On Venter envy

At one point, I think I spawned the phrase “Venter envy” to describe the feeling that many of us who are not corporate titans that can make entire chromosomes at will must feel. Surely this applies to you, yes? Well, anyway, what are the little people supposed to do? The answer to this vexing question has come, at some level, from that champion of the oppressed, George Church at Harvard. George recently published the follow-up to his most excellent MAGE paper, that showed how one can use small oligonucleotides to site-specifically alter mutiple sites in a bacterial chromosome in parallel (Wang et al. (2009), Nature 460:894). As they say in that great film, Treasure of the Silicon Valley, “Assembly? We don’ need no stinkin’ Gibson assembly.” I think MAGE is very cool, almost as cool as PACE (what is it with these Harvard folks and their acronyms? Can’t they at least invent science-y faux words, like ‘aptamer?’). I also think that it will likely be utilized in industry (certainly any industry that George starts up) and perhaps by more than a few academics.

Unfortunately, though, the efficiency of MAGE is extremely low, and while it does not require automation for success (in Texas, automation = undergraduates), it surely helps. Said automation is not beyond the reach of most academics, but it is something that requires the substantive investment of a scientific superstar like Church. This is another something I’ve been musing about for a bit, the development of upper level ’syncytia’ that freely exchange ideas, materials, personnel, and of course money between academia, industry, and government. If you can get your DNA synthesized on chips at low cost for demonstration purposes, and then hand off the products to a pendant company, it surely does help do genomic engineering projects. Sadly, I think such a model is actually a good idea for American science, but not all of us can operate that way. This really does raise the question of: what are many Universities doing in the research game? Go big, or go home (and if many of the recent “efficiency” critics of Texas Universities have their way, it will be “go home;” yeah, Massachusetts, cry me a river).

But I (typically) digress. MAGE is cool, MAGE is the nads. And CAGE, the horizontal shuffling between MAGE’d strains that the Church lab invented more recently (Isaacs et al. (2011), Science 333:348) further allows genomes to be built to spec, just like Craig Venter did, but again with many fewer oligos. To that end, Church and company are making us the Amberless Coli, where an entire stop codon is replaced with other stop codons, and thus is ‘recaptured’ into the genetic code, for other uses.

However, this all again involves very low efficiencies (“On average, 59 clones (10^-6 frequency) were observed per recombination.”). Wow. So, starting from a process where we already have typically 10^-6 frequencies of transformation / oligo, and where we’re trying to put together multiple such oligos in a single strain, we also have to pray for the outcome of the recombination event? I think a rough estimate of the probability of the Amberless Coli coming to be is therefore on the order of [10^-6]^10 (efficiency of an individual hunk o’ chromosome by MAGE) x 32 (number of segments that were MAGE’d) x [16 + 8 + 4 + 2 + 1] x 10^-6 (number of CAGE breeding events) = very, very, very small number. Obviously, this proves Intelligent Design, as do all really small numbers.

Which finally brings me to my point: we still don’t have the goddamn Amberless Coli!!! The most disappointing line in the modern scientific literature is not “Paylines have fallen to 8%” but “Thus far, 28 of 31 conjugations have been completed ….” Arrgghh! I know, I know, it’s coming. And why not get two Science papers out of it? Venter got, what, three, by teasing us through that great miracle of modern biology, the synthetic chromosome. So, I sympathize, I really do, but we all want the Amberless Coli!

Of course, we want more than the Amberless Coli. We want the Coli with the collected works of Shakespeare written in the intergenic regions, and the Coli that can set itself on fire, and whatever other weirdnesses synthetic biology comes up with. We still want to be Craig Venter. We still want to create bacterial genomes at will.

And that brings me to my sad conclusion: I am still not Craig Venter. I’m not even George Church. I’m not even sure that George Church is George Church, in the sense that … really? You’re going to do this all again? Again, back of the envelope: 15 authors (minus George) x (let’s say) 3 years of their lives = 45 FTE years / new genome? While the future is probably more rosy (as Dan Savage likes to say, “It gets better”), still! No, as much as I admire MAGE and CAGE, I think Craig and his army still win. The surety of DNA synthesis at the outset trumps the tiny numbers for MAGE / CAGE. If we are to make new bacteria it will likely be by full synthesis followed by tricks for manipulation and integration or recombination of large (100s o’ kb?) DNA pieces. We do need Gibson assembly, because it works and it’s way more likely to give you what you want in the end.

Given that I have a gift for being an anti-prophet, this almost certainly means that MAGE or its offshoots will work way better than full synthesis. Ha.


- originally posted on Friday, August 5th, 2011