Single molecule studies of protein binding and aggregation

National Institute Of Diabetes And Digestive And Kidney Diseases

Investigators

Linked publications, trials & patents

Abstract

Protein aggregation is implicated as the cause of pathology in various diseases such as Alzheimers and Parkinsons disease. Aggregation of the proteins involved in these diseases is the process of amyloid fibril formation and the oligomers, the intermediates appearing during aggregation, are thought to be toxic. However, these oligomers are relatively unstable species and a heterogeneous mixture and their characterization is very challenging using ensemble measurements. We carried out a single-molecule FRET experiment to detect stable amyloid-beta (Abeta) oligomers that appear during the aggregation process. However, quantitative analysis was difficult because of their extremely low concentration, structural heterogeneity, and a broad range of oligomer size. We collaborated with Dr. Irina Gopich in LCP to develop a molecular diffusion analysis method of diffusion-based single molecule FRET data (theory was published in Journal of Chemical Physics, Dr. Gopich corresponding author). This method can accurately determine the brightness, diffusion time, FRET efficiency, and population of oligomer species in solution. We have applied this method to the oligomer formation of 42-residue Abeta and found that the average size of oligomers appearing before and during aggregation is 70 and these oligomers should have a rod-like shape similar to amyloid fibrils. This work was published in PNAS Nexus. The method was also used for the oligomer detection in a mixture of 40- and 42-residue Abeta and a manuscript is in preparation. In collaboration with Dr. Jinwei Zhangs group in LMB, we have studied conformational dynamics and interactions of noncoding RNAs using two- and three-color single-molecule FRET spectroscopy. Compared to protein folding, much less is understood about how complex RNAs fold into tertiary structures. We have studied the T-box RNA, a large multi-domain noncoding RNA that directly binds cognate tRNAs as ligands, senses their aminoacylation status to detect starvation, and changes conformations to feedback control the transcription or translation of downstream amino acid genes. We found that T-box riboswitch has a broad conformational distribution, which can be clustered into three states based on its degree of extension: open, semi-open, and closed conformations. Interestingly all these states can bind tRNA, which indicates that binding occurs via an induced fit-like process. More importantly, we found that the conformational states of T-box before binding is retained (i.e., memory) in the bound state and many molecules come back to the original unbound conformation after dissociation of tRNA. Using a maximum likelihood analysis of single-molecule fluorescence trajectories, we could extract complex kinetic model parameters and how this memory effect plays in the efficient binding of tRNA to T-box. This work was accepted in Nature Communications.