Figure 4. ~11.5 Å cryo-EM reconstruction of a PG9-Env-8ANC195 complex.

(A) EM density fit by coordinates for PG9 (HC in dark gray; LC in light gray), gp120 subunits in dark blue, cyan, and green, gp41 in orange, and 8ANC195 Fabs in dark blue (HC) and light blue (LC). (B) Density and coordinates from the PG9-Env portion of the PG9-Env-8ANC195 reconstruction (density for 8ANC195 Fabs removed) with coordinates for the PG9 Fab from the PG9-V1V2 scaffold structure (PDB 3U4E) superimposed. When the V1V2 scaffold from the PG9-V1V2 complex structure was superimposed with the V1V2 portion of the PG9-Env-8ANC195 structure (panel A), the angle of approach of the PG9 Fab differs from the angle observed in the PG9-Env structure. (C) Hypothetical model of three PG9 Fabs bound per Env trimer. One PG9 Fab (PG9_A) is shown its experimentally-determined position from the PG9-Env-8ANC195 structure interacting mainly with gp120_A on Env trimer (view looking down the trimer three-fold axis). The V_H-V_L domains for second and third PG9 Fabs (PG9 model) were positioned onto the gp120_B and gp120_C subunits assuming the interaction observed for PG9_A with gp120_A. Predicted van der Waals clashes (red dots) were calculated using the show_bumps module in Pymol (Schrödinger, 2011).

DOI: http://dx.doi.org/10.7554/eLife.27389.013

Figure 4—figure supplement 1. Cryo-EM data processing for PG9-Env-8ANC195 complex structure.

(A) Data processing scheme. 2D classification of autopicked particles resulted in 12 K particles that were 3D classified using a low-pass-filtered negative-stain PG9-Env-8ANC195 EM structure as a reference model. The particles from the best class were refined to 11.5 Å resolution. (B) Orientation distribution of the reconstruction. (C) Local resolution estimation. The complex is shown from top (left), side (middle), and bottom (right) views. (D) Gold-standard FSC indicating 11.5 Å resolution for the reconstruction.

DOI: http://dx.doi.org/10.7554/eLife.27389.014