Our neurons (and us) would be the size of amoebas if weren’t for tubulin which forms the superhighways (microtubules) along which cargo is shipped to the end of axons. Your average NBA player has axons over 3 feet long going from his sacral spinal cord to his calf muscles. Split the difference and call it a meter. Diffusion is way too slow to get anything that far. The trucks schlepping things back and forth on the microtubular highway are called Kinesin and dynein. I think in terms of nanoMeters (10^-9 meters). Each tubulin dimer is 80 nanoMeters long, and K & D essentially jump from one to the other in 80 nanoMeter steps.
How many jumps do Kinesin and Dynein have to make to go a meter? Just 10^9/80 — call it 10,000,000. Kinesin and Dynein also have to jump from one microtubule to another, as the longest microtubule in our division is at most 100 microns (.1 milliMeter). So even in the best of cases they have to make at least 10,000 transfers between microtubules. It’s a miracle they get the job done at all.
To put this in perspective, consider a tractor trailer (not a truck — the part with the motor is the tractor, and the part pulled is the trailer — the distinction can be important, just like the difference between rifle and gun as anyone who’s been through basic training knows quite well). Say the trailer is 48 feet long, and let that be comparable to the 80 nanoMeters Kinesin and Dynein have to jump. That’s 10,000,000 jumps of 48 feet or 90,909 miles. It’s amazing they get the job done.
Now that you’re sufficiently impressed with tubulin’s importance, it’s time to see why it needs help. First a bit of history. Christian Anfinsen was a Swarthmore football player who happened to win the Nobel prize 50 years ago for his work on the protein ribonuclease, an enzyme. If you heat it, enzymatic activity is lost (the protein is said to be denatured). This is because the exact 3 dimensional path of the protein backbone forming the catalytic site of ribonuclease was lost. However if you leave the denatured protein alone (under the proper conditions) it folds back up to the correct 3 dimensional shape. His point was that the amino acid sequence of the protein was all that was needed to determine ‘the’ three dimensional shape of the protein. This was at a time when we didn’t know that most proteins have a variety of shapes not just one.
Unfortunately tubulin does not fold up to the shape found in microtubules. It needs significant help from two friends, prefoldin and TRiC. TRiC is a monster conglomerate of 2 copies each of 8 different proteins with a molecular mass over 1,000,000 Daltons (e.g. a megaDalton). What is one Dalton — it’s the mass of a hydrogen atom. TRiC is made of two back to back rings (with built in lids) each ring consisting of 8 different but related proteins). Each of the proteins has a domain which binds ATP and a domain which binds the protein to be folded. There is a central cavity 90 x 90 x 50 Angstroms in size. Since each hydrogen atom is about 1 Angstrom in diameter, it can fit 405,000 hydrogen atoms inside, or about 200,000 carbons, hydrogens, oxygens and nitrogens — enough room for most proteins.
Prefoldin is equally amazing. It basically looks like a Portuguese man o’ war — https://en.wikipedia.org/wiki/Portuguese_man_o%27_war. It is made of 2 copes of one protein and 4 of another. The tentacles are long alpha helices projecting down from the body.
The tentacles interact with tubulin, carrying it in an unstructured form, thrusting one of its tentacles into the central chamber of TRiC carrying unstructured tubulin with it. ATP addition leads to lid closure and tubulin encapsulation in the chamber.
A magnificent paper [ Cell vol. 185 pp. 4770 – 4787 ’22 ] describes what happens to tubulin in the TRiC chamber at near atomic resolution. They are literally watching tubulin fold as it passes from one of the 8 different proteins making up the TRiC ring to another. The disordered carboxy terminal chains of TRiC are postulated to function as a tethered solvent allowing the intially disordered amino acid sequence of tubulin, to slither into their correct positions more easily.
I’m sure it’s behind a paywall, but if you can look at the figures in the paper, you’ll be bound to be impressed.
So Anfinsen turned out to be wrong, and some 10% of newly translated proteins turn our to need TRiC’s help. And yet he wasn’t, because AlphaFold uses only the amino acid sequence of proteins to predict their three dimensional structure.
One further point. The ancestral bacterial protein for tubulin is called FtsZ. It happily folds to the correct structure by itself. However tubulin developed new domains, some of which are for the motor proteins Dynein and Kinesin, and others are for microtubule associated proteins such as tau, the major component of the neurofibrillary tangle of Alzheimer’s disease. These domains are on the surface of the protein, making it harder to fold by itself.
All this information would have been impossible to get 10 years ago, and it’s all due to the sharpening of our technological tools.
Chemiotics II 