Book of Abstracts: Albany 2011
June 14-18 2011
©Adenine Press (2010)
Single-molecule Fluorescence Maps the Folding Landscape of a Large Protein
A substantial body of information, both theoretical and experimental, has accumulated over the last years on protein folding, including a provocative recent idea that it is driven by amino acid stoichiometry (1), as opposed to preferred interactions (e.g., hydrophobic interactions) between specific amino acids (2, 3). Particularly advanced is our understanding of the principles that govern the folding of small, single-domain proteins (4). But can we apply the same principles in order to decipher the folding mechanisms of larger, multi-domain proteins? More than 70% of the eukaryotic proteins belong to this group, yet rather little is known about their folding reactions (5).
We propose here that high-throughput single-molecule fluorescence spectroscopy, combined with statistical analysis, can be used to study folding dynamics of large proteins (6). As a proof-of-concept, we studied the folding landscape of adenylate kinase (AK), a 214-residue protein from Escherichia coli. AK molecules were double-labeled for FRET at two sites in their CORE domain. The molecules were individually trapped within surface-tethered lipid vesicles (7, 8) in the presence of different concentrations of guanidinium chloride (GdmCl), a chemical denaturant. The trapped molecules were examined using an automated single-molecule fluorescence microscope, which allowed us to obtain data sets consisting of many thousands of short FRET trajectories of individual molecules. The availability of this massive amount of data enabled us to construct a detailed map of the folding landscape of AK, using hidden Markov modeling (HMM) (9). We found that the folding dynamics of AK could be described in terms of transitions between six quasi-stable states, one of which is probably misfolded. The folding reaction involves many parallel pathways connecting these states in jumps of various sizes (See Figure). This folding pattern differs markedly from the well-known two-state model of small proteins. Future work, involving both experiments and computation, will attempt to structurally characterize the intermediate states discovered here.
Figure: 1D projection of the energy landscape of adenylate kinase molecules, based on analysis of single-molecule fluorescence trajectories measured at 0.5 M GdmCl. Shown are the relative stabilities of the observed states, as well as free energy barriers corresponding to transitions between them. The amount of flux carried by each transition is represented by its line width. The figure shows transitions which carry at least 10% of the folding flux.
Israel Perlman Chemical Sciences Building 603