The application of support layers to cryogenic electron microscopy (cryoEM) grids can increase particle density, limit interactions with the air-water interface, reduce beam-induced motion, and improve the distribution of particle orientations. This paper describes a robust protocol for coating cryoEM grids with a monolayer of graphene for improved cryo-sample preparation.
In cryogenic electron microscopy (cryoEM), purified macromolecules are applied to a grid bearing a holey carbon foil; the molecules are then blotted to remove excess liquid and rapidly frozen in a roughly 20-100 nm thick layer of vitreous ice, suspended across roughly 1 µm wide foil holes. The resulting sample is imaged using cryogenic transmission electron microscopy, and after image processing using suitable software, near-atomic resolution structures can be determined. Despite cryoEM's widespread adoption, sample preparation remains a severe bottleneck in cryoEM workflows, with users often encountering challenges related to samples behaving poorly in the suspended vitreous ice. Recently, methods have been developed to modify cryoEM grids with a single continuous layer of graphene, which acts as a support surface that often increases particle density in the imaged area and can reduce interactions between particles and the air-water interface. Here, we provide detailed protocols for the application of graphene to cryoEM grids and for rapidly assessing the relative hydrophilicity of the resulting grids. Additionally, we describe an EM-based method to confirm the presence of graphene by visualizing its characteristic diffraction pattern. Finally, we demonstrate the utility of these graphene supports by rapidly reconstructing a 2.7 Å resolution density map of a Cas9 complex using a pure sample at a relatively low concentration.
Single particle cryogenic electron microscopy (cryoEM) has evolved into a widely used method for visualizing biological macromolecules1. Fueled by advances in direct electron detection2,3,4, data acquisition5, and image processing algorithms6,7,8,9,10, cryoEM is now capable of producing near-atomic resolution 3D structures of a fast-growing number of macromolecules11. Moreover, by leveraging the single-molecule nature of the approach, users can determine multiple structures from a single sample12,13,14,15, highlighting the promise of using the data generated to understand heterogeneous structural ensembles16,17. Despite this progress, bottlenecks in cryo-specimen grid preparation persist.
For structural characterization by cryoEM, biological samples should be well-dispersed in aqueous solution and then must be flash-frozen through a process called vitrification18,19. The goal is to capture particles in a uniformly thin layer of vitrified ice suspended across regularly spaced holes that are typically cut into a layer of amorphous carbon. This patterned amorphous carbon foil is supported by a TEM grid bearing a mesh of copper or gold support bars. In standard workflows, grids are rendered hydrophilic using a glow-discharge plasma treatment prior to the application of sample. Excess liquid is blotted with filter paper, allowing the protein solution to form a thin liquid film across the holes that can be readily vitrified during plunge-freezing. Common challenges include particle localization to the air-water interface (AWI) and subsequent denaturation20,21,22 or adoption of preferred orientations23,24,25, particle adherence to the carbon foil rather than migrating into the holes, and clustering and aggregation of the particles within the holes26. Nonuniform ice thickness is another concern; thick ice can result in higher levels of background noise in the micrographs due to increased electron scattering, whereas extremely thin ice can exclude larger particles27.
To address these challenges, a variety of thin support films have been used to coat grid surfaces, allowing particles to rest on these supports and, ideally, avoid interactions with the air-water interface. Graphene supports have shown great promise, in part due to their high mechanical strength coupled with their minimal scattering cross-section, which reduces the background signal added by the support layer28. In addition to its minimal contribution to background noise, graphene also exhibits remarkable electrical and thermal conductivity29. Graphene and graphene oxide coated grids have been shown to yield higher particle density, more uniform particle distribution30, and reduced localization to the AWI22. In addition, graphene provides a support surface that can be further modified to: 1) tune the physiochemical properties of the grid surface through functionalization31,32,33; or 2) couple linking agents that facilitate affinity purification of proteins of interest34,35,36.
In this article, we have modified an existing procedure for coating cryoEM grids with a single uniform layer of graphene30. The modifications aim to minimize grid handling throughout the protocol, with the goal of increasing yield and reproducibility. Additionally, we discuss our approach to evaluate the efficacy of various UV/ozone treatments in rendering grids hydrophilic prior to plunging. This step in cryoEM sample preparation using graphene-coated grids is critical, and we have found our straightforward method to quantify the relative hydrophilicity of the resulting grids to be useful. Using this protocol, we demonstrate the utility of employing graphene-coated grids for structure determination by generating a high-resolution 3D reconstruction of catalytically inactive S. pyogenes Cas9 in complex with guide RNA and target DNA.
1. Preparation of CVD graphene
2. Coating CVD graphene with MMA
3. Plasma etching of the graphene back-side
4. Cutting grid-sized MMA coated CVD graphene squares
5. Dissolving copper substrate from MMA coated CVD graphene
6. Removing MMA/graphene films from APS
7. Adhering graphene to grids
8. Dissolving MMA with acetone
9. Removing residual acetone with isopropanol
10. UV/ozone treatment of graphene-coated grids
11. Capturing a diffraction image
12. Assessing grid hydrophilicity
13. Single particle analysis of the dCas9 complex dataset
NOTE: All image processing described in this protocol was performed using cryoSPARC version 4.2.1.
Successful fabrication of graphene-coated cryoEM grids using the equipment (Figure 1) and protocol (Figure 2) outlined here will result in a monolayer of graphene covering the foil holes that can be confirmed by its characteristic diffraction pattern. To promote protein adsorption to the graphene surface, UV/ozone treatment can be used to render the surface hydrophilic by installing oxygen-containing functional groups. However, hydrocarbon contaminants in the air can adsorb onto the graphene surface as early as 5 min post UV/ozone treatment and counteract this effect38,39. Importantly, both the duration of UV/ozone treatment and the time elapsed between treatment and plunging can affect sample quality. We demonstrate these effects using a simple method for assessing the hydrophilic character of the coated grid based on the surface contact angle (Figure 3; see step 12).
To demonstrate the use of graphene supports in single particle cryoEM, we applied a catalytically inactive RNA-guided DNA endonuclease S. pyogenes Cas9 (H10A; C80S; C574S; H840A)40 in complex with sgRNA and target DNA to graphene-coated grids, collected a cryoEM dataset from these grids, and performed single particle analysis7. Graphene-coated grids consistently contained ~300 particles per micrograph at 0.654 Å/pix magnification using a 300 keV microscope equipped with a K3 direct electron detector (Figure 4A-E). An 8 h data collection session with a +18° stage tilt yielded 2,963 movies and 324,439 particles in a final curated stack. Using these particles, we generated a 3D reconstruction which, upon refinement, yielded a density map with an estimated resolution of 2.7 Å and adequate angular sampling to avoid anisotropic artifacts (Figure 5). An atomic model (PDB 6o0z)41 was docked into this map, and refined using ISOLDE42. Residues R63-L82 of this fitted atomic model are displayed with the refined cryoEM density map, highlighting the resolved side-chain density (Figure 5B). When comparing the same sample and concentration (250 ng/µL) applied to identical grids that lacked graphene, no particles were observed (Figure 4F,G). This observation highlights the efficacy of the graphene support in enabling the visualization of particles from low-concentration samples.
Figure 1: Required equipment. Lab equipment and tools necessary for the fabrication of graphene grids using the protocol detailed in this article. Items and their quantity are shown and labeled accordingly. Requisite reagents that are not shown include: CVD graphene, methyl-methacrylate EL-6 (MMA), ammonium persulfate (APS), acetone, isopropanol, ethanol, molecular grade water. Requisite instruments that are not shown include: spin coater, glow discharger, hot plate, vacuum desiccator, and thermometer. All requisite items are detailed in the Table of Materials. Please click here to view a larger version of this figure.
Figure 2: Schematic of the graphene grid fabrication process. Graphene is coated with a thin layer of methyl-methacrylate EL-6 (MMA) using a spin coater (step 2). Graphene on the opposite side of the copper foil is removed via plasma etching (step 3). Ammonium persulfate (APS) is then used to etch away the copper (steps 4-5). The MMA-graphene film is placed onto the grid surface (step 6-7). Lastly, MMA is dissolved during a series of washes with organic solvents (steps 8-9). Steps indicated above arrows correspond to numbered steps described in the protocols section. This method has been adapted from Han et al.30. Please click here to view a larger version of this figure.
Figure 3: Assessment of grid surface hydrophilicity as a function of duration of UV/ozone treatment and time elapsed post-treatment. (A) Measured contact angles plotted as a function of the duration of treatment. Decreased contact angles are consistent with increased hydrophilicity (untreated grid: 78°; 20 min: 37°). Contact angles measured using ImageJ43. (B) Measured contact angles plotted as a function of time, post treatment (0 min: 45°; 60 min: 74°). Grid measured in the post treatment time-course was UV/ozone treated for 12 min, as indicated by asterisk. Each post-treatment measurement was performed on the same grid, with the sample removed by wicking between measurements. Specific contact angles measured are expected to vary as a function of laboratory environmental conditions, and we recommend that users perform similar experiments in their laboratories to identify suitable conditions. Please click here to view a larger version of this figure.
Figure 4: Representative images of graphene-coated grid and uncoated control grids. (A-C) Representative atlas, grid square, and foil hole images of graphene-coated holey carbon grids taken on a 300 keV microscope equipped with a K3 direct electron detector. (D) CryoEM micrograph of S. pyogenes dCas9 in complex with sgRNA and target DNA (complex at 250 ng/µL concentration) on graphene-coated holey carbon grid. (E) Diffraction image from grid imaged in panels (A-D). Orange arrow indicates a position corresponding to a spatial frequency of 2.13 Å. An identical sample to that in panel (D) was applied to (F) UV/ozone treatment and (G) glow discharged holey carbon grids without graphene. CryoEM micrographs displayed are representative of each grid and show no particles. Please click here to view a larger version of this figure.
Figure 5: CryoEM reconstruction of a Cas9 complex from graphene-coated grids. (A) CryoEM density map from 3D reconstruction of the S. pyogenes dCas9 in complex with sgRNA and target DNA. (B) Residues R63-L82 from a fitted model are depicted within the semi-transparent cryoEM density, with a subset of visible sidechains labeled. (C) Fourier Shell Correlation (FSC) curves of the unmasked, loosely, and tightly masked maps. (D) Histogram and directional FSC plot based on the 3DFSC method23. See Supplementary Figure 2 and step 13 for more information. Please click here to view a larger version of this figure.
Supplementary Figure 1: Contact angle imaging stand. (A) A 3D printed camera stand and tabletop imaging mount to secure camera in a position that aligns the camera in-plane with the coverslip. (B) Grid is placed on the coverslip on top of a 1 cm x 1 cm square piece of paraffin film. The depicted imaging mount was 3D-printed using the provided .stl files (Supplementary Coding File 1 and Supplementary Coding File 2) and can be readily modified to accommodate most devices. Please click here to download this File.
Supplementary Figure 2: Image processing workflow. Processing workflow for dCas9 complex. Job names, job details, and non-default parameters (italicized) are indicated. Please click here to download this File.
Supplementary Coding File 1: Stereolithography CAD files in the STL format are provided to facilitate 3D printing of camera stand (camera_stand_v1.stl). Please click here to download this File.
Supplementary Coding File 2: Stereolithography CAD files in the STL format are provided to facilitate 3D printing of the tabletop imaging mount (slide_mount_v1.stl) and Please click here to download this File.
CryoEM sample preparation involves a host of technical challenges, with most workflows requiring researchers to manually manipulate fragile grids with extreme care to avoid damaging them. Additionally, the amenability of any sample to vitrification is unpredictable; particles often interact with the air-water-interface or with the solid support foil overlaying the grids, which can lead to particles adopting preferred orientations or failing to enter the imaging holes unless very high protein concentrations are applied24. Overlaying holey cryoEM grids with a continuous monolayer of graphene has shown tremendous promise in improving particle distributions on micrographs, increasing particle numbers at low concentrations, and reducing preferred orientations driven by interactions at the air-water-interface30.
A limitation of existing graphene coating protocols for cryoEM grids is the extensive manual manipulations required for the coating process, which can compromise quality and increase grid-to-grid variability. In this work we describe slight modifications to an existing protocol for coating cryoEM grids with a monolayer of graphene30.
Critical steps within this protocol include coating CVD graphene with MMA, dissolution of the CVD graphene copper substrate in APS, and the application of graphene to cryoEM grids. We modified the original protocol to minimize manual manipulations of the coated grids by exchanging solvents within the same Petri dish, instead of individually handling and transferring grids into a new solvent container for each wash step, thereby aiming to increase the yield of intact, high quality, graphene-coated grids. While we strove to reduce manipulations of graphene coated grids to a minimum, we acknowledge that the manual application of individual graphene squares to cryoEM grids is inherently challenging, and that some grid-to-grid variability is expected.
Graphene-coated grids typically require UV/ozone treatment to render the surface hydrophilic for sample application. The duration of UV/ozone treatment and the time elapsed after treatment and prior to plunging can impact grid hydrophilicity and ultimately sample quality. In addition to the grid fabrication protocol, we describe a technique for assessing grid hydrophilicity following UV/ozone treatment. In the procedure, the surface contact angle of an applied sample is used as an indicator of the hydrophilic character of the coated grid20,44. Designs are provided to inexpensively 3D-print a custom grid imaging mount that utilizes a simple cell phone camera to estimate the surface contact angle.
Finally, we describe results obtained by employing this protocol to determine the 2.7 Å cryoEM structure of the catalytically inactive RNA-guided DNA endonuclease, S. pyogenes Cas9 in complex with sgRNA and target DNA40. In the absence of graphene, no particles were observed in foil holes at the complex concentrations used (250 ng/µL). In contrast, graphene-coated grids bore particles at high density, enabling facile 3D-reconstruction of a high-resolution map from 2,961 micrographs. Taken together, these data highlight the value of applying graphene monolayers to cryoEM grids for single particle analysis.
The authors have nothing to disclose.
Specimens were prepared and imaged at the CryoEM Facility in MIT.nano on microscopes acquired thanks to the Arnold and Mabel Beckman Foundation. Contact angle imaging devices were printed at the MIT Metropolis Maker Space. We thank the laboratories of Nieng Yan and Yimo Han, and staff at MIT.nano for their support throughout the adoption of this method. In particular, we extend our thanks to Drs. Guanhui Gao and Sarah Sterling for their insightful discussions and feedback. This work was supported by NIH grants R01-GM144542, 5T32-GM007287, and NSF-CAREER grant 2046778. Research in the Davis lab is supported by the Alfred P. Sloan Foundation, the James H. Ferry Fund, the MIT J-Clinic, and the Whitehead Family.
250 mL beaker (3x) | Fisher | 02-555-25B | |
50 mL beaker (2x) | Corning | 1000-50 | |
Acetone | Fisher | A949-4 | |
Aluminum foil | Fisher | 15-078-292 | |
Ammonium persulfate | Fisher | (I17874 | |
Coverslips 50 mm x 24 mm | Mattek | PCS-1.5-5024 | |
CVD graphene | Graphene Supermarket | CVD-Cu-2×2 | |
easiGlow discharger | Ted-Pella | 91000S | |
Ethanol | Millipore-Sigma | 1.11727 | |
Flat-tip tweezers | Fisher | 50-239-60 | |
Glass cutter | Grainger | 21UE26 | |
Glass petri plate and cover | VWR | 75845-544 | |
Glass serological pipette | Fisher | 13-676-34D | |
Grid Storage Case | EMS | 71146-02 | |
Hot plate | Fisher | 07-770-108 | |
Isopropanol | Sigma | W292907 | |
Kimwipe | Fisher | 06-666 | |
Lab scissors | Fisher | 13-806-2 | |
Methyl-Methacrylate EL-6 | Kayaku | MMA M310006 0500L1GL | |
Molecular grade water | Corning | 46-000-CM | |
Negative action tweezers (2x) | Fisher | 50-242-78 | |
P20 pipette | Rainin | 17014392 | |
P200 pipette | Rainin | 17008652 | |
Parafilm | Fisher | 13-374-12 | |
Pipette tips | Rainin | 30389291 | |
Quantifoil grids with holey carbon | EMS | Q2100CR1 | |
Spin coater | SetCas | KW-4A | with chuck SCA-19-23 |
Straightedge | ULINE | H-6560 | |
Thermometer | Grainger | 3LRD1 | |
UV/Ozone cleaner | BioForce | SKU: PC440 | |
Vacuum desiccator | Thomas Scientific | 1159X11 | |
Whatman paper | VWR | 28297-216 |