Earth is basically a bacteria planet, despite humankind’s naked-eye-level profile. They’ve been here unfathomably longer than we have, they live in plenty of places where we can’t survive, and their biomass far outranks ours. This paper will show you just how adaptable the little creatures are. Wild-type E. coli (like many other bacterial species) can synthesize all its essential metabolites and cofactors from scratch (unlike us, for example, having to get several vitamins from our diet). These pathways have been worked out in some detail. For example, it’s known that aspartate 1-carboxylase (known as PanD) is an essential enzyme for the synthesis of beta-alanine. In turn, beta-alanine is essential for the synthesis of pantothenate, which is essential for the synthesis of Coenzyme A, which is essential for life itself. So what happens if you delete the PanD enzyme? All the bacteria just plotz, right?
Well, disrupting a direct pathway tied to an enzyme essential for all life on Earth is a challenge. You can generate that PanD mutant just fine, and they’ll grow happily as long as the culture medium contains beta-alanine to keep them going. The authors gradually diluted that out, though, and put the bacteria under strain to see what would happen. I have to admit that the colony of wild-type bacteria used in the study weren’t up to the challenge, and did not mutate out of the problem. But giving them a higher propensity to throw off mutations by compromising DNA proofreading gave some results. That’s the sort of mutation that normally could be troublesome for any given bacterial population, but comes in quite handy under unusual conditions, and in fact many types of bacteria are known to increase their mutation rate under stress.
Speeding things up in the manner gave a number of varieties that worked out a way around the lack of what had been essential beta-alanine supplies. One pathway that emerged did so by degradation of uracil. Interestingly, this didn’t involve changes to any enzyme structure or function, but rather in expression levels of existing ones. The mutations showed up in various repressor proteins, and allowed more uracil to be synthesized and more of it to spill over into a pathway that could generate beta-alanine, through an intermediate (malonyl semialdehyde) that’s normally considered toxic, but in this case got immediate converted to the desired product. This uracil pathway turned out to be a popular one. The group took the 11 strains that had evolved a way out of the beta-alanine stress and deleted a key enzyme in this pathway (RutABC), and that wiped out all but three of them – so in whole or in part, chewing up uracil was an attractive solution.
What if you delete both PanD and RutABC? That forces the bacteria to tap-dance even faster, but they’re up to the challenge. Under those conditions, the survivors turned out to all have the same mutation as evolved in the three leftover strains in the experiment above: they all changed G655 in ornithine carboxylase (SpeC) to either serine or alanine. This allowed the enzyme a wider scope for a decarboxylation/deamination reaction (while actually limiting another decarboxylation pathway, so there’s a tradeoff for survival). OK, what about a triple mutation? Take out PanD, RutABC, and SpeC all at the same time? The bacteria even found a way out of that corner, Houdini-style. This pathway isn’t completely worked out, but mutations appeared in arginine decarboxylase (SpeA), S-adenosylmethionine decarboxylase (SpeD), and spermidine synthase (SpeE).
These sorts of studies have been done with other nutrients and enzyme pathways, but this is the most detailed look at the adaptations that I’ve seen. It’s clear that there’s a lot of biochemistry that can be repurposed under this kind of severe pressure (betting against bacteria is like going in against a Sicilian when death is on the line). It’s important, though, to remember the problem of teleology here: it’s easy to talk about things in terms of their purpose (and that’s what I’ve been doing), but that makes it sound like that bacteria are planning their moves. They’re not, of course.
These are error-prone organisms (that was a human purpose, as described above). So there are uncounted mutated strains present in the culture; every single cell in the vial is probably a different beast. As you start to decrease the amount of beta-alanine available, the story isn’t about a wily bacterium coming up with ways to counter the problem – it’s about everything else in the culture dying because they can’t make it under the new conditions. Whoever is left gets to divide again, spinning off still more mutations, some of which will turn out to be able to deal with the next round of shrinking beta-alanine levels, and most of which won’t. What you get at the end is a particular pile of accidents that managed to hang together, survive, and reproduce. I’m one of those, too, and so are you.