the rest of the life-giving biochemical machinery intact. While it is estimated that the new organism should replace all of that in about twenty generations, it is still possible that (for instance) some important and overlooked protein is also long-lived enough to keep the cell and much of its progeny alive for quite a while before finally giving out and causing the population to crash.

In fact this points to one of the most important applications of this research. In spite of all of the attention given of late to genomics, proteomics, transcriptomics, interactomics, and a number of other "omicses", we really have only an inkling of what is actually required to keep a cell functional. One obvious future direction of this research is to produce organisms with various subsets of the current genome, to determine exactly what the minimal set of genes is to keep something alive. I hasten to add that the subset that would eventually be arrived at would be specific to the starting set of genes in the genome; other starting points will almost certainly result in other final genomes. In other words, what might be the smallest number of genes needed to keep a Mycoplasma-derived organism alive might still be more than the smallest number of genes required for (say) a Rickettsia-derived organism, even if the original genome is larger. Nevertheless, given that the Mycoplasma genome is the smallest known, this is a good starting point.

So the minor hurdle that I mentioned earlier is that, while the genome is supposed to be functional from a minimal standpoint, there is no guarantee at this stage that it is in fact indefinitely viable; Venter may already have stripped the genome past its minimal complement of genes, and omitted some critical component to the genome without which the cell can eke by for several generations. Time will tell whether this is in fact the case, of course, and if it is, the remedy is obvious. Venter and his team have proven that they can generate a synthetic bacterium once; if this one fails, they need only do it again with a different genome, and repeat as necessary until they have something that works.

That leads to one of the major hurdles. This is a new genome, yes, but as I said already, it was taken from bits of other organisms' genomes. Probably the greatest milestone in synthetic biology will be when we are able to design completely new genes. This would amount to deciding what chemical reaction we want the gene's product to perform, how we want it regulated in the cell (specifically, what other genes and gene products it must interact with), and with that information alone coming up with a previously unknown string of A's, C's, G's and T's that can be inserted into a genome to do exactly what we ask of the new gene product. A fully synthetic life-form would have all of its genes so constructed, and none taken from already-extant organisms. This is precisely opposite to Venter's new organism, and its advent is probably decades away still, barring some major and sudden advance in our understanding of protein function and generation.

Meanwhile, the fact that this technique requires that the new genome be injected into a "recipient" cell -- a previously viable organism that has left all of its life-sustaining machinery for its new genome to use -- is itself the other major hurdle to making a completely novel organism. While a truly and completely synthetic organism would need entirely novel genes, it would also need to be made without using any pre-existing cellular materials. This, and this alone, would be the actual creation of new life, similar to what happened naturally in Earth's primordial oceans some four billion years ago. We can already make coacervates, lipid vesicles similar to cell membranes, in the laboratory -- in fact, it was an exercise in my high school biology class! Putting those vesicles into the proper context for a living cell is probably not so easy, but also not impossible. Generating life in this sense will also probably require a rather different biochemistry from what Venter's techniques require. It may not be possible (or if so, prohibitively difficult) to make it work with the DNA-and-protein biochemistry used by all currently extant life. It may instead be necessary to replicate the "RNA World" that is thought to have preceded it, and probably in a progenotic sense, meaning that while all functions necessary for life would be carried out by the complete population of protocells, no individual cells would carry the full complement of genes required for those functions.

What all this amounts to is that Venter's procedure is a top-down approach: taking pre-existing components and combining them to create a new combination. The further steps down this road amount to trying to figure out which components can be removed without causing the system to collapse, and which can be simplified. This is a laudable goal, and I congratulate Venter and his team for making the advances that they have. But the truly groundbreaking work, based on the bottom-up approach, generating new life from whole cloth, has yet to be accomplished, and will not come soon.


Lucas Brouwers said...

You're absolutely right in noting that this isn't 'life from scratch', but your hypothetical 'true' synthetic lifeform is a bit of a strawman considering the goals of the Venter team.

Suppose the programmers of Adobe want to build the next version of Photoshop. What they -could- do is start by assembling a computer, down and dirty messing with electronic circuits and such. After the machinery is there, they need to build a compiler, write an assembly language and program a text editor so that they can begin to write their program. From scratch.

What I'm trying to say is that bottom-up strategies are very effective if you want to fully understand how something works. If you want to get something done however, it might make more sense to start from the top, trying to make some small hacks to see if you can understand how the system you want to manipulate works. If the Adobe programmers come up with a new and improved version of Photoshop, nobody will doubt that they came up with a new and improved (synthetic, if you will) design of a successful program.

If the Venter-team wants to build a plastic-digesting bacterium now, top down seems to be the way to go. They want to be the next Henry Ford, with synthetic bacteria rolling out of their factories en masse.

-DG said...

Great post but I offer one correction from a bit of info at the beginning. Describing the genome as having been designed on the computer isn't actually true. This is the same genome the team has used in previous work, including the big one two seats ago, and was initially constructed using classical techniques. This time they did start with a digital representation but that digital copy was the result of sequencing the genome they had peiced together earlier

Opisthokont said...

DG -- Thanks for the correction. I suppose I should be more careful with my words: "designed" carries connotations of the sort of bottom-up work that Venter's team did not do. Perhaps the word "planned" or "specified" would be better. Like I said later on in my post, this is one of the major reasons why this work is really not the big deal it can easily be spun as: the new genome is really just a reshuffling of bits of already-extant ones. However, I think the point is still relevant that the new genome existed as computer data before ever being chemically synthesised, and that is a pretty impressive accomplishment.

Lucas -- Thanks for the comment. I don't mean to be setting up strawmen, though; I see in retrospect that I should have been clearer that 'life from scratch' was not Venter's goal for this project. The things that Venter is best known for (this, the Sargasso Sea metagenome, the human genome) are actually very conservative projects, things with foreseeable outcomes using extant technology. Bottom-up synthesis of life fits none of his criteria, and I suspect that even when we get it to work it will be a very long time before we are able to get it to do anything practical. You are perfectly right: if you want to get a specific task done, the top-down approach is almost always the best one. And while my personal penchant for the interesting but impractical may hide this, I do agree that the world needs its Henry Fords every much as it does its Wright Brothers. Or should that be Wright Brotherses? Hmm.