Comment by ncmncm | Hacker News Reader

Comment by ncmncm | Hacker News Reader

Two things all articles like this fail to mention that seem essential:

1. Oceans were absolutely saturated with reactive metal complexes, particularly including iron. A billion years later oxygen poisoning started, and it all oxidized and precipitated out, but life was established by then and (some) survived the catastrophe. So, the chemical environment was very different than is easy to reproduce today, and full of stuff that would have catalyzed all kinds of reactions that go very slowly without.

2. Oceans were also absolutely lousy with nucleic and amino acids, and random RNA sequences. An RNA that copies other RNAs it bumps into was nothing special, but when two of them bumped, suddenly in a (geological) eye-blink the entire ocean was teeming with their highly-optimized descendants, along with any other RNAs that proved helpful to have around. Natural selection was already hard at work, no cells or genes needed.

So that is the environment that life arose in, utterly unlike anything found anywhere today. Any random drop of water that got caught up in a membrane would be full of those RNAs that had proven most useful at copying, and also all those random catalytic metal complexes.

I gather that nowhere in the natural world do we find any cellular process that creates a membrane from scratch; all everywhere grow from bits pinched off from membranes already on hand. Arguably, what is really essential to species is being made of bits extended from the first membrane that wrapped enough replication machinery to do the whole process inside. It would need to have been better at allowing loose nucleic acids in than letting assembled RNAs out. Life is bags of water and stuff that pack the bag with catalyzed reaction products and can extend the bag as it fills up. When a lobe gets pinched off with enough stuff inside, off it goes.

Later there would be genes, and metabolism.

(1) is correct, but I believe (2) is no longer assumed to be true (especially the part about random RNA in the ocean).

The major problem is concentration. Even if you invoke geological timescales and a favourable redox environment, concentrations of reactants would be rapidly diluted in the ocean. Since there are quite a few dehydration reactions in the synthesis of oligonucleotides (like RNA) and proteins, when diluted in the ocean, the plentiful water drives the equilibrium back towards the starting materials. Without some mechanism to accumulate the reactants to form the first R/DNA, it remains an open question how the process got started.

There used to be a homeostasis-first theory—to compete with the metabolism-first and oligonucleotide-first theories—but it hasn't gained a lot of traction as an independent theory.

Dehydration is a fact hard to argue with.

Yet, all it needs is a weak simple catalyst, in a literal ocean full or random stuff that we can't actually imagine today.

> all it needs is a weak simple catalyst,

Catalysis lowers the activation barrier, sure, but you still have to contend with the thermodynamic equilibrium in the absence of the catalyst, in addition to subsequent degradation by light, etc.

> ocean full or random stuff that we can't actually imagine today

We probably can't ever really know what was in the Earth's ocean before life, but quite a bit of research has been put into figuring out some plausible conditions and then testing the hypothesised pathways put forward by the leading theories under said conditions.

All this is to say, yes, clearly abiogenesis occurred, but there are major gaps in the leading theories that are interesting and testable. I don't think it's very useful to gloss over the details, even if in the end the best we can hope for is a plausible pathway given a specific set of assumed conditions.

Is catalysis always symmetric? Or do some catalysts only work in one direction?

I imagine one that splits a molecule could work only or mostly in the one direction, if to go the other way would require both constituent parts to come together at the same time. But that is an example favoring entropy.

What would one that favors assembling molecules look like? A ribosome seems like the extreme example; those don't disassemble proteins. But is a much simpler example very unlikely?

The difference in entropy is so small that we are talking about entropy of dissolution. It's not hard on those conditions to introduce some element that weakly bias the reactions on a different way.

I don't want to minimize the gaps on our knowledge. We don't know how any macromolecule polymerization occurred. But I also don't want to maximize the gap. Those kinds of reaction are completely mundane and occur on a huge variety of environments. The fact that we don't know what environment it was doesn't mean it's an outwordly phenomenon. And adding just a few of those mundane reactions is enough for life to appear.

(Anyway, I question your certainty about the concentration of water on our primordial oceans. Abiogenesis research usually uses a model of water origins that gives a precise estimation for its amounts, but not only we know that this model is wrong - water is cycled by geologic means like any other mineral - but we also do not have anything with near that amount of precision to the other substances that composed it.)

Articles like this tend to be heavily constrained by space, and have a target audience that's probably a bit more lay than what you would prefer.

I can recommend the author's book 'The Vital Question', which dives into some hypotheses addressing precisely these considerations in much more detail.

Reply on news.ycombinator.com