Progress after the Royal Society conference?

[Rumraket (Mikkel R.)]

@lee_merrill

@Rumraket

@lee_merrill

Well, you trimmed my response, though I subsequently admitted that I misspoke. I meant that knock-out experiments are fine, not that selecting mutations to look for ahead of time is fine

Which is what you are doing when you search for two mutations which in combination renders a gene nonfunctional and where no other mutations can compensate, deliberately insert those two mutations, and then wait for those two and only those two to reverse.

But that is not my view, I would say the specific mutations would be anticipated to be the way function would be restored, but also any other mutations which restore function would be fine, too.

Who cares what you think is fine? What matters is what was done in the actual experiment, not whether you are personally okay with some alternative possibility that wasn’t explicitly tested. LOL.

@lee_merrill

@Rumraket

Not a particularly interesting or informative experiment with respect to evolution in general. The only sort of scenario for which such an experiment can be informative would be one where you know that some adaptation could only have evolved by such a pathway, and where no other alternatives existed at the time.

But I think allowing other mutations is fine too, you will get an idea how likely it is to restore function, regardless of how this happens.

Cool, but then there are lots of examples of knock-out experiments where beneficial genes are knocked out, the organisms then evolve to compensate for the loss. Typically these involve completely deleting the gene encoding some enzyme specialized for a particular substrate conversion, and what happens is the population adapts by mutations causing upregulation of another enzyme that has some promiscuous activity on that same reaction. Here’s a recent example, open access:

Adaptive laboratory evolution recruits the promiscuity of succinate...

In this paper we see how there are multiple routes by which the bacteria can compensate for the loss of two key enzymes (the strain is engineered both Δepd and ΔgapA). Generally, almost all involve ways of increasing the output of the sad gene. Either by gene duplications (more gene copies expressed = more enzymes), straightforward upregulation of expression, or mutations that increase the promiscuous activity, or some combination of two or all three. In fact the sheer number of ways there was for mutations to cause these compensatory changes was rather surprising, even including reductions in enzyme activity for enzymes active in other pathways (interesting how this biochemically led to increased sad expression btw.)