|
Non-Native Hydrophobic Interactions in a Hidden Folding Intermediate
However, the detailed process of protein folding remains elusive. Two recent events have made the study of protein folding one of the most important issues in current biology: 1) More than 20 human diseases, termed amyloid diseases, have been found to be related to protein misfolding and precipitation in cells, including Alzheimer’s, type II diabetes, and Creutzfeldt-Jakob disease, a human version of the mad cow disease. In several cases, it was shown that partially unfolded intermediates are the major precursors of the amyloid molecule. These precursors are thought to be responsible for the amyloid diseases. 2) The genome project has accumulated a huge number of protein sequences. However, to understand the functional information encoded in these sequences, the structures of proteins are needed. Thus, it becomes highly desirable to understand the relationship between protein sequences and structures so that the structures of proteins can be predicted from their sequences by computational methods. Toward this goal, a structure genomics program has been initiated at the NIH. Physical studies of protein folding are essential because all computer programs for predicting protein structures rely on the force-field parameters derived from the physical studies of protein folding. Significant effort has been made in characterizing the structures of partially unfolded intermediates. To date, they have been mainly characterized by using amide hydrogen exchange and mutation studies. The amide hydrogen exchange method can measure the hydrogen bond formation in secondary structures through studying the rates of exchange between amide protons and protons in water. Mutation studies can produce information on side chain interactions by measuring the energetic effect of a mutation on the folding intermediate relative to the native state. These studies, however, can only yield low-resolution structures, and one often assumes that the intermediates have native-like structures in the folded regions. We identified a partially unfolded intermediate of a four-helix
bundle protein, Rd-apocytochrome b562, a redesigned
apoform of an electron transfer protein. The N-terminal helix of
this intermediate is unfolded, as determined by the hydrogen exchange
method. Given this low-resolution structure, a mutant was designed
to substitute the hydrophobic core residues in the N-terminal helix
with glycines. These substitutions selectively destabilize the native
state, making the intermediate the most stable state without affecting
the folded regions.
We have now determined the high-resolution structure of the intermediate by using multi-dimensional nuclear magnetic resonance (NMR). Although the three folded helices have native-like topology, surprisingly, there are significant non-native hydrophobic interactions among side chains (Figure 1). Nevertheless, we found the non-native intermediate buries approximately 258 ± 80 Å2 more of the solvent-accessible hydrophobic surface than does the putative native-like intermediate with the N-terminal helix unfolded. Our findings provide the first structural evidence in support of the hypothesis that evolution has minimized the exposure of hydrophobic surfaces in folding intermediates by forming non-native hydrophobic interactions to avoid aggregation. The findings also have profound implications for theoretical studies of the folding mechanism of proteins since force-field potentials that allow only native-like interactions have been widely used. The non-native features of the folding intermediate of Rd-apocytochrome
b562 also have significant implications for understanding
the general mechanism of protein folding (Feng H et al. Proc
Natl Acad Sci U S A 102: 5026–31, 2005). An important question
has been why many folding intermediates populate before rate-limiting
transition states. The current molten-globule model suggests that
early folding intermediates exist because formation of secondary
structures is intrinsically fast while tertiary packing is slow
(molten globule model). The Rd-apocytochrome b562
intermediate now shows that the repair of the mis-packed
residues cannot be rate limiting because the intermediate was previously
shown to form after the rate-limiting step. Instead, we propose
that population of early folding intermediates is likely due to
the existence of multiple domains that have different folding kinetics.
We are currently testing this hypothesis experimentally. Yawen Bai, PhD |
rotein
folding is the final step in the transfer of genetic information
from DNA to proteins. It is only after proteins appropriately fold
that they have the capability to perform their biological functions.
In the 1960s, Anfinsen and colleagues at the NIH demonstrated that
proteins could fold spontaneously from the unfolded state to the
native state under physiological conditions. It was hypothesized
that the native structure is the most stable state. This hypothesis
provided a theoretical basis for predicting protein structures from
their amino acid sequences and Anfinsen was awarded the Nobel Prize
in 1972.