Louis Hom wrote:
> Let's say you have a polypeptide that adopts a certain fold. And let's
> make the wild assumption that the order of the amino acids doesn't matter
> (i.e., only the composition of the chain is important, not the positions).
>> The limiting cases, then, would be 1) the amino acids (hydrophobic, polar,
> charged) are distributed evenly/randomly along the chain; and 2) all of
> the hydrophobics are grouped together, all the polar residues are grouped
> together, all of the charged are grouped together.
>> My question is: would you expect any interesting difference to exist
> between unfolded case 1 vs. unfoled case 2?
The problem is the definition of "unfolded state". If you define it as
completely random coil, this implies no interaction between
neighboring side chains, and then the distribution really doesn't
matter. But if you define it as "the state adopted in 8 M guanidinium
chloride (or whatever denaturing conditions you use)" it does matter.
Then there will be different amounts of residual structure left in
that "apparently unfolded" state, and this amount of course depends on
the distribution (and the sequence).
When one talks about stabilities, one should always be aware of the
possibility of residual structure in the unfolded state. But often it
is neglected, and mostly this is reasonable. It may not, however, be
neglected in two or three cases:
1. one has evidence that the unfolded state has really a lot of
residual structure
1a. one discusses the rates of the first, very fast processes in
folding: what is the starting point?
2. comparison of stabilities or folding kinetics of mutants with
different glycine content: because glycine has no side chain, its
entropy in the unfolded state is quite different from other amino
acids. But to be honest, this is mainly not a point of residual
structure, but, as I said, of entropy.
Frank