Re: Not quite as impressive as advertised.
I've looked at the paper, and I think this post warrants a point-by-point rebuttal:
> The DNA of that bacterium consists of a couple of million "base pairs",
E. coli genome size : about 4.6 million basepairs
> what they've done is replace ONE base pair with a synthetic pair which is sufficiently similar to the real deal that it doesn't break DNA replication.
True
> Even though only one base pair was changed, the protein the gene coded for was broken by the insertion (a so-called reading frame error*),
False. It was a base *replacement*, and *not* in any protein-coding sequence. Where did you get that from?
> which is why the bacterium grew more slowly
False. Because a) the above, and b) because the unnatural bases and plasmid didn't make it grow more slowly. Expression of the protein required for transport of the unnatural bases into the cell did, but did so in the absence of these bases. Adding the bases caused no significant further reduction in growth rate.
> (and presumably why they didn't let it replicate more than 15 generations - it was a death spiral).
False. They report the plasmid replicating for approx 24 (plasmid) generations (over 15 hours of growth). They analysed reversions of the modified base position at that point; this was below their limits of detection. If they didn't supply more of the unnatural bases (which degrade over time in the culture) then over the following 6 days of growth, the plasmid would either be lost from the cell or acquire a reverting mutation. This is in no sense a "death spiral" - while the necessary materials are supplied, the modified base is maintained pretty well.