With the overarching goal of improving our understanding of the kinetic mechanisms by which biomolecules fold, I have here investigated 1) an unusual, and 2) an unprecedented folding reaction, the first being the anomalously accelerated folding of a small, single-domain protein, and the second being the two-state folding of a small, single-domain DNA.

The folding rates of simple, single-domain proteins are relatively insensitive to the finest details of sequence. This said, the R48F substation in chymotrypsin inhibitor 2 (CI2) is the single most accelerating substitution seen in a set of over 700 mutations in 24 single-domain proteins for which kinetic data are available. Specifically, this mutation accelerates the folding of CI2 40-fold, an acceleration that dwarfs the less than 4-fold average change in rate produced by the point substitutions in my data set. To understand the origins of this anomalous acceleration I synthesized and characterized 10 additional substitutions at this same site, and correlated charge and hydrophobicity of the substitutions to folding rate. In doing so I find that the observed acceleration arises due to two effects that independently stabilize the folding transition state of this protein: alleviation of repulsive charge-charge interactions arising due to the wild-type arginine's positively charged guanidinyl group, and the increased hydrophobicity associated with its replacement with side chains more hydrophobic than the arginine's three methylene units. Of note, a possible reason for preservation of this biophysically unfavorable wild type arginine is a functional constraint; that is, faster folding substitutions such as isoleucine seem to pay a cost in enzymatic activity in the native role of protease inhibition.

In parallel studies I explored more generally the characteristics governing folding of sequence-specific heteropolymers. To do so, I characterized the two-state folding kinetics of a small DNA aptamer and, in an effort to determine which aspects of two-state heteropolymer folding are "universal," compared these to the folding of a collection of similarly simple, two-state proteins. Specifically, I investigated three kinetic properties commonly measured in proteins for this novel two-state aptamer---folding rate, beta Tanford (betaT), and ϕ values---and found substantial similarities and differences between the aptamer and a typical protein of the same size and complexity. First, in a model that predicts the folding rate of proteins based solely on topology, the observed folding rate of the aptamer appears to obey this same relationship.

Second, the urea-derived beta T, which is thought to reflect the fraction of hydrophobic surface area buried in the transition state, falls slightly on the low end of the range seen for a set of 38 similarly small, single domain proteins. The salt-derived betaI, which is thought to reflect the fraction of ionic interactions formed in the native state, is, in contrast, significantly higher, suggesting that the fraction of native ionic interactions formed in the transition state is higher than hydrophobic ones. Finally, of three sites for which ϕ---a residue-specific analog of betaT, indicating fraction of native-like character around an amino acid, or in this case, nucleobase, in the transition state---could be calculated, two are >0.75 while the third, despite a large error, falls unambiguously above 0.5. That three out of three ϕ values are above 0.5 suggests perhaps a more native-like transition state than is typical of proteins, which average ∼0.3.

It thus appears that 1) the large role chain entropy plays in determining folding rates of proteins may apply to the folding of cooperative heteropolymers more generally, 2) ionic interactions may be slightly better consolidated in the folding transition states of oligonucleotides, which are highly charged, than are hydrophobic interactions, and 3) the transition state for two-state oligonucleotides may be more native-like than is typical for similarly simple, two-state proteins.