Computational protocol: Time-course, negative-stain electron microscopy–based analysis for investigating protein–protein interactions at the single-molecule level

Similar protocols

Protocol publication

[…] At time points ranging from 15 s to 24 h, the Fab–trimer mixture was applied to glow-discharged, carbon-coated 400-mesh copper grids as follows. The Fab–trimer mixture was diluted with TBS to a trimer and Fab concentration of ∼91 and 600 nm, respectively. A time point was defined as the time the sample was deposited on the grid after incubation of the trimer–bnAb. For the concentration-based experiments, the trimer concentration was kept constant at 91 nm, whereas the Fab concentration was increased from 50 to 1200 nm, with each sample incubated for 1 h. Staining was performed by applying 3 μl of 2% (w/v) uranyl formate stain and blotting off immediately, followed by application of another 3 μl of stain for 45–60 s, followed by blotting. Stained grids were allowed to air-dry and stored under ambient conditions until ready for imaging. Images were collected via Leginon software (, ) using FEI Talos, Tecnai F20, or Tecnai T12 electron microscopes operated at, respectively, 200 kV ×73,000, 200 kV ×55,000, and 120 kV ×52,000 magnification. In all cases, the electron dose was 25 e−/A2. Particles were picked from the raw images using DoG Picker () and placed into stacks using Appion software (). Finally, 2D reference-free alignment was performed using iterative MSA/MRA)(). Distribution of 0, 1-, 2-, or 3-Fab bound trimers were visually determined by counting the number of total particles (in the top or bottom view) belonging to each category on the 2D class average level for each experiment. RELION 1.4 was used to cross-corroborate the 2D classification results for a subset of the data. Briefly, the particle stacks for the 15-s and 24-h time points (n = 3 for each time point) were converted from IMAGIC to RELION-formatted MRC stacks and subjected to RELION 2D classification (). [...] We prepared synthetic representations of class average data by obtaining the 3D volume of the ZM197 SOSIP.664 HIV-1 trimer with three VRC01 antibodies bound at the CD4-binding site (PDBE EMD-3059). This volume was then progressively modified in UCSF Chimera's () segmentation tool to represent 0, 1, and 2 Fab-bound trimer states. The volumes were then rotated at 10 degree intervals from 0 to 120 degrees and projected in 2D space from the top view. EMAN () was then used to apply small random translational shifts (+ −2px x/y) and a large radius Gaussian filter to simulate noise. This resulted in 960 synthetic class averages with an even distribution of antibody occupancy representation. […]

Pipeline specifications

Software tools Leginon, DoG Picker, Appion, RELION, IMAGIC, UCSF Chimera, EMAN
Databases PDBe
Application cryo-EM
Organisms Human immunodeficiency virus 1, Homo sapiens