Review article on the likelihood of functionality in amino acid sequence space?

Um, okay. I guess get back to me when this is more than a hypothesis. In the meantime I’m going to go with the evidence, which is that biological functions require a high specificity in cellular environments chock full of potential targets and therefore require a sophisticated 3D structure to perform their function properly.

You see the other references in the paper? You understand the introduction and discussion? Even the gene in question doesn’t seem to have a well defined tertiary structure. This insinuation that there’s no evidence and it’s “just a hypothesis” is false. You’re clearly letting your confirmation bias dictate your thoughts here. Snap out of it.

4 Likes

I don’t think you’ve looked at the evidence. I also note that you’ve hauled the goalposts far away from your previous claims about stability.

You’re now couching your position in easily fudgeable subjectivity (“sophisticated” has zero empirical meaning) to provide plenty of wiggle room.

As for specificity, how many receptors are there that feed into how many second-messenger pathways, and why do so many drugs have so many side-effects?

Then it appears he couldn’t see the protein for the folds, to abuse an old saw.

What matters is function. If there is function without folds then folds don’t matter. More to the point, Axe’s study wasn’t able to measure how many different folds would produce a specific function. As the abzyme studies demonstrate, finding enzyme function in randomly assembled antibodies is a lot easier than Axe’s conclusions would seem to indicate.

2 Likes

Finding a stable/functional protein will be very rare by chance, but life took 1+ billion yrs to get started, and rare events become likely over a long time. In grad school, I did a literature review on the relationship between protein sequence/structure. It is awe-inspiring to learn how changing one amino acid at a time can transition between two different 3-D structures without losing either structure or its function (1) and how related different protein sequences and structures are (2). While this does NOT prove “macro” evolution or how every protein structure evolved, once you being to comprehend how much is possible through simple single mutations, you can then begin to imagine what millions of years of evolution and rare DNA rearrangements can add up to. We can’t replay 1 Myrs of evolution in the lab, but what we have seen is remarkable.

  1. From the Cover: A minimal sequence code for switching protein structure and function
  2. A galaxy of folds
5 Likes

2 posts were split to a new topic: Welcome Brandon to PS!

There are many designed signaling cascades that have related and intended functions. Drugs are from a different designer and they are small molecules, not proteins, and much less specified.

I would be interested in looking at examples of this, other than the antifreeze proteins which are small enough and perform a very unspecific function. We can all agree those probably evolved.

Why couldn’t evolution produce new protein folds? There are many examples of new enzyme functions arising through mutations, such as the evolved beta-galactosidase enzyme that Barry Hall studied for decades:

That’s nonsensical. Why not just admit that you don’t know?

Yet another objectively false claim:

Last time I checked, monoclonal antibodies were proteins. Do you disagree?

The mAb drug cited above is pretty darn specific for the HER2 receptor, as defined by its Kd of only 5 nM. And it was designed by random genetic variation and selection in the immune system.

Yet it has side effects because of second-messenger pathways that are merely selective. That was the question you didn’t answer, while offering up two more blatant falsehoods.

2 Likes

Enzyme function is not the same thing as a fold. Functions require folds generally, but enzyme modifications in function usually involve change in the active site, not a structural change in the fold. The active site and the enzyme works exactly the same way, just with an affinity for a slightly different substrate, often promiscuous and often at much lower binding affinity than the original substrate.

If the substrate affinity is different, then the enzyme is not working in the same exact way.

Also, we have examples of antibodies that have enzymatic activity. These are derived from a random mixture of gene segments, and yet they have function. This indicates to me that enzyme function is relatively easy to find in sequence space.

1 Like

Interpreted charitably, my lack of certainty is implict in the fact that I’m speculating about alternative possibilities.

Yes, actually. Proteins are a bit more than a polypeptide. Antibodies are only partially randomized, as you know. Their basic structure, maybe you would call it a fold (I wouldn’t), is designed and a common feature of every antibody. They are designed for binding. In that way they are very similar to promiscuous enzymes.

So are antibodies, they are proteins. This isn’t something with which you disagree without making a fool of yourself.

1 Like

You weren’t. The question is, why are there so many receptors that feed into so few second-messenger pathways?

Your reply made no sense, but it was completely certain:

You also made a spectacularly false claim:

To which I pointed out a drug called Trastuzumab, which is highly specified and a protein (an antibody), completely falsifying both parts of your claim.

I asked:

For you to reply with:

was nonsensical, as you were claiming that drugs could not be proteins. Immunoglobulins are proteins.

No, I don’t know. Perhaps you should explain that.

When doing so, please explain why your immune system has to perform constant surveillance to kill off self-reactive clones, a process that causes autoimmune diseases when it fails.

This is a remarkably false claim.

You don’t seem to realize that contrary to DI misinformation, folds are structural classifications.
https://proteinstructures.com/Structure/Structure/protein-fold.html

Antibodies (aka immunoglobulins) are repeats of one of the most common folds, oddly enough called the immunoglobulin fold or domain:
http://scop.mrc-lmb.cam.ac.uk/scop/data/scop.b.c.b.b.html

If you’re interested in learning how wrong you are, scroll down through some of the families and you’ll see that many, many proteins with different functions share this fold.

So, yeah, you’re pretty much unable to communicate about protein structure if you deny that the basic structure of an antibody doesn’t represent a fold.

No, Ben, the binding subdomains (V regions) are constantly being randomized by recombination and somatic mutation. As new ones arise, they are subjected to selection against those that bind self and selection for those that bind non-self.

If antibodies have a specific fold, than that part of it is not being randomized. Agree?

We would say the fold is designed and so is the mechnanism facilitating randomization.

You can nitpick all you want. The argument Axe makes is based on folds being rare in the sequence space, not enzymatic or antibody binding sites being rare. So when you and others constantly present us with examples that aren’t novel folds forming, don’t expect us to be impressed.

Random amino acid sequences can form folds, so I don’t see how the two are related. Folds are the result of thermodynamic interactions between the peptide and its environment, and folds occur spontaneously through natural laws.

Then Axe’s argument is irrelevant to evolution. If it is relatively easy to evolve an enzyme, then it doesn’t really matter how hard it is to evolve a specific fold for one specific enzyme that has that activity. At that point, it is just the Sharpshooter Fallacy.

3 Likes

Not at all. The constant regions are not randomized. The variable regions are. Both types of region are canonical Ig folds. You don’t seem to understand the meaning of the term.

You’re still not grasping the meaning of “fold.” Antibodies are composed of 10 Ig folds, not a single one.

Pointing out that you don’t know the definition of “fold” or that you don’t know that lots of receptors feed into a much smaller number of second-messenger pathways isn’t nitpicking.

I think you may have set a record for number of false claims per unit time.

Which is gibberish. You’re also leaving out function, which he includes.

Virtually all proteins fold, so they aren’t rare.

There aren’t that many folds, which are structural classifications. Evolution doesn’t select for new folds. It’s about function.

The catalytic antibodies that have beta-lactamase activity have structural similarity to the active site of the bacterial one, but it’s a completely different fold–the Ig fold.

1 Like

Couldn’t we just say that God creates through evolution, then?

1 Like

Yes. Some do say that. Nice of you to finally notice.

1 Like