Bottom-up structural proteomics: cryoEM of protein complexes enriched from the cellular milieu

Abstract

X-ray crystallography and recombinant protein production have enabled an exponential increase in atomic structures, but often require non-native constructs involving mutations or truncations, and are challenged by membrane proteins and large multi-component complexes. We present here a bottom-up endogenous structural proteomics approach whereby near-atomic resolution cryoEM maps are reconstructed ab initio from unidentified protein complexes enriched directly from the endogenous cellular milieu, followed by identification and atomic modeling of the proteins. The proteins in each complex are identified using cryoID , a program we developed to identify proteins in ab initio cryoEM maps. As a proof of principle, we applied this approach to the malaria parasite Plasmodium falciparum , an organism that has resisted conventional structural biology approaches, to obtain atomic models of multiple protein complexes implicated in intraerythrocytic survival of the parasite. Our approach is broadly applicable for determining structures of undiscovered protein complexes enriched directly from endogenous sources.

Atomic structures of respiratory complex III2, complex IV and supercomplex III2-IV from vascular plants

“…Upon 2D classification of the particles in the micrographs, and supported by mass spectrometry data (Table 1), it became evident that the pooled fractions contained not only a CI intermediate [46] , but also CIII2 and SC III2+IV. Therefore, CIII2 and SCIII+CIV were purified in silico, in an example of "bottom-up structural proteomics" [47] on a partially purified sample. Processing of the CIII2 and SCIII2+IV particles resulted in near-atomic reconstructions of CIII2 at 3.2 Å, of CIV at 3.8 Å and of SCIII2+IV at 3.8 Å (Figure 1, Videos 1-3, Figure 1- Figure Supplements 1-3, Tables.…”

Structural snapshots of human pre-60S ribosomal particles before and after nuclear export

“…Therefore, extensive trial-and-error modeling of potential sequences from the mass spectrometry data (as well as the sequence of NMD3-NTD) were performed. In addition, a map-derived degenerate amino-acid sequence based method (CryoID) 61 was also tried. None of these efforts worked and the identity of protein X remains unknown.…”

Full-lengthde novoprotein structure determination from cryo-EM maps using deep learning

“…More and more cryo-EM maps are being generated, many of which are expected to have near-atomic resolutions (3-5Å). In addition, an advantage of cryo-EM is its ability to obtain the EM maps of multiple macromolecules in one sam-ple and therefore can be applied to structural proteomics for structure measurement in the cellular milieu 42 . Such application will generate a large number of near-atomic resolution EM maps in one experiment.…”

“…The corresponding amino acid type and second structure type marked by STRIDE 43 were assigned to each point. Twenty types of amino acids were grouped into four classes according to their sizes, shapes and distributions in their EM density maps 42 , as illustrated in Figure 2d. Specifically, GLY, ALA, SER, CYS, VAL, THR, ILE and PRO are grouped as Class I. LEU, ASP, ASN, GLU, GLN and MET are grouped as Class II.…”