Humboldt State University ® Department of Chemistry

Richard A. Paselk

Chem 431


Fall 2008

Lecture Notes: 26 September

© R. Paselk 2008


3-D Structure of Proteins 5

Symmetry in quaternary structure

Note the common symmetry relationships fopund in protein quaternary structures:

Protein Folding

Primary structure specifies tertiary (& therefore quaternary) structure. This is known from in vitro denaturation/renaturation studies of small proteins.
The classic study involved Ribonuclease: Reduce (break) -S-S- bonds, denature with urea to random coil. Now can renature by gently removing denaturant (urea) and oxidize -S-S- bonds. (text Figure 4-26) with the enzyme activity fully recovered. X-ray diffraction image is also the same! Note - no gremlins, no magic, done in "test tube."

Other small proteins, such as Myoglobin and proinsulin, fold up spontaneously in the same manner as Ribonuclease. However, insulin fails to fold correctly, since a peptide essential to folding has been cleaved off.

Accessory Folding Proteins. The ribonuclease renaturation-type experiment has not been repeated with large proteins, which seem to require the participation of "folding catalysts," to aid their folding: the Chaperones, which we will return to soon.

Thermodynamics and Folding

Let's look at folding in another way: You might guess a protein would fold to lowest free energy conformation. Problem: is there time? ("Levinthal's Paradox", formulated by Cyrus Levinthal in 1968) Stryer calculation (very conservative): Assume 100 aa residue protein with 3 possible conformations/residue; then get 3100or 5 x 1047 possible conformations. If search at a rate of one structure/10-13sec then get (5 x 1047)(10-13)= 5 x 1034 sec or 1.6 x 1027 years to search (and thus to fold protein). This is greater than the age of our Universe (13.7 x 109 yrs). [Rawn calculation (perhaps more realistic): same but assume 10 conformations, then get 1087sec or 3 x 1080 yrs!]

Obviously from these calculations not searching all possible conformations (or we have the process wrong!), so cannot say protein achieves the lowest global free energy, but rather a local free energy minima. (Like a valley in mountain range: a local energy minima, but not lowest [Marianas trench].) [sketch - note represents 'local' minimum, not global minimum)] Which valley the protein reaches will depend on folding "paths." That is folding appears to depend on kinetics as well as thermodynamics.

Some structures fold faster than others: alpha-helix seems fastest folding, due to cooperativity. That is, once a couple of H-bonds of helix are formed subsequent residues are aligned to form H-bonds, thus a cooperative process. Beta-structures are second most rapidly formed. These kinetic factors probably contribute to the commonness of these two structure types. Another factor making them common, is that helices and sheets are the most compact forms polymers can take - the proportion of a flexible chain in helix and sheet conformation increases as the chain is forced to become more compact. H-bonding and phi and psi angles then determine they will be alpha and beta structures. Thus we expect rapid formation of alpha and beta structures. However, unless they are stabilized by interacting with each other and by compaction, they may unfold and try other combinations until stable associations result.

[See D. Baker. (4 May 2000) A surprising simplicity to protein folding. Nature, 405. p 39-42.]


Stages of Protein Folding

Protein folding follows a sequence of stages (text Figures 4-27, 4-28):

Protein folding also appears to be hierarchical, that is it begins with folding of low stability local structures, which interact locally etc. (text Figures 4-28) Note that this leads naturally to a sub-domain type structure and hierarchical structures as seen in proteins.

Note that some aa residues favor one or another secondary structure. Unfortunately, such tendencies by themselves have not proved effective in predicting protein structures. However, using this kind of information in a "local" (nucleation) and "hierarchical" (extensions/higher levels) way can predict some small protein structures reasonably well.

Pathway Diagrams

C431 Home

Lecture Notes

Last modified 26 September 2008