BioPHYSICAL chemistry

The prediction of the folding pathway of a
protein of a specified sequence remains a chal-
lenge (Onuchic & Wolynes 2004; Das et al.
2005; Hubner et al. 2005; Liwo et al. 2005; Cho
et al. 2006). The folding process is complex
and different models have been proposed. The
folding can be hierarchical, with local structural
elements, such as αhelices, folding first followed
by longer-range interactions driving the folding
of helices together and eventually large domains
folding into the full structure. Alternatively, the
initial folding may be sequence-specific, with
hydrophobic amino acid residues collapsing into
a glassy state mediated by hydrophobic inter-
actions with hydrophilic residues on the exterior.
Thermodynamically, each of these conforma-
tions has a certain energy and the distribution
can be plotted (Figure 8.4). The unfolded states
are at high energy and have a large degree of
conformational entropy. As folding proceeds the
energy decreases and the distribution of states
is funneled into an ensemble of folding inter-
mediates until the single lowest-energy state is
achieved. In the course of the folding, the pro-
tein must have enough energy available to overcome barriers and to con-
tinuously sample states, even if a local minimum is reached.
In principle, for a given sequence, if all of the interactions involving
amino acid residues were properly modeled, it should be possible to
predict the folding process and hence the three-dimensional structure.
Effectively, the energy landscape shown in Figure 8.4 could be mapped and
the lowest-energy state predicted. The availability of very fast computers
has led to the development of different programs that are increasingly
becoming more effective in their predictions. The plethora of sequences
provides a testing ground for predicted structures and the possibility of
refining critical interactions for folding based upon sequence comparisons.
Furthermore, the experimental feasibility to generate peptides, whether
chemically or through expression systems, provides the opportunity to test
the effects of sequence changes on protein folding.

Prions

One of the intriguing questions in biology is how the conversion of pro-
teins with intricate folds, appearing to be at a stable energy minima, can
alter their shape and form long aggregates. These transformations are not
only of interest in understanding the energy landscapes of proteins but are

CHAPTER 8 STATISTICAL THERMODYNAMICS 169

Discrete folding

intermediates

Native structure

Molten globule

states

Beginning of helix formation and collapse

Energy

Figure 8.4The

energy landscape

for protein folding

showing the

energies of a

protein in different

configurations with

the fully folded state

having the lowest

energy.