After having looked at a lot of folds in lectures 5, we will now go back to physics and ask why there are so many - or so few - folds. What determines a "fold" in the first place, and why does nature seem to reuse folds for quite diverse functions of unrelated proteins?

We will particularly look at a few typical folds, and discuss the three main reasons why proteins might share folds

Evolutionary divergence
Functional convergence
A limited number of physically favorable folds

We will look at how nature seems to reuse components of folds to create larger structural complexes in particular for eukaryotes.

However, now we can also study folds, layers, and the complexity with tools such as entropy and free energy. Superficially this looks trivial, but there is one major concern: All the statistical mechanics you learned was under the assumption that you had an equilibrium between states, and a protein most definitely does not have an equilibrium between different sequences. We study this "multitude principle" quite a but, and you should also read about it in the book.

You might also remember that we showed that helix and sheet structures need a certain minimum size to be stable, but we asked what would happen for very long structures. In this lecture we address this part, and particularly look at the probability of a sequence of beads/cards/amino acids (anything you prefer) occurring without a break - it turns out to be pure probability that limits the length of normal sheets and helices!

The multitude principle also gives us some tools to talk about the stabilization energy of proteins, and why defects are so rare (we'll be coming back to this, so do read up on it).

Armed with this, we can now start to study what happens when a protein folds. One of the most fundamental questions is whether an entire protein folds at once (all-or-nothing), or if it's gradual? It turns out we can answer this with some fairly simple calorimetry experiments - and a bit of math.

Reading suggestions

Protein physics book, chapters 15, 16 & first part of 17.

Lecture recording

Playlist of all lecture 9 concepts

Isolated concepts:

Most sequences end up in a small fraction of folds
Why are proteins similar - evolution or physics?
Classification of protein folds
Fold databases - CATH
Fold databases - SCOP
Evolution of protein structures - hemoglobin
Bacterial vs. eukaryotic/vertebrate proteins
Evolution can happen fast - tomcod PCB resistance
What features make for stable folds?
Most folds consist of 2-4 layers
Patterns we see occur because the are stable and avoid defects
Even small defects can destabilize an entire fold
The multitude principle - Common folds can accept many sequences
Sizes of helices/sheets are limited due to sequence pattern probabilities
Recap - secondary structure elements can't be too small or large
Small defects are bad because much fewer sequences will fit the fold
Stability distribution comes from interaction randomness, not Boltzmann
Most random sequences will not fold into proteins
It is rarer with helices than sheets inside proteins
Green Fluorescent Protein (GFP)
Conformational transitions - induced & selected fit
The allosteric transition of hemoglobin is critical for O₂ transport
CD spectroscopy is useful to study unfolding
Folding of small domains is an all-or-none transition
Proteins will unfold at either high or low temperature
Study questions

Slides with still images from videos

(Some students find these useful to take notes)

Full-format slides

Handout-format slides

Study questions

What is the main reason many proteins are similar?
What is structural evolution? Give some examples.
What is the relation between structural & sequence evolution?
What is the difference between the CATH and SCOP databases? What can we use them for?
Why do protein domains usually consist of 2-4 layers? Why not 1, or very large?
What do we mean by sequence-fold fitting?
If a larger fold can accommodate more sequences, shouldn’t most domains be large?
What is the typical stabilization (free) energy of a protein?
Why is a (small) defect of only a few kcal so bad for a protein fold (in terms of sequences fitting it)?
Why is this not governed by the Boltzmann distribution?
What is the typical size of helices and sheets?
Why is it rare to see much smaller helices/sheets?
Why is it rare to see much larger helices/sheets?
What is GFP? How is it useful?
What is the difference between induced fit and selected fit in conformational transitions?
What is allosteric modulation?
What method would you pick if you wanted to study secondary structure fractions in a protein?

Scientific articles

Gene regulation of switching fetal to adult hemoglobin

Structural evolution in cod to resist PCBs

How many sequences can fold into different structures?

Monod, Wyman, Changeux classical paper on allostery