Isolation of Disease-relevant p53 for Cryo-Electron Microscopy Analysis

Tumor suppressor protein TP53 (p53) plays a multi-faceted role in all cells of the human body. Sadly, mutations in the TP53 gene are involved in nearly ~50% of tumors, spurring erratic cell growth and disease progression. Until recently, structural information for p53 remained incomplete and there are limited studies on native p53 produced in human tumors. Here, we present a highly reproducible and effective protocol to extract, enrich, and purify native p53 protein assemblies from cancer cells for downstream structural studies. This method does not introduce purication tags into the p53 gene and maintains naturally occurring modications. In conjunction with cryo-Electron Microscopy techniques, we determined new structures for p53 monomers (~50 kDa) and tetramers (~200 kDa) at spatial resolutions of ~4.8 Å and ~7 Å, respectively.1 These models revealed new insights for exible regions of p53 along with biologically-relevant ubiquitination sites. Combining biochemical and structural imaging protocols, we aim to build a better understanding of native p53’s impact in cancer formation. the DNA fragment within the tetramer assembly showed features consistent with a double stranded break in the middle of the helical strand. These results suggest that we have captured ubiquitinated-p53 tetramer assemblies primed for DNA repair. We expect complementary biochemical experiments to shed light on the mechanistic details of tetramer formation and function. Overall, these procedures highlight the versatility of the protocol and its potential impact for advancing structural studies.


Introduction
The tumor suppressor protein, TP53 (p53), is critical for cell health and is often deregulated in cancer. Its primary role is to detect stress signals and assist with DNA damage response in the cell's nucleus. As such, mutations in the p53gene can in uence tumor progression and reoccurrence. 2,3 The primary sequence of the protein consists of three main regions including the N-terminal domain (NTD), DNAbinding domain (DBD), and the C-terminal domain (CTD). Crystallographic studies have generally focused on the DBD of p53 as this region is the most stable and contains multiple cancer-related mutations. 4 Structural information for the NTD and CTD remains limited due to their exible nature. 5 Numerous biochemical studies have shed light on functional aspects of each domain and identi ed prominent binding partners. 3,6 A larger framework is still needed, however, to truly grasp how mutations in p53 disrupt in its structure and function in cells. To advance our understanding of native p53 assemblies, we developed new tools described here and in our recently publish work. 1 Cryo-electron microscopy (EM) is an optimal method to study protein structures that cannot be easily deciphered using other techniques. Advancements in hardware and equipment have contributed to the "resolution revolution" in the EM eld. To help address specimen-related challenges, here we present a highly effective protocol to isolate native p53 complexes from human cancer cells suitable for cryo-EM analysis. The samples revealed new structures of p53 monomer (~50 kDa) and functional tetramer (~200 kDa). 1 These higher-resolution structures provided a putative new model for the NTD and de nes ubiquitination sites on the protein while bound to DNA. Equally important, our protocol may be used to study native p53 structures from many different cancer cell lines.
Overview of the Procedure Immobilized metal a nity chromatography (IMAC) is a commonly used protein enrichment technique that employs Nickel-Nitrilotriacetic acid (Ni-NTA) matrices. This technique takes advantage of the strong binding a nity between the chelated Ni-NTA functional group and a stretch of continuous histidine residues found in recombinant or native proteins. Ni-NTA agarose beads or other IMAC puri cation systems can also have a more versatile use in protein puri cation. For example, proteins that are rich in post-translational modi cations (PTMs), such as phosphorylation, are known to interact with metal cations 7,8 . Heavy metal uptake in cells is a grave danger to phosphate-rich genetic material and is the cause of acute poisoning, environmental toxicity, and even cancer 9 . We can exploit the phosphorylationrich nature of p53 to extract it from the nuclear material of human cancer cells and enrich the native protein through Ni-NTA chromatography. This simple, but effective protocol may serve as a new means to investigate authentic p53 structures from many different cancer cells. Here, we provide detailed information for the extraction, puri cation and enrichment steps of the protein from glioblastoma 2. To harvest the cells, collect and wash cells in ice-cold PBS. If cells pellets will be used at a different time, use PBS with 1% HALT phosphatase inhibitor to wash cells before centrifuging and removing supernatant.
3. Flash freeze in liquid nitrogen and store at -80°C.
B. Cytoplasmic and Nuclear Cell Extraction 1. Refer to the table in the NE-PER kit manual to nd volumes for different cell pellet sizes. This protocol was created and re ned for cell pellets equal or larger than 100 µL; for the rest of this protocol, we will refer to those appropriate quantities. To the frozen cell pellet, add appropriate volume of ice cold CER I solution and vortex at highest setting for 15 seconds or until the pellet is fully suspended.
3. Add appropriate volume of the kit's ice cold CER II solution to the suspended cell pellet and once again vortex at the highest setting for 5 seconds. Incubate for 1 minute on ice.

Vortex the cells again for 5 seconds.
5. Centrifuge for 5 minutes at 17,000xg at 4°C. Collect the supernatant (cellular extract (CE); ~1 mL) and store in pre-chilled tubes. 6. The remaining pellet contains the cell nuclei and where the vast majority of p53 is located. To the pellet, add appropriate volume of ice-cold NER. Carefully resuspend with a cut pipette tip until resuspension is accomplished. 7. Vortex on the highest setting for 15 seconds and incubate on ice for 10 minutes. Repeat for a total of ve times.

Anticipated Results
A. Veri cation of sample identity Even though p53 tends to elute most frequently in fractions 2 and 3, depending on the mutations and PTMs on the molecule, it may elute in a different fraction. Veri cation through SDS-PAGE analysis in conjunction with Coomassie blue staining and western blotting is needed to determine which fractions yield the best protein sample for downstream structural studies.

B. Visualization through cryo-EM
For cryo-EM specimen preparation, we followed the protocol detailed in the compliment paper that describes the sample application and vitri cation of Silicon Nitride (SiN) microchips. 1 Here, SiN microchips were prepared with a monolayer of lipids containing 25% Ni-NTA-chelated groups and 75% DLPC. Sample was added to the monolayer (2 µL of native p53 eluate at a concentration of 0.02 mg/mL). EM specimens were imaged using a Talos F200C EM (ThermoFisher Scienti c) operating at 200 kV under low-dose conditions (~5 electrons/A 2 /sec). Images were recorded using a CMOS camera (ThermoFisher Scienti c) with a pixel size of 14 µm at a magni cation of 92,000x. Due to the polymeric identity of p53, a homogenous mixture will not be initially evident but computation software can to differentiate particle identity.

C. Structural determination of p53
SDS-PAGE analysis determined tetramer assemblies that had migrated at ~200 kDa. We used RELION software package to perform single particle image processing to determine p53 model assemblies. Tetramer structures were easily separated using 3D classi cation after 25 iterations. Tetramer dimensions were consistent with particles in the raw images and were easily visualized in class averages. The ~7-Å reconstruction was not limited in particle orientations although some preferences were noted.
The EM structure contained 4 distinct domains. Multiple models were used to interpret each domain of the tetramer assembly. 1 Initially, we used the crystal structure of the DBD to interpret the lower half of the density map (dark blue, pdb code 2AC0, all chains) 10 . Using the I-TASSER protein server, we modeled a putative NTD with a C-score of -1.81, which had a medium con dence but still allowed us to interpret the EM map. 11 These domains were placed in the biologically-relevant density adjacent to the DBD in accordance with our previously reported p53 monomer structure. 1 Finally, the upper region of the map had additional density which was attributed to ubiquitin moieties, consistent with a lower resolution model of p53 dimers. 12 These single ubiquitin chains (8 kDa) were proximal to residue K24 in the NTD of p53. Residue K24 was previously identi ed as a site of ubiquitination during DNA repair. Remaining smaller regions of density were attributed to exible PTMs that cannot be modeled at this resolution.
Finally, the DNA fragment within the tetramer assembly showed features consistent with a double stranded break in the middle of the helical strand. These results suggest that we have captured ubiquitinated-p53 tetramer assemblies primed for DNA repair. We expect complementary biochemical experiments to shed light on the mechanistic details of tetramer formation and function. Overall, these procedures highlight the versatility of the protocol and its potential impact for advancing structural oncology studies. resin used to enrich phosphorylated-p53 from the nuclear material of U87MG cells. SDS-PAGE and western blots show and enrichment of p53 monomers (~50 kDa) and tetramers (~200 kDa). (D) Enriched p53 fractions were incubated on SiN microchips decorated with Ni-NTA substrate. These microchips can be vitri ed and imaged using cryo-EM. Scale bar is 25 nm. (Adapted from previous work)1.

Figure 2
Cryo-EM structure of the p53 tetramer assembly bound to DNA. (A) Magni ed front view of the p53 tetramer structure (white) shown in different rotational views. The map was interpreted using the Nterminal domain model (cyan) along with the full tetramer structure of the DNA-binding-domain (blue; pdb code 2A0C, all chains).10 The density map accommodates two ubiquitin monomers (yellow; pdb code, 1UBQ)13 and suggests a biologically relevant con guration for attachment. The DNA-binding domain also surrounds a broken DNA helix (dark blue). (B) Sections through the structure indicate nearly full occupancy of the map. Scale bar is 20 Å. (C) Class averages (left panel) show a variety of orientations in the ~7 Å-structure according to the gold-standard Fourier shell correlation (0.143) criteria. The EM structure was not limited in its distribution of particle views. Some preferred orientations were noted in the distribution plot. (Adapted from previous work)1.