Expression and purification of LetAB
Plasmid pBEL1284, which encodes N-terminal 6×His2xQH-TEV-tagged LetA and untagged LetB, was transformed into OverExpress C43 (DE3) cells (60446-1, Lucigen; Supplementary Table 1). For protein expression, overnight cultures (LB, 100 µg ml−1 carbenicillin and 1% glucose) were diluted in LB (Difco) supplemented with carbenicillin (100 µg ml−1), grown at 37 °C with shaking to an optical density at 600 nm (OD600) of approximately 0.9, and then induced by addition of arabinose to a final concentration of 0.2%. Cultures were further incubated at 37 °C with shaking for 4 h, and then harvested by centrifugation. The pellets were resuspended in lysis buffer (50 mM Tris pH 8.0, 300 mM NaCl and 10% glycerol), flash frozen in liquid nitrogen and stored at −80 °C. Cells were lysed by two passes through an Emulsiflex-C3 cell disruptor (Avestin), then centrifuged at 15,000g for 30 min at 4 °C to pellet cell debris. The clarified lysate was subjected to ultracentrifugation at 37,000 rpm (182,460g) for 45 min at 4 °C in a Fiberlite F37L-8 ×100 Fixed-Angle Rotor (096-087056, Thermo Fisher Scientific). The supernatant was discarded and the membrane fraction was solubilized in 50 mM Tris pH 8.0, 300 mM NaCl, 10% glycerol, 25 mM n-dodecyl-β-d-maltoside (DDM) by rocking overnight at 4 °C. Insoluble debris were pelleted by ultracentrifugation at 37,000 rpm for 45 min at 4 °C. Solubilized membranes were then passed twice through a column containing Ni Sepharose Excel resin (Cytiva). Eluted proteins were concentrated using an Amicon Ultra-0.5 Centrifugal Filter Unit concentrator (MWCO 100 kDa, UFC510096) before separation on the Superdex 200 Increase 10/300 column (Cytiva) equilibrated with either Tris (20 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.5 mM DDM and 10% glycerol) or HEPES (20 mM HEPES pH 7.4, 150 mM NaCl, 0.5 mM DDM and 10% glycerol) gel-filtration buffer. Fractions containing LetAB were pooled, concentrated and applied to grids for negative-stain electron microscopy or cryo-EM. One litre of culture typically yields 30–40 µg of the LetAB complex.
Negative-stain electron microscopy
To prepare grids for negative-stain electron microscopy analysis, a fresh sample of LetAB was applied to a freshly glow-discharged (30 s) carbon-coated 400 mesh copper grid (01754-F, Ted Pella) and blotted off. Immediately after blotting, 3 µl of a 2% uranyl formate solution was applied for staining and blotted off on filter paper (Whatman 1) from the opposite side. Application and blotting of stain was repeated four times. The sample was allowed to air dry before imaging. A negative-stain grid of the soluble, periplasmic domain of LetB was prepared previously31 using a similar procedure and stored. New images from this sample were acquired for this study. Data were collected on the Talos L120C TEM (FEI) equipped with the 4K × 4K OneView camera (Gatan) at a nominal magnification of ×73,000 corresponding to a pixel size of 2.03 Å px−1 on the sample and a defocus range of −1 to −2 μm. Negative-stain dataset size was determined to be sufficient by the ability to see features in the 2D classes of picked particles. For both the LetAB and LetB datasets, micrographs were imported into cryoSPARC (v3.3.1)54 and approximately 200 particles were picked manually, followed by automated template-based picking. Particles were extracted with a 320 pixel box size. Several rounds of 2D classification were performed using default parameters, except that ‘force max over poses/shifts’ and ‘do CTF correction’ were both set to false.
Cryo-EM sample preparation and data collection
To generate the crosslinked LetAB sample, 1% glutaraldehyde was added to purified LetAB (HEPES gel-filtration buffer) at a final concentration of 0.025%. The sample was incubated on ice for 1 h and then quenched by the addition of 75 mM Tris-HCl pH 8.0. The sample was incubated for 15 min on ice before filtering using an Ultrafree centrifugal filter (UFC30GVNB) and loading onto a Superdex 200 Increase 10/300 column (Cytiva), equilibrated with buffer containing 20 mM Tris-HCl pH 8.0, 150 mM NaCl and 0.5 mM DDM, to remove aggregated LetAB. Fractions containing the LetAB complex were concentrated to 1 mg ml−1 using the Amicon Ultra-0.5 centrifugal filter unit concentrator (MWCO 100 kDa, UFC510096). Continuous carbon grids (Quantifoil R 2/2 on Cu 300 mesh grids + 2 nm Carbon, Quantifoil Micro Tools, C2-C16nCu30-01) were glow-discharged for 5 s in an easiGlow Glow Discharge Cleaning System (Ted Pella). Of freshly prepared sample, 3 µl was added to the glow-discharged grid. Grids were prepared using a Vitrobot Mark IV (Thermo Fisher Scientific). Grids were blotted with a blot force of 0 for 3 s at 4 °C with 100% chamber humidity and then plunge-frozen into liquid ethane. Grids were clipped for data acquisition.
Grids containing crosslinked LetAB were screened at the NYU cryo-EM core facility on the Talos Arctica (Thermo Fisher Scientific) equipped with a K3 camera (Gatan). The grids were selected for data collection on the basis of ice quality and particle distribution. The selected cryo-EM grid was imaged on two separate sessions at the Pacific Northwest Center for Cryo-EM (PNCC) on Krios-1, a Titan Krios G3 electron microscope (Thermo Fisher Scientific) equipped with a K3 direct electron detector with a BioContinuum energy filter (Gatan). Super-resolution movies were collected at 300 kV using SerialEM55 at a nominal magnification of ×81,000, corresponding to a super-resolution pixel size of 0.5144 Å (or a nominal pixel size of 1.029 Å after binning by 2). Movies (n = 12,029) movies were collected over a defocus range of −0.8 to −2.1 µm, and each movie consisted of 50 frames with a total dose of 50 e− Å−2. Further data collection parameters are shown in Extended Data Table 1.
The uncrosslinked LetAB complex was prepared as described in ‘Expression and purification of LetAB’, except the Superdex 200 Increase 10/300 column was equilibrated in buffer containing 20 mM Tris-HCl pH 8.0, 150 mM NaCl and 0.5 mM DDM. Continuous carbon grids (Quantifoil R 2/2 on Cu 300 mesh grids + 2 nm Carbon, Quantifoil Micro Tools, C2-C16nCu30-01) were glow-discharged for 5 s in an easiGlow Glow Discharge Cleaning System (Ted Pella). Of freshly prepared sample (1 mg ml−1) in Tris gel filtration buffer, 3 µl was added to the glow-discharged grid. Grids were prepared using a Vitrobot Mark IV (Thermo Fisher Scientific). Grids were blotted with a blot force of 0 for 3 s at 4 °C with 100% chamber humidity and then plunge-frozen into liquid ethane. Grids were clipped for data acquisition. Grids were screened at the NYU cryo-EM laboratory on the Talos Arctica (Thermo Fisher Scientific) system equipped with a K3 camera (Gatan). The grid with the best ice quality and particle distribution was imaged at the New York Structural Biology Center on Krios-1, a Titan Krios G3 electron microscope (Thermo Fisher Scientific) equipped with K3 direct electron detector with a BioContinuum energy filter (Gatan). Super-resolution movies were collected at 300 kV using Leginon56 at a nominal magnification of ×81,000, corresponding to a super-resolution pixel size of 0.5413 Å (or a nominal pixel size of 1.083 Å after binning by 2). Movies were collected over a defocus range of −2 to −5 µm, and each movie consisted of 40 frames with a total dose of 51 e− Å−2. A total of 12,455 movies were collected, consisting of 5,372 movies at 0° tilt and 7,083 movies at −30° tilt. Further data collection parameters are shown in Extended Data Table 2.
Cryo-EM structure of crosslinked LetAB
The data processing workflow for the crosslinked LetAB sample is shown in Supplementary Fig. 2e. A combination of cryoSPARC (v3.2.0–4.3.0) and RELION (v3.1.0)57 were used for data processing. Dose-fractionated movies were gain normalized, drift corrected, summed and dose weighted, and binned by 2 using the cryoSPARC Patch Motion module. Contrast transfer function (CTF) estimation for each summed image was carried out using cryoSPARC Patch CTF estimation. To generate 2D templates for auto-picking, 1,003 particles were manually picked, extracted (box size of 576 px) and subjected to 2D classification. The classes with top, tilted and side views of LetAB were selected as templates for auto-picking, which yielded 3,582,925 particles after extraction (box size of 576 px). The particles were subjected to 2D classification (200 classes) with force max over poses/shifts set to false. Well-aligned 2D classes were selected (1,793,362 particles) and a 3D reconstruction was generated using ab initio reconstruction. The 3D reconstruction was used as a template for 3D refinement in RELION, which revealed well-resolved density for MCE rings 1–4, poor and noisy density for MCE rings 5–7 and no density for LetA, probably due to rings 1–4 dominating the particle alignment. To improve resolution for LetA, local refinement was performed using a mask around LetB rings 1–4 and the TM region, followed by particle subtraction in RELION where the signal for the TM region and rings 1–2 was kept and recentred to the middle of the box (256 px). The subtracted particles were imported into cryoSPARC, where reference-free 3D classification was performed using the ab initio module (two classes) to remove misaligned and ‘junk’ particles. This resulted in one class with 1,171,725 particles with high-resolution features. The particles from this class were further cleaned using 2D classification and then subjected to non-uniform refinement (942,263 particles). The aligned particles were imported into RELION and sorted using 3D classification without alignment (eight classes), which revealed one class containing density for the TM region with high-resolution features. The particles (158,666) were then subjected to local refinement in RELION, yielding a map (Map 1a) with a nominal resolution of 3.4 Å (Supplementary Fig. 2b).
... continue reading