Preparation of Nucleosome Core Particles Complexed with DNA Repair Factors for Cryo-Electron Microscopy Structural Determination

Summary

This protocol outlines in detail the preparation of nucleosomal complexes using two methods of sample preparation for freezing TEM grids.

Abstract

DNA repair in the context of chromatin is poorly understood. Biochemical studies using nucleosome core particles, the fundamental repeating unit of chromatin, show most DNA repair enzymes remove DNA damage at reduced rates as compared to free DNA. The molecular details on how base excision repair (BER) enzymes recognize and remove DNA damage in nucleosomes have not been elucidated. However, biochemical BER data of nucleosomal substrates suggest the nucleosome presents different structural barriers dependent on the location of the DNA lesion and the enzyme. This indicates the mechanisms employed by these enzymes to remove DNA damage in free DNA may be different than those employed in nucleosomes. Given that the majority of genomic DNA is assembled into nucleosomes, structural information of these complexes is needed. To date, the scientific community lacks detailed protocols to perform technically feasible structural studies of these complexes. Here, we provide two methods to prepare a complex of two genetically fused BER enzymes (Polymerase β and AP Endonuclease1) bound to a single-nucleotide gap near the entry-exit of the nucleosome for cryo-electron microscopy (cryo-EM) structural determination. Both methods of sample preparation are compatible for vitrifying quality grids via plunge freezing. This protocol can be used as a starting point to prepare other nucleosomal complexes with different BER factors, pioneer transcription factors, and chromatin-modifying enzymes.

Introduction

Eukaryotic DNA is organized and compacted by histone proteins, forming chromatin. The nucleosome core particle (NCP) constitutes the fundamental repeating unit of chromatin that regulates accessibility to DNA-binding proteins for DNA repair, transcription, and replication¹. Although the first X-ray crystal structure of the NCP was first solved more than two decades ago² and many more structures of the NCP have been published since^3,4,5,6, DNA repair mechanisms in nucleosomal substrates have not yet been delineated. Uncovering the molecular details underlying DNA repair in chromatin will require structural characterization of the participating components to understand how local structural features of the NCP regulate DNA repair activities. This is particularly important in the context of base excision repair (BER) given that biochemical studies with BER enzymes suggest unique DNA repair mechanisms in nucleosomes that are dependent on enzyme-specific structural requirements for catalysis and the structural position of the DNA lesion within the nucleosome^{7,8,9,10,11,12,13}. Given that BER is a vital DNA repair process, there is considerable interest to fill in these gaps while also establishing a starting point from which other technically feasible structural studies involving relevant nucleosomal complexes can be carried out.

Cryo-EM is rapidly becoming the method of choice for solving the three-dimensional (3D) structure of complexes whose large-scale preparation of homogeneous sample is challenging. Although, the design and purification of NCPs complexed with a DNA repair factor (NCP-DRF) will likely necessitate tailored optimization, the procedure presented here to generate and freeze a stable NCP-DRF complex provides details on how to optimize the sample and cryo-EM grid preparation. Two workflows (not mutually exclusive) shown in Figure 1, and the specific details in the protocol identify critical steps and provide strategies for optimizing these steps. This work will propel the chromatin and DNA repair field in a direction where complementing biochemical with structural studies becomes technically feasible to better understand the molecular mechanisms of nucleosomal DNA repair.

Protocol

1. Assemble nucleosome core particles via salt-gardient dialysis

NOTE: The preparation of nucleosome core particles using recombinant histone proteins for structural studies has been extensively described in detail by others^14,15,16. Follow the purification of recombinant X. laevis histones and histone octamer assembly described by others^14,15, and assemble the nucleosomal substrate as described below.

1. Purchase three oligonucleotides (listed in Table 1) at the indicated scale: UND-197mer, 161mer, 35mer.
   1. PAGE purify ultramers (197mer and 161mer) as described¹⁷.
      1. Anneal equimolar amounts of oligonucleotides in 1x annealing buffer (10 mM Tris-Cl, pH 8.0, 50 mM NaCl). For example, mix 40 µL of 161-mer [166.7 µM]; 8 µL of 35-mer [292.4 µM]; 16.8 µL of 197-mer comp strand [408.2 µM]; 10 µL of 10x anneal buffer and 10.9 µL of dH₂O.
         NOTE: It is critical to control the annealing reaction with a temperature gradient. See Table 2 for the details on the temperature gradient.
2. Perform small-scale reconstitutions (scale by a factor of 1/57 of the example shown below for a final volume of 50 µL) to determine the ratio of DNA to histone octamer (typical starting ratios of DNA:octamer are as follows: 1:0.9, 1:1.1, 1:1.2, 1:2; shown in Figure 2A); it is critical to determine this ratio empirically with small-scale reconstitutions.
   1. After this ratio is determined, prepare a large-scale reconstitution. For example, mix 32.9 µL of DNA-197 bp [99.4 µM]; 1647.1 µL of TE (1x); 1120 µL of 5 M NaCl, and 62.3 µL of WT histone octamer [2.56 µM].
3. Make dialysis buffers: RB_low and RB_high as described in Table 3 and Table 4, respectively; set up the reconstitutions as previously described¹⁴.
   1. Transfer the dialysis tube into the beaker containing RB_high, and allow for dialysis to proceed, with gentle stirring, for 16 h or until 2 L RB_low has been transferred into the waste beaker.
4. Transfer the dialysis tube into a 1 L beaker containing RB_50mM buffer (Table 5) and allow the sample to dialyze for at least 3 h or overnight.
   1. Evaluate reconstitution efficiency on a 6% polyacrylamide nondenaturing gel, ensuring there is less than 10% free DNA and a single NCP band; store NCPs at 4 °C. See Figure 2A (lane 5) and Figure 2B for representative optimal reconstitution results.

2. Prepare NCP-DNA repair factor complex (NCP-DRF)

1. Purification by preparative-gel electrophoresis
   NOTE: This method is the most labor-intensive (of the two described), and while it is effective at removing high molecular weight (HMW) species (Figure 6A) because of the high dilution factor, it can lead to some disassembly as shown in Figure 7A (NCP prep cell out lane), releasing DNA. However, up to 25% free DNA is still compatible with cryo-EM studies. Because of the high dilution, use this method to purify the NCP only, rather than the whole complex.
   1. Prepare 70 mL of 6% nondenaturing polyacrylamide gel solution in 0.2x TBE, using polyacrylamide gel solution 37.5:1, acrylamide:bisacrylamide. Pour a cylindrical gel with outer radius of 28 mm and polymerize overnight.
   2. Assemble the preparative gel running apparatus, using a dialysis membrane with a 6-8 kDa MW cutoff. Use 0.25x TBE as the running buffer and 1x elution buffer (50 mM KCl, 10 mM HEPES, pH 7.5). Pre-run the cylindrical gel at a constant 12 W for 1 h, and collect fractions running the peristaltic pump at a flow rate of approximately 1 mL/min.
      NOTE: If a UV detector is not available to identify fractions containing DNA, a screening experiment with ³²P-NCP can be performed to identify the fractions of interest. Keeping all conditions the same, unlabeled NCPs can then be purified without a UV detector.
   3. Concentrate the 2.8 mL of NCP to 250 µL using a centrifugal filter (MW cutoff 30 kDa), and load onto the preparative gel and electrophorese for a total of 6 h. At 2 h, when the xylene cyanol has ran out of the gel, start collecting 1.5 mL fractions.
      NOTE: At 2.5 h, fractions will be clear of xylene cyanol; the DNA (197mer) elutes at approximately 3-3.5 h, and the NCP is expected to elute at 4.5-5 h mark.
   4. Analyze fractions of interest on a 6% nondenaturing polyacrylamide gel. Pool fractions containing the NCP, and immediately concentrate with a centrifugal filter (MW cutoff 30 kDa) to 1 mg/mL. The next day, evaluate the quality of the NCP on a 6% polyacrylamide nondenaturing gel before complexing with the DNA repair factor (DRF) of interest.
   5. Incubate the NCP and the DRF of interest on ice for 15 min using an optimal buffer for the specific complex. For example, mix 1,000 µL of NCP [1.2 µM]; 260 µL of 5x binding buffer (250 mM HEPES, pH 8, 500 mM KCl, 25 mM MgCl₂), 22 µL of 3 M KCl, and 18 µL of MBP- Pol β-APE1 [338 µM].
      NOTE: The temperature and ionic strength at which the complex is made may need optimization for different complexes.
   6. Add glutaraldehyde (freshly opened; EM grade) to a final concentration of 0.005%. Therefore, to this reaction, add 26.8 µL of 0.25% glutaraldehyde with 13.2 µL of dH₂O. Mix well, and incubate at room temperature for 13 min.
   7. Quench with 1 M Tris-Cl, pH 7.5, to a final concentration of 20 mM Tris-Cl, pH 7.5. Concentrate to ~50 µL and exchange the buffer with 1x freezing buffer: 50 mM KCl, 10 mM HEPES, pH 7.5 using a desalting column.
      CAUTION: Glutaraldehyde can cause severe skin burns and eye damage; it is harmful if swallowed, and it is toxic if inhaled. Wear a lab coat, goggles, gloves, and handle glutaraldehyde in a fume hood and wear a mask.
      NOTE: It is critical to perform the crosslinking reaction in a large volume with the NCP at this lower concentration. Previous attempts of crosslinking, even at this concentration of glutaraldehyde, but at a 10x higher concentrated NCP led to aggregation of the sample and over-crosslinking as indicated by SDS PAGE and size exclusion chromatography. These grids had no discernable particles with just clumps (Figure 5D,E).
   8. Determine the absorbances at OD₂₈₀ and OD₂₆₀ and concentrate further if needed to reach 1.3-3 mg/mL, based on OD₂₈₀ (typical ratio of OD₂₆₀/ OD₂₈₀= 1.7-2 yields good particles). For this small volume of 50 µL, the best method to reduce sample loss is to make a dialysis button with a lid of a 1.7 mL tube covered by a dialysis membrane (6-8 kDa MW cutoff), cutting the bottom of the tube and using the rim of the tube to seal the membrane on top of the lid.
   9. Place the dialysis button containing the sample on top of a polyethylene glycol bed, with the membrane facing down (check progress every 2 min). This method is preferred where the concentration needed is less than or equal to 2.5-fold to avoid diluting or losing the sample in the concentrator. It is critical to immediately prepare cryo-EM grids of the complex after this step.
2. Purification by size exclusion chromatography
   1. The day before freezing, wash and equilibrate a size exclusion column with 60 mL of dH₂O, followed by 80 mL of freezing buffer (50 mM KCl, 10 mM HEPES, pH 8) overnight at a rate of 0.4 mL/min. Using NCPs straight after reconstitution, prepare the same complex using 2.5x greater amounts as follows: mix 2,500 µL of NCP [1.2 µM]; 650 µL of 5x binding buffer; 45 µL of MBP-Pol β-APE1 [338 µM] and 55 µL of 3M KCl.
   2. Incubate the mixture on ice without crosslinker for 15 min. Then, add glutaraldehyde to a final concentration of 0.005% as described in the preparative gel method. In this case, however, concentrate the complex in a pre-equilibrated (with freezing buffer) centrifugal filter to approximately 120 µL.
   3. Immediately analyze the peak fractions and concentrate those fractions containing the histones and MBP-Pol β-APE1 as indicated previously. See Figure 5A,B,C for successful results using the size exclusion method and Figure 7 using both methods.

3. Freeze nucleosomal complex

1. After turning on the plunge freezer and filling the humidifier with 50 mL of dH₂O, set the chamber temperature to 22 °C and H_R (humidity) to 98%. Place the ethane cup and liquid nitrogen cup (containing a labeled grid box) in their respective space holders.
2. Cover the ethane cup with the ethane lid dispenser, and carefully pour liquid nitrogen (LN2) over it, while also filling the LN₂ cup with LN₂. When the LN₂ level has stabilized and reaches 100% and a temperature of -180 °C, carefully open the ethane valve and fill the ethane cup until it forms a bubble on the clear lid. Remove the ethane dispenser.
3. Place two filter papers onto the blotting device and secure them with a metal ring. Go to the set up and use the following parameters for blotting: 0 pre-blot, 3 s blot, 0 post-blot; select A-plunge and click on OK.
4. On the main screen, click on Load Forceps, and load them with a grid with the carbon/application side facing toward the left (prepared as before). Calibrate the forceps adjusting the Z-axis to ensure a solid blot. Click on Lower Chamber and apply 3 µL of the complex (1.3-3 mg/mL; OD₂₈₀). Click on Blot/A-plunge. This will rotate the grid to blot from the front and will plunge freeze it.
5. Transfer and store the grid in the grid box in the LN₂ chamber. When all four grids have been frozen and placed in the grid box, rotate the lid to neutral position, where all the grids are covered by the lid, and tighten the screw. Grids can be stored in LN₂ until screening is initiated.

4. Screen grids

1. Before loading the samples in the microscope, place the vitrified grids on a ring and secure them using a C-clip. Perform this clipping process under LN₂ in a humidity-controlled room, to avoid ice contamination.
2. When loading the microscope, insert the grids in a 12-slot cassette. Shuttle the cassette into a nanocab capsule and load it into the autoloader. The autoloader robotic mechanism initiates the process.
3. Transfer the grid from the cassette to the microscope stage. Adjust the stage to eucentric height by wobbling the stage 10° and simultaneously moving the Z-height until minimal planar shift is observed in the images.
4. Once the eucentric height is reached, start imaging. First, obtain the atlas by taking a 3 x 3 montage image of the grid, in which each montage is taken at 62x magnification. Pick three squares of varying sizes; take the eucentric height, and then image at 210x magnification.
5. Once a square is imaged, pick one hole each from the edge, center, and in between within the squares. Image each hole at 2600x magnification.
6. Before taking this high magnification image, the autofocusing occurs at an offset of the imaging area. Take the high magnification image from the center of the hole at 36,000x magnification and at 7.1 s exposure time, 60 frames, 1.18-pixel size, and 3 µm defocus. See representative images of screening in Figure 5, Figure 6, and Figure 7.

Representative Results

Properly assembled NCPs (Figure 2) were used to make a complex with a recombinant fusion protein of MBP-Polβ-APE1 (Figure 3). To determine the ratio of NCP to MBP-Polβ-APE1 to form a stable complex, we performed electrophoretic mobility shift assays (EMSA) (Figure 4), which showed a singly shifted band of the NCP with 5-fold molar excess of MBP-Polβ-APE1. During the optimization of making this complex, crosslinking with glutaraldehyde was critical to prevent the NCP from falling apart. Initially, assembly of the complex of NCP-Polβ-APE1 was performed in a smaller volume, at approximately 10 µM NCP, using 0.005% final glutaraldehyde concentration. Under these conditions, the sample was overly crosslinked (Figure 5D-E) and resulted in aggregates without discernable individual complexes. Crosslinking at approximately 10-fold diluted NCPs (1.2 µM NCP) resulted in significantly reduced aggregation and improved particle stability (Figure 5A-C). Figure 6 illustrates that both methods shown in Figure 1 can be combined with the preparative gel improving the quality of the NCPs when the NCP contains a significant amount of high molecular weight (HMW) aggregates (Figure 6A), followed by the purification of the complex via size exclusion. Although this showed great success in the preparation of the sample (Figure 5B-C) and grids, yielding stable particles (Figure 6D), the suboptimal formation of the complex (Figure 6A) yielded a 3D map of the NCP alone (data not shown). These results show that both methods (size exclusion and preparative gel) can be used independently to generate stable complexes (Figure 7A-C). Indeed, data have been collected from both grids and show almost identical 2D classes (Figure 7D) and 3D maps (Figure 7E) at approximately 3.2 Å resolution.

Figure 1: Workflow of the two methods presented in this protocol to prepare and freeze an NCP-Polβ-APE1 complex via 1) preparative gel and 2) size exclusion. Although not shown here, these methods can be combined (Figure 6), starting with (1) and purifying the concentrated complex from (1) using a sizing column. This will require doubling the starting material of NCPs. Please click here to view a larger version of this figure.

Figure 2: Representative NCP reconstitutions. (A) DNA was titrated with increasing amounts of histone octamer at ratios of 1:0.9, 1:1.1, 1:1.2, 1:2, DNA:octamer. (B) Duplicate large-scale assembled NCPs (1:2, DNA:octamer). Reconstitutions were electrophoresed in a 6% nondenaturing polyacrylamide gel, followed with sybr green I staining. Please click here to view a larger version of this figure.

Figure 3: Schematic of genetically fused Polβ-APE1 with MBP at its N-terminal. MBP-Polβ-APE1 fusion protein was enzymatically characterized for both activities (data not shown). Please click here to view a larger version of this figure.

Figure 4: MBP-Polβ-APE1 binding to NCPs. Nucleosomal substrate (100 nM) was incubated with increasing amounts of MBP-Polβ-APE1 for 15 min on ice. Bound and unbound NCP were separated in a 6% polyacrylamide nondenaturing gel, followed by sybr green I staining. Please click here to view a larger version of this figure.

Figure 5: Analysis of successful and unsuccessful crosslinking results. The elution profile of the complex is shown in (A) with the peak of interest labeled. Fractions, denoted by "F", collected from this elution were analyzed in a 16% tricine polyacrylamide gel and shown in (B). (C) Sample from (B) was frozen using holey carbon support grids and imaged using the 200 kV field emission cryo-transmission electron microscope. Corresponding unsuccessful experiments are shown in (D), (E), and (F). Note these unsuccessful experiments contain the fusion protein without MBP. HO_c and HO_D denote histone octamer concentrated and diluted, respectively; RT denotes room temperature; "X" corresponds to crosslinking with glutaraldehyde. Scale bars = 100 nm (C, F). Please click here to view a larger version of this figure.

Figure 6: Preparation of NCP-Polβ-APE1 complex using the two methods in tandem. (A) Unbound and bound (1:1, NCP- Polβ-APE1) were analyzed in 6% polyacrylamide nondenaturing gel (note that at this ratio only 50% is bound). (B) Elution profile of NCP-Polβ-APE1 complex from a sizing column. (C) Analysis of eluted fractions in a 16% tricine polyacrylamide gel; fractions were pooled and concentrated, as indicated. (D) Sample was frozen and imaged as described in Figure 5B. Data processing did not show an additional density corresponding to Polβ-APE1 (not shown). "X" in lanes 5 and 6 denotes crosslinking with glutaraldehyde as indicated in the protocol. Scale bar = 100 nm (D). Please click here to view a larger version of this figure.

Figure 7: Analysis of NCP-MBP-Polβ-APE1 complex. (A) Samples from the different steps in the methods illustrated in Figure 1 were electrophoresed in a 6% polyacrylamide nondenaturing gel. Notice that complex formation is reproducible by the two methods. (B,C) Representative grid holey carbon support grid images taken using a 300 kV transmission electron microscope where the red boxes highlight different orientations of the complex (some of which correspond to the 2D classes shown in (D). (E) Data processed from (C) show a 3D map at 3.2 Å resolution of the NCP-MBP-Polβ-APE1 complex.). Scale bars = 100 nm (B, C). Please click here to view a larger version of this figure.

Oligonucleotide name	Sequence (5'-->3')			Purification scale	Method of purificaiton	Number of vials ordered
UND complementary strand	TGATGGACCCTATACGCGGCCGCCCTGGAGAAT CCCGGTGCCGAGGCCGCTCAATTGGTCGTAGA CAGCTCTAGCACCGCTTAAACGCACGTACGCGC TGTCCCCCGCGTTTTAACCGCCAAGGGGATTAC TCCCTAGTCTCCAGGCACGTGTCAGATATCAAC ATCCTGTGCATGTATTGAACAGCGACCTTGCCG			20 nmole	PAGE purified	6
161mer	/5Phos/GATATCTGACACGTGCCTGGAGACTAGG GAGTAATCCCCTTGGCGGTTAAAACGCGGGGG ACAGCGCGTACGTGCGTTTAAGCGGTGCTAGA GCTGTCTACGACCAATTGAGCGGCCTCGGCA CCGGGATTCTCCAGGGCGGCCGCGTATAGGG TCCATCA			20 nmole	PAGE purified	4
35mer	CGGCAAGGTCGCTGTTCAATACATGCACAGG ATGT			250 nmole	HPLC	1

Table 1: DNA substrate. Three oligonucleotides were annealed to generate the dsDNA substrate. The underlined sections are 25 bp of linker DNA flanking the 147 bp 601 DNA positioning sequence containing a single nucleotide gap.

Step	Temperature (°C)	Time (min)/step
1	95	10
2-5	90, 85, 80, 75	5
6-41	69 decrease by 1	3
42	4	hold

Table 2: Annealing temperature gradient. Oligos were incubated in a thermocycler at these temperatures to generate the dsDNA substrate.

Components (stock)	Volume/amount	Final concentration
3M KCl	167 mL	0.250 M
1 M HEPES, pH 8.0	20 mL	10 mM
0.5 M EDTA	4 mL	1 mM
1 M DTT	2 mL	1 mM
CHAPS	2 g	1.6 mM
QS with dH2O to make 2L

Table 3: RB_low reconstitution buffer. Instructions to make 2 L as previously reported¹⁴, except 1.6 mM CHAPS was added.

Components (stock)	Volume/amount	Final concentration
3M KCl	267 mL	2 M
1 M HEPES, pH 8.0	4 mL	10 mM
0.5 M EDTA	800 μL	1 mM
1 M DTT	400 μL	1 mM
CHAPS	0.4 g	1.6 mM
QS with dH2O to make 400 mL

Table 4: RB_high reconstitution buffer. Instructions to make 400 mL as previously reported¹⁴, except 1.6 mM CHAPS was added.

Components (stock)	Volume	Final concentration
3M KCl	16.7 mL	50 mM
1 M HEPES, pH 8.0	10 mL	10 mM
0.5 M EDTA	2 mL	1 mM
1 M DTT	1 mL	1 mM
QS with dH2O to make 1L

Table 5: RB_50mM reconstitution buffer. Instructions to make 1 L of reconstitution buffer compatible for freezing.

Discussion

A specific protocol for purifying the DNA repair factor will be dependent on the enzyme of interest. However, there are some general recommendations, including the use of recombinant methods for protein expression and purification¹⁸; if the protein of interest is too small (<50 kDa), structure determination by cryo-EM had been nearly impossible until more recently through the use of fusion systems¹⁹, nanobody-binding scaffolds²⁰, and optimizing imaging strategies²¹.

It is quite common to obtain poor quality grids in the initial stages. Initial attempts to determine NCP homogeneity and stability without any crosslinker, using uranyl acetate and carbon-coated copper grids for negative staining, showed stable, nicely dispersed NCPs with minimal aggregation. However, once this sample was vitrified (at 20x higher concentration) in cryo-EM grids, no particles were detected. Sample optimization to keep the NCPs from falling apart or aggregating was the biggest hurdle. Additionally, two methods of preparation are described separately; however, they are not mutually exclusive and can be used back-to-back by first purifying the NCP via a preparative gel, complexing it with the repair factor, and purifying the complex with a size exclusion column. This requires twice as much material for the initial step, and it is also more time-consuming, but helpful for difficult samples (Figure 6).

Finding a starting point for the concentration of crosslinker was difficult due to the lack of consistency in methods of crosslinking and the amounts of crosslinker used by others^22,23. Attempts to replicate the less-controlled crosslinking reaction on the bench top described by Anderson et al. failed to give positive results, and the complex was overly crosslinked. This is likely due to the source of glutaraldehyde (freshly opened ampule compared to long-term stored glutaraldehyde). Because this level of detail is often omitted in non-methods, research articles, it is difficult to replicate the exact conditions as a starting point. Optimal results were obtained when the complex was crosslinked at a concentration of 1.2 µM in 150 mM KCl ionic strength compared to crosslinking the complex at 10 µM NCP, at the same glutaraldehyde concentration of 0.005% for 13 min at room temperature. The crosslinking efficiency is also likely dependent on the complex that is being formed as some chromatin-interacting factors may induce/require greater destabilization influencing where the crosslinking will occur and the overall effect on NCP stability. Therefore, while optimization of these conditions is inevitable, finding critical steps such as optimal NCP and salt concentration during the crosslinking reactions will be critical.

Another important parameter was the ratio of NCP:DRF that yielded enough molecules on the grid containing the DRF bound to the NCP. Indeed, this parameter has also been identified by others as one that is critical to optimize for structural studies¹⁶ given that even for cryo-EM homogeneity of the way the DRF binds to the NCP is important for the 3D map. Not optimizing this may yield unbound or heterogenous, non-specific binding. When a 1:1 molar ratio was used, yielding only 50% the NCP bound by the DRF (Figure 6), the 3D map did not show any additional density corresponding to the DRF. Therefore, it is important to determine the ratio that yields a single band at the same location on an EMSA and shows a single peak on the chromatogram of a sizing column, corresponding to the complex. Several parameters for freezing of the grids, including the blot time, added several rounds of optimization.

Taken together, this protocol provides detailed instructions for the first time on two different methods to prepare nucleosomal complexes for cryo-EM structural determination. The baseline parameters for the grid preparation provided can be used as a starting point. And while not a single method works equally well for all macromolecular complexes, focusing on optimizing the parameters described in this protocol will narrow down the strategy for tailored optimization of other nucleosomal complexes.

Materials

Name	Company	Catalog Number	Comments
1 M HEPES; pH 7.5	Thermo Fisher Scientific	15630080
1 M MgCl2	Thermo Fisher Scientific	AM9530G
10x TBE	Bio-rad	1610733
25% glutaraldehyde	Fisher Scientific	50-262-23
3 M KCl	Thermo Fisher Scientific	043398.K2
491 prep cell	Bio-rad	1702926
Amicon Ultra 15 centrifugal filter (MW cutoff 30 kDa)	Millipore Sigma	Z717185
Amicon Ultra 4 centrifugal filter (MW cutoff 30 kDa)	Millipore Sigma	UFC8030
AutoGrid Tweezers	Ted Pella	47000-600
Automatic Plunge Freezer	Leica	Leica EM GP
C-1000 touch thermocycler	Bio-rad	1851148
C-clips and rings	Thermo Fisher	6640--6640
Clipping station	SubAngostrom	SCT08
Dialysis Membrane (MW cufoff 6-8 kDa)	Fisher Scientific	15370752
Diamond Tweezers	Techni-Pro	758TW0010
dsDNA	Integrated DNA techonologies	N/A
FEI Titan Krios	Thermo Fisher	KRIOSG4TEM
FPLC purification system	AKTA Pure	29018224
Fraction collector Model 2110	Bio-rad	7318122
Glow Discharge Cleaning System	Ted Pella	91000S
Grid Boxes	SubAngostrom	PB-E
Grid Storage Accessory Pack	SubAngostrom	GSAX
Liquid Ethane	N/A	N/A
Liquid Nitrogen	N/A	N/A
Minipuls 3 peristaltic two-head pump	Gilson	F155008
Nanodrop	Thermo Fisher Scientific	ND-2000
Novex 16%, Tricine, 1.0 mm, Mini Protein Gels	Thermo Fisher Scientific	EC6695BOX
Pipetman	Gilson	FA10002M
Pipette tips (VWR) Low Retention	VWR	76322-528
Polyacrylamide gel solution (37.5:1)	Bio-rad	1610158
polyethylene glycol (PEG)	Millipore Sigma	P4338-500G
Pur-A-lyzer Maxi 3500	Millipore Sigma	PURX35050
Purified recombinant DNA repair factor	N/A	N/A
R 1.2/1.3 Cu 300 mesh Grids	Quantifoil	N1-C14nCu30-01
Recombinant histone octamer	N/A	N/A
Spring clipping tools	SubAngostrom	CSA-01
Superdex 200 column 10/300	Millipore Sigma	GE28-9909-44
Transmission Electron Microscope	Thermo Fisher	Talos Arctica 200 kV
Tweezers Assembly for FEI Vitrobot Mark IV-I	Ted Pella	47000-500
UltraPure Glycerol	Thermo Fisher Scientific	15514011
Vitrobot	Thermo Fisher	Mark IV System
Whatman Filter paper (55 mM)	Cytiva	1005-055
Xylene cyanol	Thermo Fisher Scientific	440700500
Zeba Micro Spin Desalting Columns, 7K MWCO, 75 µL	Thermo Fisher Scientific	89877