Freshly prepared sample was applied to 2 nm pre-coated Quantifoil R3/3 holey carbon supported grids and vitrified using a Vitrobot Mark IV (FEI Company) and visualized on a Spirit TEM (FEI Company) with about 20e− Å−2 at a nominal magnification of × 105,000 with a nominal defocus between −1 μm and −3.5 μm. Automatic particle detection was performed by the programme SIGNATURE49 (link). Initial in silico sorting of the data set consisting of 54,800 particles in total was performed using the SPIDER software package49 (link). Classes were obtained by competitive projection matching in SPIDER50 (link)51 (link). The final 30S·ABCE1 data set contained 19,500 particles and the final resolution was 17 Å (Fourier shell correlation 0.5).
For interpretation of the 30S·ABCE1 electron density at a molecular level, the models for the Pyrococcus furiosus 30S subunit (4V6U)52 (link) and ribosome-bound ABCE1 in (3J15)8 (link) were fitted as rigid bodies using UCSF Chimera. The FeS cluster domain was repositioned by a rotation of ∼160° around a hinge (residues 76–78) into an unaccounted electron density near ribosomal protein S12. This repositioning results in a close contact between lysine 60 of ABCE1 (Lys64 in P. furiosus) and lysine 40 of S12 and is consistent with above described XL-MS data.
For interpretation of the 30S·ABCE1 electron density at a molecular level, the models for the Pyrococcus furiosus 30S subunit (4V6U)52 (link) and ribosome-bound ABCE1 in (3J15)8 (link) were fitted as rigid bodies using UCSF Chimera. The FeS cluster domain was repositioned by a rotation of ∼160° around a hinge (residues 76–78) into an unaccounted electron density near ribosomal protein S12. This repositioning results in a close contact between lysine 60 of ABCE1 (Lys64 in P. furiosus) and lysine 40 of S12 and is consistent with above described XL-MS data.
Free full text: Click here
