The expression vector for the recombinant GST-tagged form of the DNA-binding domain of p53 (residues 94–312) was constructed. To prepare a plasmid for p53C-wt expression, the genomic sequence encoding the corresponding amino acid sequence was amplified with two primers designed from human adult normal kidney cDNA (BioChain Institute, Hayward, CA, United States) and placed into pGEX-6P-3 (Cytiva, Tokyo, Japan) using an In-Fusion HD Cloning Kit (Takara bio, Kusatsu, Japan) according to the manufacturer’s protocol. To prepare a plasmid for p53C-R175H expression, two complementary oligonucleotides with the mutated sequence were used as primers to introduce a mutation that replaced the amino acid encoding Arg175 with His. The protein was prepared by expression in Escherichia coli BL21 (DE3), and the cells were cultured in LB medium. Isopropyl-β-D-thiogalactopyranoside was added at an OD₆₀₀ of approximately 0.6 to a final concentration of 0.2 mM, and the cells were incubated at 20°C overnight. The harvested cells were resuspended in lysis buffer (20 mM phosphate buffer, pH 6.0, 300 mM NaCl) and disrupted by sonication. The supernatant was applied to a DEAE-Sepharose (Cytiva) column and affinity purified using Glutathione Sepharose 4 Fast Flow (Cytiva) chromatography. The GST tag was removed by HRV 3C protease on beads. The protein was subsequently purified by size exclusion chromatography using a HiLoad 26/60 Superdex 75 pg (Cytiva) in 20 mM phosphate buffer, pH 6.0, 140 mM NaCl, and 1 mM dithiothreitol (DTT). After purification, the sample was concentrated on an ultrafiltration membrane and stored at −80°C. For use, the samples were dialyzed in an assay buffer of 20 mM Tris-HCl and 1 mM DTT at pH 7.3. After dialysis, the supernatant was separated by centrifugation at 15,000 ×g at 4°C for 2 min and placed on ice until just before use.

The aggregation of p53C was monitored by fluorescence enhancement upon both ANS (Sigma-Aldrich; Merck KGaA) and ThT (Sigma-Aldrich) binding (Yoshimura et al., 2012; Adachi et al., 2015; So and Yoshimura, 2020). ANS and ThT were added simultaneously to contain 10 and 20 μM, respectively, and the concentration of the p53 protein was 4 μM. The measurement samples were prepared at room temperature within 2 min, aliquoted into a quartz cell pre-warmed with the cell holder to 37°C in an F-7000 fluorescence spectrophotometer (Hitachi High-Tech Co., Tokyo, Japan), and measured immediately using the FL Solutions program of the instrument. The fluorescence intensity was measured every 2 min for 1 h. The fluorescence wavelength was set to 485 nm, the excitation spectrum was scanned, and the values at 375 and 445 nm were extracted as the fluorescence of ANS and ThT, respectively.

The ANS/ThT values after 60 min of aggregation were measured with an EnSpire Multimode Plate Reader (PerkinElmer Inc., Waltham, MA, United States). The p53C solution at a concentration of 4 μM with 10 μM ANS and 20 μM ThT was prepared on a black 96-well plate (OptiPlate-96 F; PerkinElmer Inc.) on ice and measured before incubation to use as the blank. After incubation at 37°C for 1 h, the reaction was stopped by placing the plate on ice. The excitation wavelength was set to 375 and 445 nm for ANS and ThT, respectively, and the detection wavelength was set to 485 nm for both fluorophores. Each sample was measured three times, and the average of the measurements was used as a single datum. The samples were n = 3, and the mean and standard deviations were calculated.

p53C solution at a concentration of 4 μM was incubated on a transparent 96-well microplate (AGC Techno Glass Co., Ltd., Shizuoka, Japan) for 2 h at 37°C. Both ThT and ANS were added to a concentration of 40 μM, and observation was performed. The fluorescence images were obtained at ×10 magnification using fluorescence microscopy (IX-71; Olympus, Tokyo, Japan).

After aggregation, all samples were diluted 100 times in a buffer of 20 mM Tris-HCl and 1 mM DTT at pH7.3 to a final volume of 1 ml. Measurements on Nanosight NS300 (Malvern, United Kingdom) were executed according to the manufacturer’s software manual. The measurements were performed at 25°C. The sample solution was collected in a 1 ml disposable syringe and measured while pumping with a syringe pump. The NTA 2.3 software was used for video capture and data analysis. The data of the five measurements were averaged, and the standard deviation was calculated.

The DNA to bind p53 was prepared by introducing AGG​CAT​GCC​TAG​GCA​TGC​CT (Chen et al., 2010) into pGEX-6P-3 using an In-Fusion HD Cloning Kit. DNA of 100 bp length containing this sequence was amplified using PCR and purified using a Wizard SV Gel and PCR Clean-Up System (Promega, Madison, WI, United States). A solution of 4 μM p53C was incubated at 37°C, and a part was sampled at the appropriate time and placed on ice. When all sampling was completed, 20 ng of purified 100 bp DNA was added, and then the samples were run on a 6% TBE gel. The gel was stained with ethidium bromide and detected on a LED transilluminator (GELmieru; FUJIFILM Wako, Osaka). The band intensities were analyzed using ImageJ software.

A solution of 15 μM p53C in a buffer of 20 mM phosphate and 10% (v/v) D₂O at pH 7.3 was incubated at 37°C and placed on ice. A deuterated glucose (Cambridge Isotope Laboratories, Inc., Andover, MA, United States) was used as a glucose for NMR measurements. One-dimensional ¹H-NMR was measured at 15°C on a 900 MHz NMR spectrometer (Avance III; Bruker, Karlsruhe, Germany) equipped with a cryogenic triple-resonance probe. All NMR data were obtained using Topspin software and analyzed using NMRPipe software packages.

The fluorescence intensity of tyrosine and tryptophan to detect the structural transition of p53C was monitored on an F-7000 fluorescence spectrophotometer. The concentration of the p53 protein was 4 μM. The measurement samples were prepared at room temperature within 2 min, entered into a quartz cell pre-warmed to 37°C with the cell holder in the fluorescence spectrophotometer, and immediately measured using the FL Solutions program of the instrument. The fluorescence intensity was measured every 2 min for 1 h. The excitation wavelength was set to 280 nm, the fluorescence spectrum was acquired, and then the ratio of the value at 333 nm to the value at 305 nm was calculated.

The wild-type of p53C (p53C-wt) and R175H mutant of p53C (p53C-R175H), which is a type of hotspot mutation, were reported to aggregate (Butler and Loh, 2003). First, we monitored the aggregation of p53C using two kinds of fluorophores: ANS and ThT. The fluorescence intensity of ANS increases in hydrophobic environments and that of ThT increases by binding to cross-β sheet structures, which is characteristic of amyloid structures (Khan et al., 2015). As a result, the fluorescence intensity of both ANS and ThT increased with time; we were able to reproduce these previous reports (Figures 1A–D). We then examined the effect of the presence of KCl, a type of charged osmolyte in cells, on the aggregation of p53C-wt and p53C-R175H. The increase in the fluorescence intensity of ANS was reduced (Figures 1A,B). Although p53C-wt could not increase or decrease the fluorescence intensity of ThT (Figure 1C), p53C-R175H caused a increase in ThT fluorescence intensity (Figure 1D). There was concern that these fluorescence fluctuations were due to the ionic effect of 400 mM KCl. Thus, we examined the effect of KCl in the presence of either p53C-wt or p53C-R175H. For ANS fluorescence, approximately 180% and 140% fluorescence enhancement were observed against non-aggregated p53C-wt and p53C-R175H, respectively. In contrast, approximately 55% and 45% reduction of ThT fluorescence were observed (Supplementary Figures S2A,B). We showed the normalized fluorescence changes upon aggregation of p53C-wt and p53C-R175H monitored by ANS or ThT (Supplementary Figures S2C,D). Taking all these effects into consideration, we confirmed the tendency of the salt-dependent suppression of the ANS fluorescence and the salt-dependent enhancement of ThT fluorescence. To analyze the change in the ratio of ANS and ThT based on the concentration of KCl, we measured the increased ANS and ThT fluorescence values on the microplate for high-throughput analysis. p53C solution was aggregated for 1 h at 37°C on a 96-well microplate, and we plotted the calculated value of ThT/ANS on the vertical axis and the concentration of KCl on the horizontal axis. The results showed that the value of ThT/ANS increased in a KCl concentration-dependent manner and that the values were similar for p53C-wt and p53C-R175H (Figure 1E).

To clarify the difference in the shape of aggregates with and without 400 mM KCl, we tried to observe the aggregates stained with ANS and ThT by fluorescence microscopy. Aggregates stained with ThT were so few in both p53C-wt and p53C-R175H that they were not visible under fluorescence microscopy at 10-fold magnification, but aggregates stained with ANS were visible. Under conditions without KCl, the aggregates of p53C-R175H were larger than those of p53C-wt (Supplementary Figures S1A,B). In the 400 mM KCl condition, both p53C-wt and p53C-R175H mutants showed much less aggregation (Supplementary Figures S1C,D), which was consistent with the results of the fluorescence spectrophotometer experiments.

Fluorescence microscopy suggested that the morphologies of the p53C-wt and p53C-R175H aggregates differed and were dependent on salt concentrations. Therefore, to analyze the aggregate size in detail, an NTA assay was performed. We incubated the p53C solution at a concentration of 4 μM for 1 h at 37°C and then placed it on ice to stop aggregation. To allow for proper NTA analysis, the samples were diluted 100-fold after aggregation and then assayed (Supplementary Figure S3). To facilitate understanding of the measurement results, the number of particles observed at each 200 nm was summed (Figure 1F). The results showed that in 0 mM KCl, p53C-wt had approximate equal amounts of 0–199 nm and 200–399 nm particles, whereas in 400 mM KCl, the majority of p53C-wt particles were 0–199 nm. The aggregates of p53C-R175H under 0 mM KCl conditions were larger than those of p53C-wt, and there were more particles larger than 200 nm than those under other conditions. The aggregates of p53C-R175H and p53C-wt in 400 mM KCl showed mostly 0–199 nm particles.

For further analysis, the number of particles with a size of 0–199 nm was plotted on the horizontal axis and the number of particles with a size of 200 nm or larger was plotted on the vertical axis (Figures 1G,H). Both p53C-wt and p53C-R175H shifted to the lower right; that is, the particle size became smaller with an increase in salt concentration.

The relationship between the aggregation conditions and the size of the aggregates was consistent with that observed by fluorescence microscopy, although particles with a size of less than 20 μM observed in the fluorescence microscopy could not be detected because they were beyond the detection limit of the NTA assay.

The aggregation-inhibiting and aggregation-promoting effects of salt follow the Hofmeister series. To clarify the Hofmeister effect on the aggregation of p53C, chloride salts of Na⁺, NH₄ ⁺, and Rb⁺ or potassium salts of CH₃COO⁻ and Br⁻ were added. According to Hofmeister (1888), anions Br⁻, Cl⁻, and CH₃COO⁻, in that order, and cations NH₄ ⁺, Rb⁺, K⁺, and Na⁺, in that order, are more effective in salting out proteins. Upon aggregation in a 96-well microplate at 37°C for 60 min, both p53C-wt and p53C-R175H exhibited a similar fluorescence ratio of ThT/ANS in the presence of each tested salt (Figure 2A; Supplementary Figure S4). Further, the order of the strength of the salting-out effect did not correspond to the order of the increase or decrease in ANS or ThT, suggesting that the inhibition of aggregation was different from that observed in other amyloids due to the salting-in and salting-out effects of ions (Rabbani et al., 2012).

We then examined the effects of glucose (Glc) and trehalose (Trh) on p53 aggregation; these uncharged osmolytes reportedly affect protein aggregation (Iwaya et al., 2020). Since Glc and Trh are monosaccharide and disaccharide, respectively, a concentration of 200 mM was adopted for Trh to match the monosaccharide concentration (Rabbani and Choi, 2018). As in the case of charged osmolytes, we calculated the ratio of ThT/ANS. The results showed that neither Glc nor Trh had any effect on the ratio of ThT/ANS after 60 min of aggregation of p53C (Figure 2B). The changes in the fluorescence intensity of ANS and ThT were measured over time using a fluorophotometer, and we found that the values of both ANS and ThT showed an increasing tendency (Figures 2C–F). Since the fluorescence intensity of ANS and ThT increased in a similar manner, the ratios were close.

Since p53 is a transcription factor, its binding to DNA is essential for its function. To evaluate the effect of KCl on this protein–DNA binding, we performed a gel shift assay. A graphical overview of the experiment is shown in Figure 3A. A p53C solution at a concentration of 4 µM was prepared on ice and incubated at 37°C for 30, 60, 90, and 120 min for the p53C-wt and 5, 10, 15, and 20 min for the p53C-R175H, and then placed on ice again to stop p53C aggregation. The DNA for the assay was a 100-bp sequence that contained a p53C binding site, which was amplified by PCR and purified. The sample was placed on ice and mixed with the 100 bp DNA, and the KCl concentration was adjusted so that the final concentration of KCl was 200 mM, a concentration at which the salt had no effect on p53C-DNA binding (Ishimaru et al., 2009). The DNA was also mixed after aggregation because binding to DNA and RNA stabilizes p53 and inhibits aggregation (Paleček et al., 1997; Ishimaru et al., 2009; Silva et al., 2010; Kovachev et al., 2017). After electrophoresis of the mixture of p53C and DNA on a TBE gel, the gel was stained with ethidium bromide to detect DNA bands, and the protein complex fraction and the unbound DNA fraction were analyzed and normalized by binding rate (Figures 3B,C; Supplementary Figure S5). Unless they were exposed to 37°C, we confirmed that not only p53C-wt but also p53C-R175H retained their DNA-binding capacity. We also found that the DNA-binding activities of p53C-wt and p53C-R175H in samples with no additives were lost because of aggregation. As expected, both p53C-wt and p53C-R175H with KCl retained their DNA-binding ability. The addition of Glc also kept the DNA binding activity of p53C-wt equivalent to that of KCl. However, Glc was not as effective as KCl in maintaining the DNA-binding activity of p53C-R175H. Intriguingly, the DNA-binding activity of p53C-wt was maintained under the conditions of a high ratio of ThT/ANS or increased ThT fluorescence.

To detect the aggregation details of p53C-wt and p53C-R175H with and without KCl or Glc, we obtained one-dimensional ¹H-NMR spectra. In the NMR experiment, the structure of the monomer should be detectable because the aggregates increase in molecular weight, with the peaks broadening, and should not be detected. Almost all experiments were performed at a protein concentration of 4 μM, but NMR was performed at a concentration of 15 μM to increase the signal-to-noise ratio. To suppress the appearance of peaks other than p53C in the 1D-NMR spectrum, phosphate buffer and deuterated glucose were substituted for Tris-HCl buffer and glucose, respectively, and DTT was not added. Samples were prepared on ice and incubated at 37°C for 15 or 30 min for the p53C-wt, 5 or 10 min for the p53C-R175H, and then placed back on ice to stop aggregation, followed by NMR measurements at 15°C. Note that we checked separately that the NMR spectrum of p53C did not change for at least 16 h at 15°C. To capture slight changes in proteins, we focused on the amide proton region, which reflects subtle changes around the amino acids that constitute the protein. We initially monitored the structural changes associated with the aggregation of p53C-wt and p53C-R175H without any small additives. A comparison of p53C-wt and p53C-R175H without additives revealed no significant differences in the peak patterns, but sharper peaks were observed for p53C-wt (Figures 4A,D). The peak intensities of both p53C-wt and p53C-R175H decreased markedly over time, probably due to the increase in molecular weight caused by aggregation and the marked decrease in mobility.

With the addition of Glc, there was an increase in peak intensity in both p53C-wt and p53C-R175H, prominent in p53C-R175H (Figures 4B,E). The peaks of p53C-wt with Glc were still detectable after 30 min. However, the peak intensity of p53C-R175H with Glc decreased at the 5-min time point, as with the samples with no additives.

These results of 1D ¹H-NMR experiments for Glc are consistent with those of the DNA binding assay, which showed that p53C-wt retained its DNA-binding ability in the presence of Glc, whereas p53C-R175H with Glc or samples with no additives lost that ability.

The patterns of the spectra of both p53C-wt and p53C-R175H under 400 mM KCl conditions were different from those under non-additive conditions before aggregation (Figures 4C,F). In the 400 mM KCl condition, a decrease in peak intensity due to aggregation was observed, as in the non-additive condition, but there was no change in the pattern of peaks due to aggregation for both p53C-wt and p53C-R175H. The peaks of p53C-R175H with KCl were still observable after 5 min.

Considering these results together, we propose that the DNA-binding ability of both p53-wt and p53-R175H can be retained by the addition of KCl, and that both the residual peaks and the peak pattern observed in this study might be related to the DNA-binding structure of p53C.

The substitution of the 175th arginine with histidine makes it difficult for the Zn ion to coordinate (Butler and Loh, 2003). The structural transition due to Zn removal can be monitored by the change in the ratio of the fluorescence intensity at 305–333 nm when excited at 280 nm. We monitored the structural transition of p53C-wt in the presence and absence of EDTA, as well as that of p53C-R175H without additives (Figure 5A). The profile of p53C-wt with EDTA was close to that of p53C-R175H. The plateau value of p53C-R175H seemed to be the endpoint of the Zn-dislocated structure.

Subsequently, we also monitored p53C-wt and p53C-R175H with 400 mM KCl and 400 mM Glc to reveal the effects of those additives on the structural transition (Figures 5B,C). The results showed that KCl suppressed the structural transition in both p53C-wt and p53C-R175H. Glc also suppressed the structural transition, but the ratio of the final structural transition in p53C-wt was also reduced, whereas in p53C-R175H, the structural transition slowed down, but the ratio of the final structural transition was almost the same as that with no additives. These results are consistent with the observations of NMR experiments, although the protein concentrations were different. This result showed that KCl and Glc could interfere with the structural change due to the loss of Zn.