“Going with Richard Dawkins’s observation on how evolution works, ie, that even a very poor eye is better than no eye at all, we’ll have created an internal system in our petri dish that says, ‘Right, you, not-very-well-functioning fluorescing cell, you’ve done better than the others, so you get to survive to the next generation,’” he says.
Those successful cells, he says, will replicate lots of copies of their “good-at-the-job” DNA, which will then be taken up by other cells and used to piece together slightly better solutions next time round.
“What we hope to see is the fitness of the population improving as the cells get better at solving the problem of how to do the job,” he says.
Apart from bio-remediation of drinking water, what could these clever bacteria be used for?
“This is really speculative,” says Amos, reluctantly, aware of the dangers of promising too much, “but there’s the possibility, way in the future, of taking cells from a patient and re-engineering the cell so it can detect a problem and construct its own chemical response. Once reinserted, that means that the cell — which wouldn’t be rejected, as it’s the patient’s own — could act at the site of the problem. You wouldn’t need to blast someone with antibiotics, for instance — it’s a much more exquisite solution.”
What does he anticipate the scientists on his project will struggle with?
“The main risk is that the comparator doesn’t work,” he says. “That would mean the cells can’t tell how well they’re doing at the task, so the ones that perform well couldn’t be ‘rewarded.’ And then we’re stuffed.”
And what would success look like — on a trivial note, might we see fluorescing bacterial Christmas tree lights?
“Well, actually, bacterial Christmas tree lights have already been done,” he says with a laugh. “But they were made by researchers manually engineering cells to do the job. Success for us would be to get our cells ‘learning’ how to be Christmas tree lights all by themselves.”



