A grid-optimization matrix is screened: 2 grid types (e.g., Quantifoil R1.2/1.3 and R2/1 holey carbon on 300-mesh Cu) x 2 protein concentrations (1 and 3 mg/mL) x 3 blot times (3, 5, 7 s) at fixed 4 C and 100% humidity, with blot force fixed at a mid value, giving 12 conditions, each frozen in duplicate (24 grids). Grids are screened on a 200 kV microscope for ice quality and particle distribution before selecting the best 1-2 for high-resolution collection on a 300 kV system. Surface treatment (with/without detergent additive such as a low CMC fluorinated surfactant) is compared on the best two conditions to address air-water-interface and preferred orientation.
BSL-1 for non-pathogenic recombinant samples. Major hazards are cryogenic and flammable: liquid ethane is extremely flammable and an asphyxiant/narcotic vapor; condense and use only in a fume hood or well-ventilated area away from ignition sources, never in a sealed room. Liquid nitrogen causes cryogenic burns and asphyxiation in enclosed spaces; use an O2 monitor, cryo-gloves, face shield, and lab coat, and never seal LN2 in a closed vessel (explosion risk). Dispose of ethane by controlled evaporation in the hood. The TEM and FEG carry high-voltage and X-ray interlocks; follow facility SOPs.
Positive control: a standard sample (e.g., apoferritin or mouse heavy-chain apoferritin, or GroEL) vitrified in the same session to confirm cryogen quality, blotting performance, and that the microscope/detector resolve high-resolution information. Buffer-only control grid: identifies contaminating ice/ethane crystals and tests for empty-hole ice quality independent of protein. Untreated vs glow-discharged grid comparison controls for surface hydrophilicity. The non-additive condition serves as the reference for evaluating any surfactant's effect on orientation.
Optimal grids show a gradient of ice across the grid with a usable region of thin vitreous ice (~20-50 nm) where individual particles are visible inside holes, ~50-300 particles per hole depending on size, and Thon rings extending toward the detector's Nyquist on power spectra. Good data yields broad Euler-angle coverage and, after motion correction and CTF estimation, micrographs supporting sub-4 Å 2D/3D. Thick ice (gray, no particle contrast) or crystalline/hexagonal ice (reflections in the FFT) indicate failed vitrification or devitrification; empty holes indicate too-low concentration or over-blotting.
To vitrify a purified macromolecular complex in a thin layer of amorphous ice across the holes of holey carbon grids using an automated plunge-freezer, optimizing glow-discharge, blot time, blot force, and particle concentration to achieve thin ice (~20-50 nm), even particle distribution, and minimal preferred orientation for downstream high-resolution reconstruction.
Independent: grid type/hole size, protein concentration, blot time, blot force, surfactant additive presence. Dependent: ice thickness, particles per hole, orientation-distribution uniformity (Euler angle coverage), high-resolution CTF fit (Thon ring extent), final reconstruction resolution. Controlled: chamber temperature (4 C) and humidity (100%), ethane temperature, glow-discharge settings, operator, sample batch, total dose during screening.
A monodisperse complex confirmed by negative-stain EM, applied at ~1-4 mg/mL to a freshly glow-discharged holey grid and blotted under controlled humidity/temperature, will partition into the holes as single particles embedded in vitreous ice, giving micrographs with visible secondary structure at the detector and yielding a 3D reconstruction at sub-4 Å resolution after data collection.
Screening micrographs are assessed in the microscope software and cryoSPARC/RELION live for ice thickness (using the aperture-limited scattering or an ice-thickness estimate), particles per hole, and Thon-ring resolution. Full datasets undergo motion correction (MotionCor2/Relion), CTF estimation (CTFFIND4/patch CTF), particle picking, 2D classification, ab initio + heterogeneous refinement, and homogeneous/non-uniform refinement; resolution is reported by gold-standard FSC at 0.143. Orientation distribution is quantified by the cryoSPARC/relion Euler-angle plot and an orientation-bias metric (e.g., efficiency).
Optimization is descriptive/comparative: report mean +/- SD particles-per-hole and ice-thickness across n >= 20 holes per condition, and compare conditions by one-way ANOVA with Tukey HSD (alpha = 0.05). Orientation uniformity compared between additive and no-additive grids by a chi-square goodness-of-fit against a uniform angular distribution. Final map resolution reported as a single FSC=0.143 value with the gold-standard half-map convention; local resolution reported as a distribution. No conventional power calculation; reproducibility established by independently freezing and reconstructing from two separate grids.