Acquired On-Target Alterations Drive Clinical Resistance to p53-Y220C Reactivators

On-target TP53 secondary mutations arise during progression on the p53-Y220C reactivator rezatapopt, either abolishing p53 function or reshaping the Y220C pocket to block drug binding, defining translational resistance classes.

Abstract

The tumor-suppressor TP53 is the most frequently altered gene in cancer, and the Y220C hotspot, found in 1.8% of TP53-mutant tumors, creates a druggable cavity that destabilizes p53. Rezatapopt, a first-in-class, orally bioavailable reactivator of Y220C-mutant p53, has demonstrated promising initial efficacy in the phase 1/2 PYNNACLE trial. We report the first clinical mechanisms of resistance to this therapeutic class. Profiling of circulating tumor DNA, tumor biopsies, and rapid autopsy specimens upon rezatapopt progression revealed multiple heterogenous secondary TP53 alterations in cis with Y220C, including (i) DNA-binding domain mutations or frameshift/nonsense mutations that abolish transcriptional activity and (ii) mutations within the Y220C-binding surface predicted to hinder drug binding. Functional modeling confirmed that these double mutants eliminate p53 reactivation and target gene induction by rezatapopt. These findings establish a molecular framework for resistance to p53 Y220C reactivators and inform strategies to overcome resistance with next-generation agents.

Significance:

This study illustrates how pan-cancer resistance to Y220C-mutant p53 reactivators emerges in patients, indicating that on-target acquired alterations can represent a major mechanism of clinical resistance. These insights establish a molecular basis for therapeutic failure and provide a framework for developing next-generation agents to overcome resistance.

See related commentary by Liu and Gu, p. 620

Introduction

Loss-of-function mutations affecting TP53 are among the most frequent alterations observed in cancer, making it a potentially important target for cancer therapy. However, targeting of altered tumor-suppressor genes with small molecules has been challenging, as it requires an agent capable of restoring tumor-suppressor gene function (1). The mutation of tyrosine 220 (Y220) to cysteine (Y220C) in p53 represents a unique case creating a cryptic pocket that is not present in the wild-type (WT) protein conformation (2). Importantly, this cryptic pocket is located adjacent to the DNA-binding domain of p53 and has enabled the design of small molecules that can bind to this pocket, interact with the mutant cysteine, and revert Y220C p53 back to an active conformation, thus restoring p53 function (3). Rezatapopt (4), a p53-Y220C reactivator currently in clinical development, has demonstrated promising clinical efficacy across multiple tumor types, with an acceptable safety profile. In the ongoing phase 1/2 PYNNACLE trial, which has enrolled 112 patients (data cutoff September 4th, 2025), the overall response rate was 34% with a median duration of response of 7.6 months (5–7). However, as with most targeted therapies, acquired resistance eventually develops. In this study, we performed an initial characterization of potential mechanisms of clinically acquired resistance to p53-Y220C reactivators by evaluating serial circulating tumor DNA (ctDNA) in patients receiving rezatapopt.

Results

Two patients with advanced solid tumors and a confirmed TP53 Y220C mutation by genomic analysis of tumor tissue [DNA whole-exome sequencing (WES)] and blood (ctDNA) enrolled in the monotherapy arm of the PYNNACLE trial (NCT04585750) and received rezatapopt (PC14586). Patient 1 (Fig. 1A) was a 58-year-old female with poorly differentiated, HPV-negative head and neck carcinoma, metastatic to the lung, liver, and lymph nodes, which had progressed despite prior chemotherapy and immunotherapy. The patient achieved a confirmed radiographic partial response based on RECIST v1.1 criteria, lasting approximately 5 months. Upon progression, imaging revealed enlargement of known liver metastases. Subsequent profiling at progression demonstrated the emergence of as many as 94 mutations in TP53, the vast majority identified in ctDNA, all of which encompassed the mutations detected by tissue-based sequencing (Fig. 1B; Supplementary Fig. S1). Although serial blood samples were not available for this patient, a ctDNA sample at baseline determined that mutations of interest were not detected prior to rezatapopt treatment (Supplementary Fig. S2).

Figure 1.

Clinical history of the patient cohort. A, Axial enhanced CT images of patient 1’s abdomen showing measurable target liver lesions at baseline, on-treatment, and at progression. B, Lollipop graph depicting acquired TP53 mutations for patient 1 at disease progression. C, Axial CT images of patient 2’s pelvis showing measurable target iliac lymph node lesion at baseline, on-treatment, and at progression. D, Lollipop graph depicting acquired TP53 mutations for patient 2. E, ddPCR analysis of serial ctDNA samples from patient 2. Acquired mutations depicted were initially observed by ctDNA analysis at the time of progression or tissue sequencing at the time of rapid autopsy. Treatment course with rezatapopt is indicated by shading.

Patient 2 (Fig. 1C) was a 65-year-old male diagnosed with metastatic squamous cell carcinoma of the ureter (retroperitoneal lymph node and bone involvement) which had progressed despite chemotherapy and immunotherapy. In response to rezatapopt, the patient experienced rapid clinical symptom improvement, whereas radiographic evaluation showed stable disease (−4%) with a reduced tumor burden. He remained on therapy for approximately 6 months before developing new liver metastases. Similarly, postprogression ctDNA analysis as well as tissue-based sequencing obtained during a rapid autopsy procedure revealed the emergence of 13 TP53 mutations not detectable prior to rezatapopt exposure (Fig. 1D). Serial ctDNA analysis throughout therapy showed that the allele frequency of Y220C TP53 initially decreased but then rebounded sharply as the patient’s tumor progressed. The rebound in Y220C TP53 levels was accompanied by the emergence of new TP53 mutations, not detectable prior to rezatapopt (Fig. 1E).

Taken together, these observations suggest that emergence of heterogeneous on-target secondary mutations could be an important mechanism of clinical acquired resistance to Y220C-mutant p53 reactivators. Interestingly, we noted that acquired TP53 alterations could be categorized into distinct classes. A first class comprises mutations that impair p53 function, including loss-of-function point mutations acting as a “second hit” that inactivate the restored rezatapopt-bound p53, as well as frameshifts or nonsense mutations resulting in nonfunctional protein products. A second class of alterations includes mutations occurring within the Y220C-binding crevice (Fig. 2). Notably, some of these alterations are not known to impair p53 function but may alter the ability of rezatapopt to bind p53-Y220C.

Figure 2.

Acquired mutations cluster within the Y220C crevice. Representation of Y220C-mutant p53 in complex with PC-9859, a precursor of the clinical compound rezatapopt (PC14586). Highlighted are the amino acids located in the Y220C crevice found mutated in our patient cohort (PDB code: 9BR4).

We selected representative mutations from both classes and introduced them in p53-null cell models to functionally characterize their effect on response to rezatapopt. Expression of these mutations in cis with Y220C in non–small cell lung cancer (NSCLC) or head and neck squamous cell carcinoma cell models consistently abolished rezatapopt’s effects on cell viability, increasing the IC₅₀ values greater than ten times compared with single-mutant Y220C-expressing cells (Fig. 3A–D). In p53-Y220C expressing cells, p53 reactivation by rezatapopt results in transcriptional upregulation of p53-target genes such as the cell-cycle inhibitor p21, the p53-negative regulator MDM2, or the proapoptotic gene PUMA (7). Consistent with the lack of in vitro activity in double-mutant–expressing cells, rezatapopt treatment did not result in upregulation of p53-target genes (Fig. 3E; Supplementary Figs. S3, S4A and S4B). The Y220C mutation can drive protein instability, compromising its thermal stability (8). Therefore, we tested whether the lack of drug sensitivity emerged from enhanced protein instability upon adding a second mutation. Notably, these secondary alterations did not affect Y220C-mutant protein levels unless causing a shift in the reading frame (Supplementary Figs. S3 and S4B). As a proof of principle, we demonstrated that whereas the double-mutant Y220C/N288fs was transcribed in our cell models, p53 protein levels remained undetectable, indicating that secondary frameshifts define a mechanism of acquired resistance mediated by drug target loss (Supplementary Fig. S4C).

Figure 3.

Acquired TP53 mutations drive resistance to rezatapopt. A, Cell viability of H1299 cells engineered to express the indicated p53 alterations and treated with rezatapopt for 72 hours. Data are normalized to vehicle-treated cells. Error bars, SD (n = 3). e.v., empty vector. B, Effective concentration resulting in 50% growth inhibitory effect (EC₅₀) is depicted for indicated cell line. Each dot represents an independent experiment. C, Cell viability of SCC9 cells engineered to express the indicated p53 alterations and treated with rezatapopt for 72 hours. Data are normalized to vehicle-treated cells. Error bars, SD (n = 3). e.v., empty vector. D, Effective concentration resulting in 50% growth inhibitory effect (EC₅₀) is depicted for indicated cell line. Each dot represents an independent experiment. E, Gene expression changes in H1299 cells treated with rezatapopt for 24 hours. Data are normalized to vehicle-treated cells. Error bars, SD (n = 3).

The first class of acquired mutations introduced in our cell models (R175H, R248W, R273C, and R282W) constitutes a group of well-characterized deleterious TP53 point mutations (9). Although they are structurally located outside of the Y220C crevice, these residues are critical to interact and bind to DNA regions in which p53 acts as a transcription factor. Therefore, they can potentially inactivate p53 even if rezatapopt could physically interact and restore p53 to its WT conformation. These “second hit” alterations are likely to cause universal resistance to small molecules aimed to restore p53 function by binding to the Y220C crevice.

In contrast, we identified a second class of acquired mutations affecting residues F109, L145, V147, P151, and P152 that are physically located within the pocket in which rezatapopt binds (Fig. 2). Mutations in these residues are not frequently observed in tumors (10), and we hypothesized that they could alter the binding affinity of rezatapopt to the Y220C crevice. Indeed, when expressed in cis with Y220C in our cell models, they all invariably lead to insensitivity to rezatapopt (Fig. 3A–D).

To elucidate their role in resistance, we initially characterized whether expressing these mutations in the absence of the Y220C mutation result in a p53 loss-of-function phenotype. To do so, we evaluated whether they were able to bind to chromatin (Fig. 4A) as well as to promoter regions of p53-target genes, such as CDKN1A (encoding p21) and MDM2 (Fig. 4B). Interestingly, we found that many of the mutations observed in the binding crevice also abrogated p53 function. However, cells transfected with p53-V147L behaved very similarly to p53-WT–expressing cells, indicating that this acquired mutation does not prevent p53 from binding to DNA target regions and does not notably alter p53 function (Fig. 4A and B). Additionally, p53-P152R retained partial binding, suggesting that some alterations affecting the P152 residue may not fully inactivate p53. We confirmed that p53-V147L–expressing cells showed transcriptional upregulation of CDKN1A, MDM2, and PUMA (Fig. 4C; Supplementary Fig. S5), as well as decreased cell viability 24 hours after transfection, similarly to our p53-WT control cells. However, p53-P152R–expressing cells failed to show transcriptional activity of p53 downstream genes, suggesting that whereas this mutation permits promoter binding to some extent, it still results in p53 inactivation (Fig. 4C; Supplementary Fig. S5). Thus, whereas some alterations in the Y220C-binding crevice also drive loss of p53 function, others such as V147L do not impair p53 function and likely function by reducing the binding affinity of rezatapopt in the Y220C crevice.

Figure 4.

Some acquired mutations located in the Y220C crevice act as drug-binding disruptors. A, p53-bound to chromatin in H1299 cells transfected with the indicated p53 alterations for 24 hours (top). Whole cell fraction is included in the bottom panel. B, CUT&RUN coupled to qPCR for MDM2 and CDKN1A in H1299 cells transfected with the indicated p53 alterations for 24 hours (shown is the mean and standard error, n = 5). C, Gene expression changes in H1299 cells transfected with the indicated p53 alterations for 24 hours. Data are normalized to GFP control-transfected cells. Errors bars, SD (n = 3).

Interestingly, we identified additional substitutions affecting residues P151 and P152 in patient 1 at disease progression (Supplementary Fig. S1). As P152R retained partial binding at promoter regions of p53-target genes, we interrogated whether a different substitution affecting the same residues could still result in a functional protein. In the case of residue P151, both substitutions into either an alanine (P151A) or an arginine (P151R) impaired transcriptional activity of p53-regulated genes. In contrast, modifying a proline for a leucine at residue 152 (P152L) resulted into a p53 form retaining partial functionality (Supplementary Fig. S6A), an observation which did not occur when the proline was substituted by an arginine (P152R; Fig. 4C). Despite this, alterations in P151 or P152 expressed in cis with Y220C invariably abrogated the effect of rezatapopt on cell viability (Supplementary Fig. S6B and S6C). In conclusion, similarly to the V147L acquired alteration, certain mutations at proline 152 may also induce rezatapopt resistance by structurally modifying the Y220C crevice and preventing drug binding.

Discussion

In this study, we provide the first clinical evidence that on-target secondary TP53 mutations can represent a major mechanism of acquired resistance to the p53-Y220C reactivator rezatapopt. Although our analysis was limited to two patients, both individuals exhibited numerous newly emergent TP53 alterations at relapse, strongly implicating selective pressure from Y220C-directed therapy in shaping tumor evolution. Notably, one patient developed nearly one hundred de novoTP53 variants, an extraordinary degree of genomic diversification, highlighting the intense evolutionary drive to abrogate p53 function and evade drug-induced cell death.

By functionally characterizing a representative subset of these alterations, we establish proof-of-principle that Y220C-directed selective pressure can yield at least two mechanistic classes of resistance. The first comprises deleterious TP53 mutations, including well-characterized DNA-binding domain substitutions and frameshifts or nonsense mutations, which eliminate p53 transcriptional activity and are predicted to confer cross-resistance to all agents within this therapeutic class. The second includes mutations clustering within the Y220C-binding surface, which likely remodel the druggable crevice to disrupt rezatapopt binding. Importantly, some members of this second class do not invariably abolish p53 activity, suggesting that resistance here may be drug-specific rather than universally applicable across all Y220C reactivators.

These structural insights have immediate translational implications. If some resistance mutations selectively impair rezatapopt binding without destabilizing the mutant protein, alternative compounds, particularly those with different steric profiles, higher binding affinities, or irreversible covalent chemistry may retain efficacy. Rational structure-guided drug design and medicinal chemistry optimization will be essential to generate next-generation Y220C reactivators capable of circumventing such resistance.

By contrast, secondary alterations that abolish p53 function pose a formidable therapeutic challenge. Addressing this type of resistance may require clinical strategies that combine p53 reactivators with agents targeting vulnerabilities conferred by p53 loss, such as dependencies in DNA damage response or cell-cycle regulation. Rigorous preclinical validation will be critical to inform rational trial designs that can effectively evaluate such approaches.

Future studies in larger patient cohorts are warranted to define the prevalence, timing, and evolutionary dynamics of these resistance classes. Integration of functional assays with patient-derived genomic data will be critical to distinguish between universally deleterious TP53 alterations and drug-specific escape mutations, thereby guiding clinical decision-making. Ultimately, these insights provide a molecular framework for understanding and overcoming therapeutic resistance in the emerging field of allele-specific p53 reactivation.

Methods

Patients

Patients with advanced solid tumors with a confirmed TP53 Y220C alteration were enrolled at MGH/DFCI starting in 2021 on clinical trial NCT04585750. All clinical data and tumor specimens were collected and analyzed in accordance with the Institutional Review Board–approved protocol (DFHCC 13-416), to which patients provided written informed consent, and all studies were conducted in accordance with the Declaration of Helsinki.

Cell Culture and General Reagents

H1299 cells were cultured in RPMI medium with 10% fetal bovine serum (FBS) and 1% sodium pyruvate (Thermo Fisher Scientific, #11360070). SCC-9 cells were maintained in DMEM/F12 medium with 10% FBS. HEK293T cells were grown in DMEM medium with 10% FBS. All cell lines were obtained from the ATCC or from the MGH Center for Molecular Therapeutics cell line collection, which performs routine short tandem repeat genotyping and Mycoplasma testing, and were grown in the presence of penicillin–streptomycin at 37°C and 5% CO₂. Lentiviral particles were produced in HEK293T cells. Mutated TP53 genes were introduced in H1299 and SCC-9 cells by lentiviral transduction followed by puromycin selection. Transient expression of TP53 genes in H1299 cells were carried out by Lipofectin 2000–mediated transfection.

All antibodies were purchased from Cell Signaling Technologies. Rezatapopt was kindly provided by Kevan Shokat.

TP53_WT construct (#82754) was obtained from Addgene. To generate mutants, QuikChange II Site-Directed Mutagenesis Kit (Agilent Technologies) was used. Introduced point mutations were verified by Sanger sequencing, and mutants were shuttled into a gateway-compatible pLENTI-puro vector obtained from Addgene.

Dose–Response Experiments

Viability assays were carried out using the CellTiter-Glo assay (Promega). Cells were counted and seeded in 96-well plates in triplicates. The next day, compounds were added, and 3 days later, cell viability was assessed. EC₅₀ value determination was performed using GraphPad Prism.

Western Blot Analysis

Cell lines were treated with rezatapopt for 24 hours, and lysates were prepared as described previously (11). All antibodies were diluted in 5% BSA + 0.02% sodium azide in TBST as follows: p53 (1:1,000), p21 (1:1,000), GFP (1:1,000), histone H3 (1:5,000), and GAPDH (1:2,500).

Quantitative PCR

RNA was extracted using a RNeasy kit (Qiagen) per the manufacture’s protocol. Reverse transcription was performed using qScript cDNA SuperMix (Quantabio). qPCR analysis was performed using TaqMan Gene Expression Master Mix (Thermo Fisher Scientific) on the Roche Light Cycler 480. β-Actin (Thermo Fisher Scientific; 4325788) was used as endogenous control.

Chromatin Fractionation

Cells were washed once with ice-cold PBS and collected by scraping in 1 mL chilled PBS on ice. Following centrifugation at 13,000 g for 2 minutes at 4°C, the supernatant was aspirated, and the cell pellet was resuspended in 200 µL hypotonic buffer (10 mmol/L HEPES-KOH pH 7.5, 5 mmol/L KCl, 1.5 mmol/L MgCl₂, 1 mmol/L DTT, and 0.5% NP-40) supplemented with protease inhibitor cocktail (PI, Roche, 11836170001). The suspension was incubated on an end-over-end rocker for 10 minutes at 4°C and centrifuged at 4,000 g for 5 minutes at 4°C. The supernatant was removed, and the pellet was washed once with 200 µL hypotonic buffer and then gently resuspended in 200 µL hypotonic buffer supplemented with PI and 0.5 mol/L NaCl. After a second 10-minute incubation on an end-over-end rocker at 4°C, the sample was centrifuged at 13,000 g for 5 minutes at 4°C. The pellet was washed once more in 200 µL hypotonic buffer and then resuspended in 100 µL 2× SDS sample buffer and boiled at 95°C for 10 minutes. Lysates (10–20 µL) were used for SDS-PAGE analysis.

CUT&RUN

For each sample, 3 × 10⁵ cells were harvested and washed twice in ice-cold Wash Buffer [20 mmol/L HEPES-KOH, pH 7.5, 150 mmol/L NaCl, 0.5 mmol/L Spermidine, and PI (Roche, 11836170001)]. After the final wash, cells were resuspended in 200 μL of Wash Buffer per reaction and kept on ice until further processing.

BioMag Plus Concanavalin A–coated magnetic beads (10 μL/reaction) were activated by washing twice with 1.5 mL of Binding Buffer (20 mmol/L HEPES-KOH, pH 7.9, 10 mmol/L KCl, 1 mmol/L CaCl₂, and 1 mmol/L MnCl₂) and resuspended to the original bead slurry volume (10 μL per reaction). For each reaction, 10 μL of bead slurry was added to 200 μL of cells (3 × 10⁵ cells per reaction) and incubated at room temperature for 10 minutes with gentle rotation to allow binding of cells to the beads.

The bead-bound cells were washed twice with 1.5 mL of Wash Buffer and resuspended in 200 μL of Antibody Buffer (Wash Buffer containing 4 μL/mL of 0.5 mol/L EDTA and 0.0125% digitonin (EMD Millipore, 300410). Anti-IgG antibody or anti-p53 antibody were added at a dilution of 1:100 in Antibody Buffer and incubated overnight at 4°C with rotation.

On the following day, cells were washed twice in 1.5 mL of digitonin Wash Buffer (0.0125% digitonin in Wash Buffer). Two μL/reaction of pAG-MNase (Cell Signaling Technology, 40366S) in digitonin Wash Buffer was added to the cell–bead complexes (200 μL per reaction). The mixture was incubated at 4°C for 2 hours with gentle rotation to allow binding of the pAG-MNase to the antibody–protein complexes.

After pAG-MNase binding, cells were chilled in an ice bath for 5 minutes. Chromatin digestion was initiated by adding 3 μL of 100 mmol/L CaCl₂ to each reaction while vortexing, and reactions were incubated in the ice bath for 30 minutes to allow chromatin cleavage by MNase. The digestion was stopped by adding 150 μL of 2× STOP Buffer [200 mmol/L NaCl, 20 mmol/L EDTA, 4 mmol/L EGTA, 0.05% digitonin, 50 μg/mL RNase A (Thermo Fisher Scientific, EN0531), and 40 μg/mL glycogen (Sigma-Aldrich, 10901393001)].

Reactions were incubated at 37°C for 10 minutes to release soluble chromatin fragments. After incubation, samples were centrifuged at 16,000 g for 3 minutes at 4°C, and the supernatant containing the chromatin fragments was transferred to a fresh 1.5-mL microcentrifuge tube.

DNA was extracted by adding 3 μL of 10% SDS and 2.5 μL of proteinase K (20 mg/mL, New England BioLabs, P8107S) to each sample, followed by incubation at 70°C for 10 minutes. After cooling, 300 μL of phenol–chloroform–isoamyl (25:24:1) was added to each sample, vortexed for 30 seconds, and centrifuged at 16,000 g for 5 minutes at room temperature. The aqueous phase was transferred to a phase-lock gel tube and extracted again with 300 μL of chloroform. After centrifugation, the aqueous phase was transferred to a new tube containing 2 μL of 2 mg/mL glycogen. DNA was precipitated with 750 μL of 100% ethanol, chilled on ice for 10 minutes, and centrifuged at 16,000 g for 10 minutes at 4°C. The DNA pellet was washed twice with 1 mL of 100% ethanol, air-dried, and resuspended in 100 μL of sterile water. Samples were run on a 384-well plate using Applied Biosystems PowerUp SYBR Green Master Mix for qPCR (Fisher Scientific, A25742) on the LightCycler 480 Instrument II using indicated primers.

Cell-Free DNA Extraction and Droplet Digital PCR

Whole blood was collected by routine phlebotomy in two 10-mL Streck tubes. Plasma was separated within 1 to 4 days of collection through two different centrifugation steps (the first at room temperature for 10 minutes at 1,600 × g and the second at 3,000 × g for the same time and temperature). Plasma was stored at −80°C until cell-free DNA (cfDNA) extraction. cfDNA was extracted from plasma using QIAamp Circulating Nucleic Acid Kit (QIAGEN) with 60 minutes of proteinase K incubation at 60°C. All other steps were performed according to the manufacturer’s instructions. For Droplet Digital PCR (ddPCR) experiments, DNA template (up to 10 μL, with a total of 20 ng for each specimen) was added to 10 μL ddPCR Supermix for Probes (Bio-Rad) and 2 μL custom primer/probe mixture. This reaction mix was added to a DG8 cartridge together with 60 μL Droplet Generation Oil for Probes (Bio-Rad) and used for droplet generation. Droplets were then transferred to a 96-well plate (Eppendorf) and then thermal cycled with the following conditions: 10 minutes at 95°C, 40 cycles at 94°C for 30 seconds, 55°C (with a few grades of difference among assays) for 1 minute, followed by 98°C for 10 minutes (Ramp Rate 2°C/s). Droplets were analyzed using the QX200 Droplet Reader (Bio-Rad) for fluorescent measurement of FAM and HEX probes. Gating was performed based on positive and negative controls, and mutant populations were identified. The ddPCR data were analyzed with QuantaSoft analysis software (Bio-Rad) to obtain a fractional abundance of the mutant DNA alleles in the WT/normal background. The quantification of the target molecule was presented as the number of total copies (mutant plus WT) per sample in each reaction. An allelic fraction was calculated as follows: AF % = [Nmut/(Nmut + Nwt) × 100], in which Nmut is the number of mutant alleles and Nwt is the number of WT alleles per reaction. ddPCR analysis of normal control plasma DNA and no DNA template controls was always included. Probe and primer sequences are available upon request.

Mutation Analysis

Using nine tumor and three normal samples, we performed tumor–normal paired mutation calling using MuTect1 (12) and VarDict (13). We used these variant callers with less stringent parameters than default settings to maximize sensitivity to obtain an initial set of candidate mutations. These variants are then further filtered using our machine learning classifier which utilizes read-level features and functional annotations to stratify any single-nucleotide variants into somatic and artifact categories. Using the classifier generated prediction scores as well as multisample information, we obtained a final set of mutations based on the following selection criteria: (i) if the classifier prediction score was greater than 0.02 (with an expected FPR of 4.08%) even in a single sample, (ii) mutations detected in more than three tumor tissue samples were retained if they had either a minimum classifier prediction score greater than 0.001 or a maximum classifier score greater than 0.01 with expected FPR of 2.57% assuming the artifacts are independent. Some artifacts can occur at hotspots; therefore, we further excluded mutations in which the alternate allele was observed in one of the three matched normal samples. We further excluded synonymous mutations and those with unannotated or unavailable protein-level effects. To ensure the final set of mutations represented true somatic events, we manually inspected IGV plots for each mutation site in the corresponding tumor tissue sample and three matched normal samples.

Supplementary Material

Supplementary Figures S1-S6

Supplementary Figure S1|List of p53 alterations detected in patient 1 at disease progression. Supplementary Figure S2|Mutations surrounding the Y220C crevice were not present before treatment with rezatapopt. Supplementary Figure S3|Acquired p53 alterations render cells insensitive to rezatapopt treatment. Supplementary Figure S4|Acquired p53 alterations render HNSCC cells insensitive to rezatapopt treatment. Supplementary Figure S5|Not all acquired mutations in the Y220C crevice result in p53 loss-offunction. Supplementary Figure S6|Alterations affecting P152 invariably led to resistance to rezatapopt, while some substitutions may retain partial p53 functionality.

Data Availability

The breakdown of acquired TP53 alterations found at progression to rezatapopt is shown in Supplementary Figure S1 (patient 1) and Fig. 1D (patient 2). The accession number for the publicly available crystal structure used in Fig. 2 is provided in the figure legend. PCR probes are shown in Fig. 1 and Supplementary Fig. S2 and are available through direct communication with the corresponding author. Raw data are available upon reasonable request from the corresponding authors. WES data from autopsies are deposited at dbGaP under the accession number phs001853.v1.p1.

Authors’ Disclosures

A. Varkaris reports other support from PMV Pharmaceuticals during the conduct of the study. L. Pappas reports other support from Eli Lilly and Moderna Therapeutics and personal fees from Kestrel Therapeutics, CARIS Life Sciences, Astellas Pharma, Merus, Tallac Therapeutics, Bristol Myers Squibb, and Takeda Pharmaceuticals outside the submitted work. D. Juric reports grants and personal fees from Blueprint, Novartis, Roche, and Lilly, personal fees from Relay Therapeutics and PIC Therapeutics, and grants from AstraZeneca and Nikang Therapeutics outside the submitted work. J.L. Hopkins reports previous employment at Nested Therapeutics and currently employment at Acrivon Therapeutics. K.Z. Guiley reports other support from Nested Tx and Rezo Tx outside the submitted work; in addition, K.Z. Guiley has a patent for WO2021087096A1 issued. K.M. Shokat reports grants from Emerald Foundation during the conduct of the study and personal fees from BridGene Biosciences, Erasca, G Protein Therapeutics, Genentech, Kumquat Biosciences, Kura Oncology, Lyterian, Merck, Nextech, Pfizer, Revolution Medicines, Rezo, Tahoe Therapeutics, Type6 Therapeutics, and Wellspring Biosciences outside the submitted work; in addition, K.M. Shokat has a patent for WO2021087096A1 pending. D.C. Gulhan reports a patent for SigMA pending and licensed to Pfizer and Takeda. A.R. Parikh has held equity in C2i Genomics, Khora, OneCell, XGenomes, Cadex, and Parithera. In the last 36 months, she has served as an advisor/consultant for Zola, CVS, Phesi, Xilio, 3T Biosciences, Do More Diagnostics, Summit Therapeutics, Pfizer, Regeneron, GSK, Foundation Medicine, Careset, Value Analytics Labs, Naterara, Adroya, AstraZeneca, Scare, Hookipa, Guardant, AbbVie, Seagen, Mirati, Takeda, PMV, Kahr, Sirtex, Eli Lilly, Merck, Amgen, Delicate, Exact, Caris, Bristol Myers Squibb, Incyte, Pheon, Neogenomics, J & J, Exact, Boehringer Ingelheim, Third Rock Ventures, MPM Capital, and Science For America. She is the chief scientist of Reversing Early Recurrence. She receives fees from Up to Date. She has received travel fees from Karkinos Healthcare. She has received research funding to the Institution from PMV Pharmaceuticals, Bristol Myers Squibb, Mirati, Erasca, Genentech, Daiichi Sankyo, Syndax, Revolution Medicine, Lily, Xilio, and Parthenon. R.B. Corcoran reports personal fees and other support from Remix Therapeutics, Alterome Therapeutics, Sidewinder Therapeutics, C4 Therapeutics, Nested Therapeutics, Pheon Therapeutics, Cogent Biosciences, grants, personal fees, and other support from Revolution Medicines, other support from Erasca and Avidity Biosciences, personal fees from Genentech/Roche, and grants from Novartis, Relay Therapeutics, OnKure, and Parabilis Medicines outside the submitted work. No disclosures were reported by the other authors.

Authors’ Contributions

F. Fece de la Cruz: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, methodology, writing–original draft, writing–review and editing. A. Varkaris: Conceptualization, data curation, formal analysis, validation, investigation, methodology, writing–original draft, writing–review and editing. P.S. Patel: Conceptualization, data curation, formal analysis, validation, investigation, methodology, writing–original draft, writing–review and editing. E.W. Kushner: Formal analysis, investigation, writing–review and editing. A.A. Morales-Giron: Investigation, writing–review and editing. S.S. Lee: Investigation, writing–review and editing. A. Singh: Investigation, writing–review and editing. C.T. Kim: Investigation, writing–review and editing. B.L. Norden: Investigation, writing–review and editing. S. Ehnstrom: Investigation, writing–review and editing. J.M. Riedl: Formal analysis, investigation, writing–review and editing. J.M. Curtis: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, writing–original draft, writing–review and editing. H. Barnes: Investigation, writing–review and editing. A.M. Kehlmann: Investigation, writing–review and editing. N.J. Chevalier: Data curation, formal analysis, investigation, writing–review and editing. H.S. Okuma: Data curation, formal analysis, investigation, writing–review and editing. M. Patel: Investigation, writing–review and editing. L.J. Wirth: Investigation, writing–review and editing. B. Connell: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, writing–original draft, writing–review and editing. F. Nugent: Investigation, writing–review and editing. L. Pappas: Investigation, writing–review and editing. K. Lau: Investigation, writing–review and editing. D. Juric: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, writing–review and editing. J.L. Hopkins: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, writing–review and editing. K.Z. Guiley: Investigation, writing–review and editing. K.M. Shokat: Conceptualization, investigation, writing–review and editing. D.C. Gulhan: Conceptualization, data curation, formal analysis, investigation, methodology, writing–review and editing. A.R. Parikh: Conceptualization, data curation, formal analysis, supervision, investigation, methodology, writing–original draft, writing–review and editing. R.B. Corcoran: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, methodology, writing–original draft, writing–review and editing.