Study of protein structure by electron microscopy and single particle image processing

Dandey, Venkata Prasad. Study of protein structure by electron microscopy and single particle image processing. 2015, PhD Thesis, University of Basel, Faculty of Science.


Official URL: http://edoc.unibas.ch/diss/DissB_11230


The molecular assemblies form an important role in the understanding the machinery of living cells. These assemblies with thickness less than 1000nm and exhibiting different shapes, size and biochemical states are studied using Transmission Electron Microscopy (TEM). The ordered arrays of assemblies such as 2D crystals of membrane proteins and Helices are first reconstructed to near atomic resolution by image processing of its EM images. High-resolution structures of 2D crystals and helices using EM are very rare, due to inability of many complexes to form highly ordered arrays. If the proteins are large enough, it can be prepared as single particles and can be studied using EM. Single particle based approach is widely applicable approach for any symmetry and in the range of 0.1 - 100 MDa molecular weight complexes. This approach has caught up crystallography to reach near atomic resolution after the advancements in electron detector. In single particle approach many identical particles are expected to present on an EM grid and then 3D structure is calculated from the many views of the same molecule. The quality of the 3D structure is highly dependent on homogenous preparations of single particles.
The aim of this thesis is to determine the structure of molecular assemblies through single particle approach, focusing on their image processing. Also, the single particle based helical image processing techniques, applied to very thin and weakly ordered helical filaments, are also studied. In the methods section (Chapter~\ref{ch:two}-\ref{ch:three}), both methods related to conventional single particle reconstruction (SPR) for non-homogenous single particles and single particle based Iterative helical real space reconstruction (IHRSR) approaches are presented. The workflow of semi automated tool chain for SPR applied to heterogeneous sample, from automated particle picking to the final 3D density maps, is explained. The limitation of IHRSR technique is its dependence on initial guess helical symmetry parameters for the helical filaments. So, determining the helical symmetry by analysis of diffraction the pattern of the helical objects is explained in detail.\\
The second part of the thesis is the application of these methods in structural studies of unknown molecular assemblies. The SPR techniques are applied to study the heterogeneity of the Silent Information Regulatory (SIR) protein complexes. The flexibility among the different subunits of the SIR complex and existence of mixed structures are revealed using single particle based random conical tilt reconstruction (RCT) approach, which supports the results of MALDI-TOF mass spectrometer results. To analyze this sample, a semi-automated EM tool chain has been developed in house to automate Tecnai T12 electron microscope for collecting thousands of images per day and followed by semi automated SPR techniques.The structural studies of Pyrin domain of Apoptisis Speck like CARD domain (ASC-PYD) by Cryo-EM using IHRSR technique revealed the structure of its filaments to 3.7{\AA}. A novel hybrid approach combining the solution NMR, solid-state NMR and Cryo-EM techniques are used to obtain precise structural details of PYD filaments to atomic resolution. This novel method is expected to become a routine to complement both Cryo-EM and NMR results.
In Chapter~\ref{ch:six}-\ref{ch:seven}, the semi-automatic image acquisition and analysis methods used for TEM investigations in combination with isolation methods are used to extract the quantitative information on protein abundances.
Advisors:Stahlberg, Henning
Committee Members:Roth, Volker
Faculties and Departments:05 Faculty of Science > Departement Biozentrum > Structural Biology & Biophysics > Structural Biology (Stahlberg)
Item Type:Thesis
Thesis no:11230
Bibsysno:Link to catalogue
Number of Pages:115 S.
Identification Number:
Last Modified:30 Jun 2016 10:57
Deposited On:08 Jun 2015 13:45

Repository Staff Only: item control page