Computational protocol: Structural basis of epilepsy-related ligand–receptor complex LGI1–ADAM22

Similar protocols

Protocol publication

[…] Diffraction data sets were collected at 100 K at BL41XU in SPring-8 (Hyogo, Japan) and processed with HKL2000 and the CCP4 program suite. The LGI1 LRR structure was determined by molecular replacement using the program Balbes, which selected the Slit2 D4 LRR structure (PDB 2WFH) as the optimal reference model. The structure of the LGI1 EPTP–ADAM22 complex was determined by molecular replacement using the program Molrep. The apo ADAM22 ECD structure (PDB 3G5C) was used as the search model. The atomic model of the entire LGI1 EPTP domain could be obtained after iterative cycles of model building and structure refinement. Information from the 3D structure prediction of LGI1 EPTP was useful for interpretation of the electron density map at the initial stage of the model building. The structure of the LGI1–ADAM22 complex was also determined by molecular replacement using the program Molrep. The structures of LGI1 LRR and the LGI1 EPTP–ADAM22 ECD complex were used as the search models. The programs Coot and Phenix were used for the model building and structure refinement, respectively. Data collection and refinement statistics are shown in Table . [...] A 4.0-µL of purified sample of the LGI1R470A–ADAM22 ECD complex (95 ng µL−1) was applied to a glow-discharged holey carbon grid (R1.2/1.3, Quantiofoil), and the grid was plunge-frozen into liquid ethane by using a semi-automated vitrification device (Vitrobot Mark IV, FEI) with 3-s blotting with 0 blot offset in 100% humidity at 4 °C. Data acquisition was performed by using 200-kV field emission cryo-EM (Tecnai Arctica, FEI) at 23,500-fold nominal magnification with a Gatan K2 summit direct electron detector under low-dose condition using the data acquisition software Serial EM. All the data were collected as a movie with 36 subframes of 1.4 e−1 per Å2 in super-resolution mode with a total electron dose of 50 e−1 per Å2 at a pixel size of 0.785 Å per pixel. The defocus range of the data set was set to a range of −1.5 – −3.5 μm.Movie processing was performed using the software MotionCor2 and CTFFIND4. After the particle picking performed with Gautomatch (http://www.mrc-lmb.cam.ac.uk/kzhang/) for the particles with 200 Å diameter, image processing was performed with RELION2,. Particles were extracted with relatively larger box size, 192 × 192 pixels to have more than one 2:2 LGI1R470A–ADAM22 ECD complex in it. An initial 2D classification was performed with 280 Å diameter mask, which is large enough to accommodate the oligomer particles larger than the dimer particles. The second 2D classification was performed to roughly assign the classes to monomers, dimers, and trimers. Several steps for 2D classification were performed to extract monomers from the dimer classes or dimers from the trimer classes. The final classification for the monomer, which corresponds to the 1:1 LGI1R470A–ADAM22 ECD complex, was performed to 50 classes with the mask of 150 Å. Thirty-nine classes were then selected as the 1:1 LGI1R470A–ADAM22 ECD complex (46,153 particles). The final classification for the dimer, which corresponds to the 2:2 LGI1R470A–ADAM22 ECD complex, was performed to 50 classes with the mask of 280 Å to select only dimer classes. Twenty-four classes were then selected as the 2:2 LGI1R470A–ADAM22 ECD complex (21,163 particles). The final classification for the trimer, which corresponds to the 3:3 LGI1R470A–ADAM22 ECD complex, was performed to 40 classes with the mask of 280 Å to select only trimer classes. Thirteen classes were then selected as the 3:3 LGI1R470A–ADAM22 ECD complex (3780 particles). [...] SEC-SAXS data were collected on beamline BL45XU at SPring-8 (Hyogo, Japan). The LGI1–ADAM22 complex at two different concentrations (5.7 g L−1 or 8.2 g L−1) was applied onto an ENrich SEC 650 (10 × 300 mm) column (Bio-Rad) with 20 mM Tris-HCl (pH 7.5) buffer containing 150 mM or 500 mM NaCl. Scattering intensities were measured at 293 K on PILATUS 3 × 2 M detector with sample to detector distance of 3.5 m. Data collection and structural parameters are listed in Supplementary Table . Four scattering data (0.25 s radiation for each measurement) at elution top peak were averaged by the program PRIMUS and used for the subsequent analysis. The molecular modeling of the 3:3 LGI1–ADAM22 assembly and the subsequent rigid body fitting to the experimental scattering curve were performed using the program SASREF. In this fitting, the LGI1 EPTP–ADAM22 complex was treated as a single rigid body, while LGI LRR was treated as another rigid body. In addition, the Cα–Cα distance restraints were applied on the basis of the crystal structure as follows: 4 Å distance between Ile222 and Ile223 (the two-residue linker between LGI1 LRR and EPTP) and 12 Å distance between Arg474 of one LGI1 and Glu123 of the other LGI1. A simulated annealing protocol was employed to minimize the difference between the experimental and theoretical scattering curves. This fitting was performed using the program CRYSOL. […]

Pipeline specifications

Software tools CCP4, Molrep, Coot, PHENIX, MotionCor2, CTFFIND, Gautomatch, RELION
Applications cryo-EM, Protein structure analysis
Organisms Homo sapiens, Mus musculus
Diseases Brain Diseases, Epilepsy