High-speed atomic force microscopy (AFM) is a new experimental technique to study the dynamics of biomolecules. It can produce a series of images observing their motions in real-time and in solutions; however, their resolution is at nanometer scales, and they can only provide information on the surface of the molecules. To further understand biomolecular functions, information on three-dimensional (3D) structures is beneficial. Therefore, we have developed an algorithm to recover 3D information from an AFM image by computational modeling. Since the AFM image includes only low-resolution information of a molecule, we represent the structures by a coarse-grained model (Gaussian mixture model). Using Monte-Carlo sampling, we generate candidate models for which the AFM images simulated from the models are similar to the target AFM image. We show that the proposed algorithm could produce a low-resolution 3D model of the target molecule using only a simulated AFM image as a reference. In addition, we examined the effect of molecular orientations captured in AFM images on the 3D modeling performance and showed that similar accuracy could be obtained for many orientations.<br/><br/>Furthermore, we applied this algorithm to recent HS-AFM data on ClpB, which shows several predominant conformational classes. ClpB belongs to the cellular disaggregate machinery involved in rescuing misfolded or aggregated proteins during heat or other cellular shocks. We reconstruct 3D models of ClpB from HS-AFM images in different conformational classes. After conformational sampling, we obtained models that reflect conformational variety embedded within the AFM images. From these reconstructed 3D models, we described, in terms of relative domain arrangement, the different types of ClpB oligomeric conformations observed by HS-AFM experiments.<br/><br/>In conclusion, this new AFM-based hybrid modeling method would be a valuable tool for retrieving 3D information of the entire biomolecules from HS-AFM experimental data, even with limited resolution, which is helpful for annotations of the different conformational states involved functions.

Reconstruction algorithm of low-resolution three-dimensional conformations of biomolecules from atomic force microscopy images

ACS Spring 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.