PRISM-EM: Template interface-based modelling of multi-protein complexes guided by cryo-electron microscopy density maps

Guray Kuzu, Ozlem Keskin*, Ruth Nussinov, Attila Gursoy

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

16 Scopus citations

Abstract

The structures of protein assemblies are important for elucidating cellular processes at the molecular level. Three-dimensional electron microscopy (3DEM) is a powerful method to identify the structures of assemblies, especially those that are challenging to study by crystallography. Here, a new approach, PRISM-EM, is reported to computationally generate plausible structural models using a procedure that combines crystallographic structures and density maps obtained from 3DEM. The predictions are validated against seven available structurally different crystallographic complexes. The models display mean deviations in the backbone of <5 Å. PRISM-EM was further tested on different benchmark sets; the accuracy was evaluated with respect to the structure of the complex, and the correlation with EM density maps and interface predictions were evaluated and compared with those obtained using other methods. PRISM-EM was then used to predict the structure of the ternary complex of the HIV-1 envelope glycoprotein trimer, the ligand CD4 and the neutralizing protein m36.

Original languageEnglish
Pages (from-to)1137-1148
Number of pages12
JournalActa Crystallographica Section D: Structural Biology
Volume72
DOIs
StatePublished - 2016

Funding

FundersFunder number
National Cancer InstituteZIABC010441

    Keywords

    • PRISM-EM
    • modelling protein assemblies
    • multimolecular protein complexes
    • protein structure prediction
    • threedimensional electron microscopy

    Fingerprint

    Dive into the research topics of 'PRISM-EM: Template interface-based modelling of multi-protein complexes guided by cryo-electron microscopy density maps'. Together they form a unique fingerprint.

    Cite this