The majority of TP53 missense mutations identified in cancer patients are in the DNA-binding domain and are characterized as either structural or contact mutations. These missense mutations exhibit inhibitory effects on wild-type p53 activity. More importantly, these mutations also demonstrate gain-of-function (GOF) activities characterized by increased metastasis, poor prognosis, and drug resistance. To better understand the activities by which TP53 mutations, identified in Li–Fraumeni syndrome, contribute to tumorigenesis, we generated mice harboring a novel germline Trp53R245W allele (contact mutation) and compared them with existing models with Trp53R172H (structural mutation) and Trp53R270H (contact mutation) alleles. Thymocytes from heterozygous mice showed that all three hotspot mutations exhibited similar inhibitory effects on wild-type p53 transcription in vivo, and tumors from these mice had similar levels of loss of heterozygosity. However, the overall survival of Trp53R245W/+ and Trp53R270H/+ mice, but not Trp53R172H/+ mice, was significantly shorter than that of Trp53+/ mice, providing strong evidence for p53-mutant–specific GOF contributions to tumor development. Furthermore, Trp53R245W/+ and Trp53R270H/+ mice had more osteosarcoma metastases than Trp53R172H/+ mice, suggesting that these two contact mutants have stronger GOF in driving osteosarcoma metastasis. Transcriptomic analyses using RNA sequencing data from Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ primary osteosarcomas in comparison with Trp53+/– indicated that GOF of the three mutants was mediated by distinct pathways. Thus, both the inhibitory effect of mutant over wild-type p53 and GOF activities of mutant p53 contributed to tumorigenesis in vivo. Targeting p53 mutant–specific pathways may be important for therapeutic outcomes in osteosarcoma.

Significance:

p53 hotspot mutants inhibit wild-type p53 similarly but differ in their GOF activities, with stronger tumor-promoting activity in contact mutants and distinct protein partners of each mutant driving tumorigenesis and metastasis.

The TP53 gene encodes a tetrameric transcription factor, p53, that binds a specific DNA sequence and activates hundreds of genes, many in a tissue-specific manner (1). Under normal conditions, p53 activity is low due to inhibition by MDM2 and MDM4 proteins (2). Under DNA damage and other stress signals, p53 is posttranslationally modified and stabilized to activate a transcriptional repertoire that induces cell-cycle arrest, senescence, apoptosis, and metabolic changes to inhibit cell transformation (3, 4).

Given these growth-suppressive pathways, the TP53 tumor suppressor is frequently mutated or deleted in most human cancers. For example, 96% of high-grade serous ovarian cancers (5) and 80% of basal-like breast cancers (6) have TP53 alterations. Additionally, 60% to 80% of “classic” Li–Fraumeni syndrome (LFS) patients have TP53 germline mutations with high risk of developing cancer at an early age (7). Osteosarcoma is a major cancer in LFS patients, especially in children. It is a rare cancer for which treatment options are limited (8).

Remarkably, among the TP53 mutations found in human cancers, missense mutations predominate (9), including several hotspot p53 mutations located in the DNA-binding domain defined as either contact (R273H, R248Q, and R248W) or structural (R175H, G245S, R249S, and R282H) mutations based on a defective protein structure (10, 11). These mutant p53 proteins exhibit a range of activities that include: loss of wild-type (WT) activity and an inhibitory effect (IE) that reduces WT p53 activity (12), and gain-of-function (GOF) activities (13). Mutant and WT p53 proteins form mixed tetramers to inhibit WT p53 activity, often in a context-dependent manner (12). In addition to inhibition of WT p53 activities by mutant p53, compelling evidence shows that mutant p53 proteins have novel GOF activities that contribute to tumorigenesis and metastasis. Trp53R172H/+ and Trp53R270H/+ mice (mimicking human LFS TP53R175H and TP53R273H mutations, respectively) develop tumors with increased metastasis as compared with Trp53+/– mice (14, 15). Numerous studies identified multiple mechanisms of GOF activities of mutant p53 that contribute to increased invasiveness and metastatic potential (13, 16). In general terms, p53-mutant proteins can interact with other cellular proteins (some of which are tissue-specific) disrupting their normal functions or usurping their transcriptional functions to alter the transcriptome. Although mutant p53 proteins drive tumor development by multiple mechanisms, these mechanisms are not necessarily mutually exclusive. In addition, these mechanisms are cancer type– and cell context–dependent (13, 17).

Unequal tumorigenic effects of TP53 missense mutations in LFS patients have been noted (18). Knock-in mouse models that mimic patients with LFS have also been developed (19). Trp53R172H (a structural mutant) and Trp53R270H (a contact mutant) mice were generated with the Trp53R270H mouse model exhibiting a distinct spontaneous tumor spectrum from the Trp53R172H mouse, indicating that different missense mutations may drive tumorigenesis differently (15). Although a humanized TP53R258W knock-in (HUPKI) mouse was generated (20), a comparable Trp53R248W mouse model has not been developed to date. To better understand the mutant-specific phenotypes of mice mimicking the human TP53R248W mutation, and to explore the IE and GOF in vivo, a germline Trp53R245W allele was generated. The IE and GOF of p53R245W were compared with two previously published LFS mutations, p53R172H and p53R270H. All three showed comparable IE on WT p53 activities. Contact mutations p53R245W and p53R270H showed stronger GOF than the structural mutation p53R172H. In addition, osteosarcomas for all three alleles showed distinct GOF mechanisms drive tumorigenesis.

Mice and tumor analyses

Genotyping was performed by polymerase chain reaction (PCR) as previously described (21). Zp3-cre and Trp53R270H/+ mice were purchased from The Jackson Laboratory. Trp53R270H/+ mice (15) were backcrossed to C57BL/6J for three generations to obtain a background (>92% C57BL/6J) similar to Trp53R172H/+ and Trp53R245WH/+ mice. Mouse cohorts were monitored daily for tumorigenesis. Moribund mice were euthanized, and tissues were fixed in 10% v/v formalin and embedded in paraffin. Sections were stained with hematoxylin and eosin for pathologic analyses. All mouse experiments were approved by the Institutional Animal Care and Use Committee of MD Anderson Cancer Center and in compliance with the US Public Health Service Policy on Humane Care and Use of Laboratory Animals.

p53 and cleaved caspase-3 (CC3) IHC was performed as described previously with CM5 and CC3 (22) antibodies, respectively. The Vector DAB Substrate Kit (Vector Laboratories) was used for chromogenic detection. The percentage of p53-positive nuclei was determined using ImageJ software. To examine loss of heterozygosity (LOH) of p53, primers spanning the missense mutations were used to amplify tumor DNA samples by PCR; the PCR products were then sequenced. LOH was measured as loss of 80% of WT p53 peaks as previously described (21). The primers for detecting LOH in Trp53R270H/+ tumors were the following: In7FW: CCAGCTTTCTTACTGCCTTGTGC; Ex8Rev: GCAGTTCAGGGCAAAGGACTTCC.

Cell culture, immunoprecipitation, Western blot analysis, and reverse transcription quantitative PCR

Male mice (age 4–8 weeks) or mouse embryonic fibroblasts (MEF) were irradiated at 2.5 or 6 Gy, and protein lysates and total RNA were prepared from the mouse tissues or MEFs 4 hours later. The protein lysates were prepared by homogenizing mouse tissues directly in SDS-PAGE loading buffer, followed by sonication for 1 to 2 minutes. Cell lines 0263 and H318 were generated previously (23), H222 was generated from Trp53R172H/+, 14W and 55W from Trp53R245W/+, 26R and 752R from Trp53R270H/+ primary osteosarcomas. The cell lysates were prepared in NP-40 buffer and immunoprecipitated using Glial fibrillary acidic protein (Dako Z033401) or p53 antibody (Leica; cat. #NCL-L-p53-CM5p). Antibodies used for western blots were: Cdkn1a (p21; 1:1,000) (Cell Signaling Technology; cat. #2947, RRID:AB_823586); vinculin (1:5,000; Sigma-Aldrich, V9264); Egr1(1:1,000; Thermo Fisher, MA5-15008); signal transducer and activator of transcription 3 (Stat3; 1:1,000; Thermo Fisher, MA5-15712). Total RNA was prepared from cells or tissues by TRIzol reagent (Invitrogen 15596-026) and then treated with DNase I (Roche). Reverse transcription quantitative PCRs (RT-qPCR) of p53 downstream target genes Eda2r, p21, Mdm2, Bbc3 (Puma), and Pmaip1(Noxa) were performed as reported previously (1, 24). Data were normalized to Gapdh or Rplp0.

Transcriptomic analyses of murine osteosarcomas and siRNA-treated Trp53R245W/+ tumor cell line

The primary osteosarcomas from Trp53+/, Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ mice were collected at the time of necropsy, snap-frozen using liquid nitrogen, and stored at −80°C. The tumors were crushed to powder using a mortar and pestle, and approximately 100 mg of powder was used to isolate RNA using a Direct-zol RNA Microprep Kit (Zymo Research, R2062). Osteosarcoma cell line 14W was transfected with either p53siRNAs (Si-1 and Si-2) or control siRNA (MilliporeSigma, SASI_Mm02_00310137, SASI_Mm02_00310139 and SiC001), total RNAs were isolated 48 hours after the transfection using RNeasy Mini Kit (Qiagen 74004). RNA was submitted to the Advanced Technology Genomics Core at The University of Texas MD Anderson Cancer Center for bulk RNA-sequencing as described (25). Read mapping was performed using STAR RRID:SCR_004463 and GRCm38 (MM10) was used as the reference genome. The program FastQC, RRID:SCR_014583 (v. 0.11.5) was used to check for quality of FASTQ reads. Annotation of genes was carried out using the GENCODE, RRID:SCR_014966 (v. 3.6.0) and the Bioconductor, RRID:SCR_006442 package was used for analysis.

Differentially expressed genes (DEG) were identified using DESeq2, RRID:SCR_000154. The read count was first prefiltered to keep genes that had 5 or more reads in 3 or more samples. DESeq2 modeled the counts using a negative binomial distribution, followed by the Wald test. The final P value was adjusted using the Benjamini and Hochberg method. Significant DEGs were selected based on the criteria of adjusted P value < 0.05 (Padj < 0.05) when the mutant Trp53 cohorts were compared with the heterozygous Trp53 cohort. To identify potential mutant p53-associated transcription factors that lie upstream of differentially upregulated genes, data from JASPAR, TRANSFAC, ENCODE, ChEA, and Targetscan databases were used by considering the overlap between Enrichr, oPOSSUM, and Ingenuity Pathway Analysis (Ingenuity Pathway Analysis, RRID:SCR_008653, Invitrogen). This approach allowed us to avoid experimental bias and discrepancies introduced due to z-score and P value cutoffs in different algorithms. IPA was also used to identify dysregulated pathways in the mutant Trp53 cohorts using DEGs.

Statistical analysis

Student t tests and Kaplan–Meier survival analyses were performed with PRISM, RRID:SCR_005375, version 9). P values of < 0.05 were considered to be statistically significant.

Data availability statement

RNA-seq data sets generated in this study have been deposited in Gene-Expression Omnibus under accession number GSE198802.

Generation and characterization of p53R245W mice

To generate a knock-in mouse model with the p53R245W mutation (representing the hotspot p53R248W mutation in humans), we took advantage of the conditional Trp53wm-R245W allele previously characterized (21). Trp53wm-R245W expresses WT p53 but converts to p53R245W-mutant expression upon Cre-mediated deletion of WT sequences. Trp53wm-R245W mice were crossed with Zp3-Cre mice, and subsequently to C57BL/6J mice to generate germline Trp53R245W/+ mice (Supplementary Fig. S1A). After removal of the Cre transgene, interbreeding of Trp53R245W/+ mice produced a normal Mendelian ratio of expected genotypes, indicating no lethal consequences of Trp53R245W/R245W homozygosity (Supplementary Table S1).

To test whether the Trp53R245W allele is a loss-of-function allele in vivo, we examined its ability to rescue the Mdm2-null early lethal phenotype as does p53 deletion but not p53 heterozygosity (26, 27). In several crosses, the expected number of Mdm2/Trp53R245W/R245W mice were observed at weaning, indicating that homozygosity of the Trp53R245W allele completely rescues the embryonic lethality of Mdm2/ mice (Supplementary Table S2), demonstrating little or no WT activity of the p53R245W protein. To examine the transcriptional activity of p53R245W after DNA damage, the relative RNA levels of canonical p53 downstream targets cyclin-dependent kinase inhibitor 1A (Cdkn1a, or p21) and Bbc3 (Puma) were examined by qRT-PCR in the spleens of mice treated with ionizing radiation (IR). Trp53R245W/R245W mice showed no transcriptional activity after 6 Gy IR exposure (Supplementary Fig. S1B), as would be expected from an allele that lost WT p53 activity. p21 protein was also not induced in Trp53R245W/R245W mouse spleens or in MEFs after 6 Gy IR (Supplementary Fig. S1C and S1D). These results indicated that p53R245W has little if any WT p53 activity.

Similar IE from different p53 mutations in vivo

To investigate whether p53R172H, p53R245W, and p53R270H proteins exert IE over WT p53 in vivo, heterozygous Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ mice (collectively Trp53Mut/+) were irradiated with 2.5 Gy, a dose that distinguishes small differences in WT p53 activity (28). The relative mRNA levels of p53 downstream target genes Eda2r, Cdkn1a (p21), Bbc3 (Puma), Bax, and Ccng1 in thymuses were measured by RT-qPCR. In comparison with the Trp53+/ mice, all Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ mice showed comparable decreased activation of these downstream target genes, which were not statistically different from each other (Fig. 1A). No activation of target genes was observed in Trp53/ mice after irradiation, demonstrating that the assay was specific for p53 activity (Fig. 1A). Moreover, the activation of all genes among heterozygous mice was higher than in Trp53/ mice, indicating partial inhibition of WT p53 activity by mutant p53. To explore the IE on p53-induced apoptosis, we performed IHC staining with CC3 antibodies on IR-treated thymuses. Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ mice showed less CC3 staining than Trp53+/ mice but more than Trp53/ mice again indicative of a similar IE of mutant p53 on WT p53 function (Fig. 1B and C). Thus, two different assays indicate that mutant p53 exerts an IE on WT p53.

Figure 1.

Similar IE among p53 mutations in vivo. A, RNA levels of p53 downstream target genes Eda2r, Cdkn1a (p21), Bbc3 (Puma), Bax, and Ccng1 were determined by RT-qPCR 4 hours after 2.5 Gy irradiation in WT (+/+; N = 4 for both non-IR and IR treatment), Trp53+/− (+/−; N = 3), Trp53R172H/+ (R172H/+; N = 5), Trp53R245W/+ (R245W/+; N = 4), Trp53R270H/+ (R270H/+; N = 4), and Trp53–/ (−/−; N = 3 for non-IR and N = 5 for IR) mouse thymuses. B, IHC staining of CC3 was performed on thymus tissues from mice with the indicated genotypes. Scale bar, 100 µm. C, CC3-stained cells were quantified in five random fields. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by the t test.

Figure 1.

Similar IE among p53 mutations in vivo. A, RNA levels of p53 downstream target genes Eda2r, Cdkn1a (p21), Bbc3 (Puma), Bax, and Ccng1 were determined by RT-qPCR 4 hours after 2.5 Gy irradiation in WT (+/+; N = 4 for both non-IR and IR treatment), Trp53+/− (+/−; N = 3), Trp53R172H/+ (R172H/+; N = 5), Trp53R245W/+ (R245W/+; N = 4), Trp53R270H/+ (R270H/+; N = 4), and Trp53–/ (−/−; N = 3 for non-IR and N = 5 for IR) mouse thymuses. B, IHC staining of CC3 was performed on thymus tissues from mice with the indicated genotypes. Scale bar, 100 µm. C, CC3-stained cells were quantified in five random fields. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by the t test.

Close modal

Trp53neo/ mice express approximately 7% of the WT p53 protein (due to insertion of the Neo gene into the WT p53 locus within intron 4 near the internal promoter), yet show delayed tumorigenesis as compared with Trp53/ mice, indicating that even a low level of p53 is tumor suppressive (22). Additionally, Trp53neo/R172H mice have significantly shorter survival than Trp53neo/ mice, indicating an IE of p53R172H on WT p53 during tumor development (22). To compare the IE among all three p53 mutants in this model in vivo, we generated Trp53neo/R172H, Trp53neo/R245W, and Trp53neo/R270H mice (collectively Trp53neo/Mut) and compared with Trp53neo/ mice. To measure transcriptional activity, mice were irradiated at 6 Gy at about 1 month of age, and p53 downstream target genes in the thymuses were examined by RT-qPCR. The 6 Gy dose was required because of the low p53 levels produced by the Trp53neo allele. The expression of Cdkn1a (p21), Bax, and Bbc3 (Puma) was clearly less activated (if at all) in Trp53neo/Mut mice than in Trp53neo/ mice indicative of an IE of mutant on WT p53 (Fig. 2A). To further compare the effects of mutant p53 on tumor development, mice were monitored for spontaneous tumorigenesis. The major tumor types that developed in these mice were lymphomas (about 60%). The median overall survival of Trp53neo/R172H, Trp53neo/R245W, Trp53neo/R270H mice was similar (136, 141, 153 days, respectively) to Trp53/ mice (160 days) and had no statistically significant difference. However, all three cohorts exhibited significantly shorter survival than the Trp53neo/– mice (218 days; Fig. 2B), indicating all three p53 hotspot mutations had a similar IE on WT p53 activity during spontaneous tumorigenesis.

Figure 2.

IE of mutant p53 in vivo. A, IE of mutant p53 by IR treatment in Trp53neo/− and Trp53neo/mut mice. Relative RNA levels of Cdkn1a (p21), Bax, and Bbc3 (Puma) were determined by RT-qPCR in thymuses of 1-month-old Trp53neo/ (Neo/−), Trp53neo/R172H (Neo/R172H), Trp53neo/R245W (Neo/R245W), Trp53neo/R270H (Neo/R270H), and Trp53/ (−/−) mice, 4 hours after 6 Gy IR treatment. At least three mice were used for each genotype. B, Kaplan–Meier survival curves of Trp53neo/ (N = 48), Trp53neo/R172H(N = 36), Trp53neo/R245W (N = 22), Trp53neo/R270H (N = 23), and Trp53/ (N = 19) mice. C, Expression of p53 target genes in Trp53neo/Mut tumors compared with Trp53neo/ tumors was determined by RT-qPCR. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by the t test.

Figure 2.

IE of mutant p53 in vivo. A, IE of mutant p53 by IR treatment in Trp53neo/− and Trp53neo/mut mice. Relative RNA levels of Cdkn1a (p21), Bax, and Bbc3 (Puma) were determined by RT-qPCR in thymuses of 1-month-old Trp53neo/ (Neo/−), Trp53neo/R172H (Neo/R172H), Trp53neo/R245W (Neo/R245W), Trp53neo/R270H (Neo/R270H), and Trp53/ (−/−) mice, 4 hours after 6 Gy IR treatment. At least three mice were used for each genotype. B, Kaplan–Meier survival curves of Trp53neo/ (N = 48), Trp53neo/R172H(N = 36), Trp53neo/R245W (N = 22), Trp53neo/R270H (N = 23), and Trp53/ (N = 19) mice. C, Expression of p53 target genes in Trp53neo/Mut tumors compared with Trp53neo/ tumors was determined by RT-qPCR. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by the t test.

Close modal

Tumors frequently undergo LOH in the context of WT Trp53 alleles. Even though this Trp53neo allele is a hypomorph, we evaluated if there were any differences in Trp53neo allele retention between the Trp53neo/Mut and Trp53neo/ tumors. We assayed for its presence in 30 tumors (6 from Trp53neo/ mice, 10 from Trp53neo/R172H mice, 7 from Trp53neo/R245W mice, and 7 from Trp53neo/R270H mice). All tumors, except 1 Trp53neo/R245W tumor, retained the Trp53neo allele as determined by Sanger sequencing. Because tumors retained a WT p53 allele (albeit a weak one), we were able to examine whether the WT p53 transcriptional program was repressed by mutant p53 in tumors at endpoint. The expression levels of p53 downstream target genes Bax, Ccng1, and Eda2r were expressed significantly lower in Trp53neo/R172H, Trp53neo/R245W, and Trp53neo/R270H tumors in comparison with Trp53neo/ tumors as measured by RT-qPCR (Fig. 2C), supporting IE activity of all three p53-mutant alleles in vivo.

Tumorigenesis among different mutant p53 alleles

To compare the roles of p53R172H, p53R245W, and p53R270H-mutants versus Trp53 loss in spontaneous tumorigenesis with a genotype consistent with LFS patients, a cohort of Trp53R172H/+ (N = 35), Trp53R245W/+ (N = 105), Trp53R270H/+ (N = 38), and Trp53+/− (N = 33) mice with a similar genetic background were generated. Consistent with previous studies (14, 15), the overall survival of Trp53R172H/+ mice did not differ significantly from that of Trp53+/− mice (median survival at 488 and 486 days, respectively). However, the overall survival of Trp53R245W/+ and Trp53R270H/+ mice (median survival at 431 and 419 days, respectively) were significantly shorter than that of the Trp53+/− mice (P = 0.036 and 0.012, respectively) but not that of Trp53R172H/+ mice when the cohorts were monitored for up to 22 months (Fig. 3A). Additionally, when surveyed at 14 months of age, the differences in overall survival between both Trp53R245W/+ and Trp53R270H/+ mice, as compared with Trp53R172H/+ mice were significant (P = 0.005 and 0.0007, respectively; Fig. 3B). These data suggested that Trp53R245W/+ and Trp53R270H/+ mice were more prone to tumor development than Trp53R172H/+ mice, indicating p53R245W and p53R270H-mutants were more potent than the p53R172H mutant in driving spontaneous tumorigenesis in vivo. The major cancer these mice developed were sarcomas, which varied from 37% to 60% (Table 1). Osteosarcomas, in particular, occurred in 17% to 33% of mice and were more prevalent in female mice (see below). Lymphomas were the second most common tumor observed (20%–30%). More carcinomas were observed in Trp53R172H/+ (15%), Trp53R245W/+ (27%), and Trp53R270H/+ (23%) mice as compared with Trp53+/– (8%) mice. In particular, lung adenocarcinomas were found in Trp53R172H/+ (3%), Trp53R245W/+ (11%), and Trp53R270H/+ (17%) mice but not in Trp53+/– mice, indicating mutation-specific tumor types compared with p53 loss. Importantly, metastases of sarcomas and carcinomas were observed only in Trp53Mut/+ and not in Trp53+/− mice indicative of GOF activities for mutant p53 proteins.

Figure 3.

Distinct tumor phenotypes among mutant p53 mice. A and B, Kaplan–Meier survival curves of Trp53+/ (N = 33), Trp53R172H/+ (N = 35), Trp53R245W/+ (N = 105), and Trp53R270H/+ (N = 38) mice at 22 months (A) and 14 months (B). C, LOH analysis in Trp53R172H/+ (N = 31), Trp53R245W/+ (N = 26), and Trp53R270H/+ (N = 14) tumors. D, Kaplan–Meier survival curves of Trp53/ (N = 19), Trp53R172H/R172H (N = 34), Trp53R245W/R245W (N = 139), and Trp53R270H/R270H (N = 44) mice. E, p53 levels were detected by IHC staining in Trp53R172H/+ (N = 19), Trp53R172H/R172H (N = 6), Trp53R245W/+ (N = 16), Trp53R245W/R245W (N = 25), Trp53R270H/+ (N = 20), and Trp53R270H/R270H (N = 13) tumors. Positively stained nuclei were counted by ImageJ software. F, Kaplan–Meier survival curves of homozygous Trp53R245W/R245W and Trp53R270H/R270H mice with 0%–40% and 60%–100% positive p53 nuclei by IHC staining. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by the t test.

Figure 3.

Distinct tumor phenotypes among mutant p53 mice. A and B, Kaplan–Meier survival curves of Trp53+/ (N = 33), Trp53R172H/+ (N = 35), Trp53R245W/+ (N = 105), and Trp53R270H/+ (N = 38) mice at 22 months (A) and 14 months (B). C, LOH analysis in Trp53R172H/+ (N = 31), Trp53R245W/+ (N = 26), and Trp53R270H/+ (N = 14) tumors. D, Kaplan–Meier survival curves of Trp53/ (N = 19), Trp53R172H/R172H (N = 34), Trp53R245W/R245W (N = 139), and Trp53R270H/R270H (N = 44) mice. E, p53 levels were detected by IHC staining in Trp53R172H/+ (N = 19), Trp53R172H/R172H (N = 6), Trp53R245W/+ (N = 16), Trp53R245W/R245W (N = 25), Trp53R270H/+ (N = 20), and Trp53R270H/R270H (N = 13) tumors. Positively stained nuclei were counted by ImageJ software. F, Kaplan–Meier survival curves of homozygous Trp53R245W/R245W and Trp53R270H/R270H mice with 0%–40% and 60%–100% positive p53 nuclei by IHC staining. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by the t test.

Close modal
Table 1.

Tumor spectrum in Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ mice.

Genotype
Trp53+/Trp53R172H/+Trp53R245W/+Trp53R270H/+
Tumor type(N = 25)(N = 34)(N = 46)(N = 24)
Lymphoma 6 (24%) 10 (29%) 19 (30%) 6 (20%) 
Sarcoma 15 (60%) 15 (44%) 23 (37%) 16 (53%) 
 Osteosarcoma 7 (28%) 8 (24%)a 11 (17%)b 10 (33%)c 
 Angiosarcoma 1 (4%) 1 (3%) 6 (10%) 1 (3%) 
 Spindle cell sarcoma 3 (12%) 3 (9%) 3 (5%) 1 (3%) 
 Sarcoma (synovia)    1 (3%) 
 Sarcoma, NOS 4 (16%) 3 (9%) 3 (5%) 3 (10%) 
Carcinoma 2 (8%) 5 (15%) 17 (27%) 7 (23%) 
 Adenocarcinoma     
  Mammary   3 (5%)d 1 (3%) 
  Lung  1 (3%) 7 (11%) 5 (17%) 
  Other 1 (4%) 3 (9%) 6 (10%) 1 (3%) 
 Hepatocellular   1 (2%)  
 Basaloid  1 (3%)   
 Squamous 1 (4%)    
Other tumors 2 (8%) 4 (12%) 4 (6%) 1 (3%) 
 Neuroendocrine tumor   1 (2%) 1 (3%) 
 Leukemia   1 (2%)  
 Baso-squamous tumor  1 (3%)   
 Histiocytic sarcoma 2 (8%) 1 (3%)   
 Myeloma  1 (3%)   
 Granuloma  1 (3%)   
 Hemangioma   2 (3%)  
 Tumor totals 25 34 63 30 
Genotype
Trp53+/Trp53R172H/+Trp53R245W/+Trp53R270H/+
Tumor type(N = 25)(N = 34)(N = 46)(N = 24)
Lymphoma 6 (24%) 10 (29%) 19 (30%) 6 (20%) 
Sarcoma 15 (60%) 15 (44%) 23 (37%) 16 (53%) 
 Osteosarcoma 7 (28%) 8 (24%)a 11 (17%)b 10 (33%)c 
 Angiosarcoma 1 (4%) 1 (3%) 6 (10%) 1 (3%) 
 Spindle cell sarcoma 3 (12%) 3 (9%) 3 (5%) 1 (3%) 
 Sarcoma (synovia)    1 (3%) 
 Sarcoma, NOS 4 (16%) 3 (9%) 3 (5%) 3 (10%) 
Carcinoma 2 (8%) 5 (15%) 17 (27%) 7 (23%) 
 Adenocarcinoma     
  Mammary   3 (5%)d 1 (3%) 
  Lung  1 (3%) 7 (11%) 5 (17%) 
  Other 1 (4%) 3 (9%) 6 (10%) 1 (3%) 
 Hepatocellular   1 (2%)  
 Basaloid  1 (3%)   
 Squamous 1 (4%)    
Other tumors 2 (8%) 4 (12%) 4 (6%) 1 (3%) 
 Neuroendocrine tumor   1 (2%) 1 (3%) 
 Leukemia   1 (2%)  
 Baso-squamous tumor  1 (3%)   
 Histiocytic sarcoma 2 (8%) 1 (3%)   
 Myeloma  1 (3%)   
 Granuloma  1 (3%)   
 Hemangioma   2 (3%)  
 Tumor totals 25 34 63 30 

aOne of 8.

bFive of 11.

cFour of 10 osteosarcomas had metastasis.

dTwo of 3 mammary adenocarcinomas had metastasis.

To investigate whether LOH of the WT Trp53 allele contributes to differences in tumorigenesis among these Trp53Mut/+ cohorts, LOH analyses were performed in Trp53R172H/+ (N = 31), Trp53R245W/+ (N = 26), and Trp53R270H/+ (N = 14) tumors, including lymphomas, carcinomas, and sarcomas (Table 1). For all three cohorts, the WT Trp53 allele was retained in 65% to 86% of tumors while completely lost in 14% to 35%, and the percentages of zero, partial, or complete LOH were statistically similar among the three cohorts (Fig. 3C). Thus, it is unlikely that LOH of WT p53 in tumors contributes to the difference in overall survival among Trp53Mut/+ mice. The shorter overall survival of Trp53R245W/+ and Trp53R270H/+ mice suggested that contact mutations p53R245W and p53R270H are more tumorigenic than p53 loss in a heterozygous background.

Additionally, homozygous Trp53R172H/R172H, Trp53R245W/R245W, and Trp53R270H/R270H mice were monitored for tumorigenesis as well, and the median overall survival (134.5, 154, and 155.5 days, respectively) was similar to Trp53/ (160 days; Fig. 3D). Although similar percentages of lymphoma were observed among all three homozygous cohorts, 4 of 89 Trp53R245W/R245W mice and 2 of 29 Trp53R270H/R2700H mice also developed osteosarcoma, and one mouse from each genotype even had osteosarcoma metastasis (Table 2), which was not observed in Trp53R172H/R172H mice, indicating a distinct role of p53R245W and p53R270H in osteosarcomagenesis. Of note, more tumor types were observed in Trp53R245W/R245W and Trp53R270H/R270H mice, including glioblastoma multiforme, malignant peripheral nerve sheath tumor, and melanoma (Table 2), indicating tumor differences among homozygous mutant mice. Further, the average percentage of p53-positive nuclei (as measured by IHC) among all tumors from heterozygous and homozygous mutant mice was similar (Fig. 3E). To determine if the presence of p53 protein affected overall survival, tumors from Trp53R245W/R245W and Trp53R270H/R270H homozygous mice were categorized into two groups: one group with 0% to 40% p53-positive nuclei, and another group with 60% to 100% p53-positive nuclei (Supplementary Fig. S2A). Strikingly, the overall survival of mice with a higher percentage of p53-positive nuclei in tumors was significantly shorter (median survival at 132 days vs. 205 days; P = 0.0059; Fig. 3F), indicating that detection of p53R245W and p53R270H in tumor cells accelerates tumorigenesis in homozygous mice, thereby supporting GOF contributions to disease progression.

Table 2.

Tumor spectrum in Trp53R172H/R172H, Trp53R245W/R245W, and Trp53R270H/R270H mice.

Genotype
Trp53–/Trp53R172H/R172HTrp53R245W/R245WTrp53R270H/R270H
Tumor type(N = 18)(N = 21)(N = 89)(N = 29)
Lymphoma 15 (83%) 13 (62%) 67 (57%) 17 (59%) 
Sarcoma 3 (17%) 7 (33%) 42 (36%) 8 (28%) 
 Osteosarcoma   4 (3%)a 2 (7%)c 
 Angiosarcoma 3 (17%) 4 (19%) 10 (9%) 1 (3%) 
 Spindle cell sarcoma  2 (10%) 4 (3%) 1 (3%) 
 Rhabdomyosarcoma    1 (3%) 
 Fibrosarcoma    1 (3%) 
 Hemangiosarcoma   6 (5%)  
 Synovial sarcoma   1 (1%)  
 Sarcoma, NOS  1 (5%) 17 (15%) 2 (7%) 
Carcinoma  1 (5%) 2 (2%) 2 (7%) 
 Adenocarcinoma    1 (3%) 
 Other carcinoma  1 (5%) 2 (2%)b 1 (3%) 
Other tumors   6 (5%) 2 (7%) 
 Glioblastoma multiforme   1 (1%)  
 Leukemia   2 (2%)  
 Neuroblastoma   1 (1%)  
 Teratoma   1 (1%)  
 Malignant peripheral nerve sheath tumor    1 (3%) 
 Melanoma   1 (1%) 1 (3%) 
 Tumor totals 18 21 117 29 
Genotype
Trp53–/Trp53R172H/R172HTrp53R245W/R245WTrp53R270H/R270H
Tumor type(N = 18)(N = 21)(N = 89)(N = 29)
Lymphoma 15 (83%) 13 (62%) 67 (57%) 17 (59%) 
Sarcoma 3 (17%) 7 (33%) 42 (36%) 8 (28%) 
 Osteosarcoma   4 (3%)a 2 (7%)c 
 Angiosarcoma 3 (17%) 4 (19%) 10 (9%) 1 (3%) 
 Spindle cell sarcoma  2 (10%) 4 (3%) 1 (3%) 
 Rhabdomyosarcoma    1 (3%) 
 Fibrosarcoma    1 (3%) 
 Hemangiosarcoma   6 (5%)  
 Synovial sarcoma   1 (1%)  
 Sarcoma, NOS  1 (5%) 17 (15%) 2 (7%) 
Carcinoma  1 (5%) 2 (2%) 2 (7%) 
 Adenocarcinoma    1 (3%) 
 Other carcinoma  1 (5%) 2 (2%)b 1 (3%) 
Other tumors   6 (5%) 2 (7%) 
 Glioblastoma multiforme   1 (1%)  
 Leukemia   2 (2%)  
 Neuroblastoma   1 (1%)  
 Teratoma   1 (1%)  
 Malignant peripheral nerve sheath tumor    1 (3%) 
 Melanoma   1 (1%) 1 (3%) 
 Tumor totals 18 21 117 29 

aOne of 4 osteosarcomas had metastasis.

bOne of 2 carcinomas had metastasis.

cOne of 2 osteosarcomas had metastasis.

p53 mutant–specific differences in tumorigenesis and metastasis

Emerging evidence shows that different p53 mutants may have distinct GOF activities (11, 21). To examine allele-specific difference in tumorigenesis, lymphoma-specific survival was first assessed. Lymphomas were prominent in Trp53R172H/+ and Trp53R245W/+ mice, yet very few Trp53R270H/+ mice had lymphoma as the only cancer, which is consistent with a previous study reporting that Trp53R270H/+ in a 129S4/SvJae genetic background showed high incidence of carcinomas and multiple tumors per mouse (15). Therefore, the lymphoma-specific survival of Trp53R245W/+ mice was compared with that of the Trp53R172H/+ mice. Trp53R245W/+ mice with only lymphoma (median survival, 421 days) died significantly faster than Trp53R172H/+ mice (median survival, 518 days; P = 0.0298; Fig. 4A), suggesting p53R245W was more potent in driving lymphoma-specific lethality than p53R172H.

Figure 4.

Mutant allele-specific differences in tumorigenesis. A, Lymphoma-specific survival curves of Trp53R172H/+ (N = 10) and Trp53R245W/+ (N = 9) mice (lymphomas as the only cancer were not observed in Trp53R270H/+). B, Female mouse-specific survival curves for the Trp53R172H/+ (N = 19), Trp53R245W/+ (N = 60), and Trp53R270H/+ (N = 20) mice. C, p53 IHC staining in Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ osteosarcomas. D, Percent metastasis observed in Trp53+/− (0/6), Trp53R172H/+ (1/8), Trp53R245W/+ (5/11), and Trp53R270H/+ (4/10) mice. *, P < 0.05 by the t test.

Figure 4.

Mutant allele-specific differences in tumorigenesis. A, Lymphoma-specific survival curves of Trp53R172H/+ (N = 10) and Trp53R245W/+ (N = 9) mice (lymphomas as the only cancer were not observed in Trp53R270H/+). B, Female mouse-specific survival curves for the Trp53R172H/+ (N = 19), Trp53R245W/+ (N = 60), and Trp53R270H/+ (N = 20) mice. C, p53 IHC staining in Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ osteosarcomas. D, Percent metastasis observed in Trp53+/− (0/6), Trp53R172H/+ (1/8), Trp53R245W/+ (5/11), and Trp53R270H/+ (4/10) mice. *, P < 0.05 by the t test.

Close modal

We also examined survival differences in mice with osteosarcoma as it is the second most common tumor type in Trp53Mut/+ mice, notably, osteosarcomas occur with a higher frequency in female than male heterozygous mice (6 females vs. 2 males for Trp53R172H/+ osteosarcomas, 8 females vs. 3 males for Trp53R245W/+ osteosarcomas, and 10 females for Trp53R270H/+ osteosarcomas), so we chose to restrict the downstream analysis to just female mice. Intriguingly, the overall survival of Trp53R270H/+ female mice (median survival at 417 days) was significantly shorter than that of Trp53R172H/+ female mice (median survival at 456 days; P = 0.046; Fig. 4B; Supplementary Fig. S2B), indicating that the p53R270H mutation is more tumorigenic than p53R172H in female mice. As all osteosarcomas lost the WT Trp53 allele (determined in Fig. 3C), we were able to examine mutant p53 levels by IHC. The number of nuclei with detectable p53-mutant levels was similar among osteosarcomas from Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ mice (Fig. 4C). However, the number of mice with metastasis was clearly higher in Trp53R245W/+ (5/11, 45%) and Trp53R270H/+ (4/10, 40%) mice than in Trp53R172H/+ (1/8, 12.5%) mice (Fig. 4D; Table 1), indicating p53R245W and p53R270H-mutant proteins had stronger GOF contributing to osteosarcoma metastasis than p53R172H. The data combined from analyses of lymphomas and osteosarcomas indicated allele-specific differences in mice with different p53 missense mutations.

The transcriptome of mutant p53-specific osteosarcomas supports allele-specific differences in GOF

The presence of osteosarcoma is a defining feature in LFS patients, and osteosarcomas develop more frequently than any other sarcomas in Trp53Mut/+mice, with more metastasis in Trp53R245W/+ and Trp53R270H/+ mice than in Trp53R172H/+ and Trp53+/− mice. These mouse models provided a platform to examine mutant-specific differences in a single tumor type. One mechanism of mutant p53 GOF is mediated by the interaction with other transcription factors to upregulate tumor-promoting pathways (13). Therefore, RNA-seq analysis was performed to identify mutant p53-driven transcriptomes of primary osteosarcoma tumor samples from Trp53+/− (n = 4), Trp53R172H/+ (n = 6), Trp53R245W/+ (n = 3), and Trp53R270H/+ (n = 5) mice. DEGs were identified in Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ tumors compared with Trp53+/– tumors by significance criteria of adjusted P < 0.05 and fold change > 2. We identified 88 DEGs for Trp53R172H/+, 115 DEGs for Trp53R245W/+, and 180 DEGs for Trp53R270H/+ osteosarcomas (Supplementary Fig. S3A). The majority of DEGs from Trp53R172H/+ (62/88; 70%), Trp53R245W/+ (89/115; 77%) and Trp53R270H/+ (164/180; 91%) osteosarcomas were nonoverlapping (Supplementary Fig. S3B). In addition, distinct individual clusters of DEGs among mutant p53 groups were discernible when compared with the heterozygous p53 group by supervised clustering based on genotype (Fig. 5A). Principal component analysis revealed that osteosarcomas from contact mutations Trp53R245W/+ and Trp53R270H/+ were separated from osteosarcomas from Trp53+/− and Trp53R172H/+ (Fig. 5B). IPA was performed with these three groups of DEGs to identify dysregulated pathways that could potentially contribute to tumorigenesis and metastasis. Again, very few overlapping pathways were observed among these groups (Supplementary Table S3). These data suggest that different p53-mutants drive distinct transcriptomes that may contribute to osteosarcoma progression.

Figure 5.

Different pathways contribute to osteosarcoma metastasis among p53 mutants. A, Supervised clustering based on genotype using the Pearson distance and Ward linkage. B, Principal component analysis of all osteosarcoma RNA-seq data. C, Upregulated DEGs for each mutant were identified by DESeq2 through comparing the individual Trp53Mut/+ tumors with Trp53+/− tumors. The significance criteria were adjusted P < 0.05 and fold change > 2. D, Immunoprecipitation experiments were performed in primary osteosarcoma cell lines with LOH of the p53 WT allele (designated as O). Top, input. Ctl, control antibody for IP; p53, CM5 antip53 antibody. N.S., nonspecific; IP, immunoprecipitation; IB, immunoblot. E, IPA and Enrichr analysis results of top upstream transcription factors regulating the promoters of genes differentially expressed in 14W Trp53Si-1 cells. Ctl, control siRNA; Si-1 and Si-2, Trp53siRNA.

Figure 5.

Different pathways contribute to osteosarcoma metastasis among p53 mutants. A, Supervised clustering based on genotype using the Pearson distance and Ward linkage. B, Principal component analysis of all osteosarcoma RNA-seq data. C, Upregulated DEGs for each mutant were identified by DESeq2 through comparing the individual Trp53Mut/+ tumors with Trp53+/− tumors. The significance criteria were adjusted P < 0.05 and fold change > 2. D, Immunoprecipitation experiments were performed in primary osteosarcoma cell lines with LOH of the p53 WT allele (designated as O). Top, input. Ctl, control antibody for IP; p53, CM5 antip53 antibody. N.S., nonspecific; IP, immunoprecipitation; IB, immunoblot. E, IPA and Enrichr analysis results of top upstream transcription factors regulating the promoters of genes differentially expressed in 14W Trp53Si-1 cells. Ctl, control siRNA; Si-1 and Si-2, Trp53siRNA.

Close modal

As mutant p53 proteins interact with other transcription factors to upregulate genes to promote tumorigenesis and metastasis, a Venn diagram was generated using only upregulated DEGs (Fig. 5C). No common DEGs was found among the three different mutants, further suggesting distinct mechanisms drive GOF in these different p53-mutant mice. To identify transcription factors that might account for these differences, a promoter analysis was performed using three tools (Enrichr, oPOSSUM, and IPA). Due to the differences in the algorithms deployed by these programs, we selected common transcription factors identified from all 3 programs as the strongest probability of binding mutant p53. Using this method, we identified 4 potential cofactors for Trp53R172H/+ and 8 potential transcriptional cofactors each for Trp53R245W/+ and Trp53R270H/+ tumors (Table 3; Supplementary Table S4). Stat3 was identified to interact with both contact mutants, p53R245W and p53R270H, consistent with recent findings that human p53R248Q binds to Stat3, leading to GOF, in colorectal cancers (29). In addition, Yy1, which was previously shown to interact with WT p53 (30), was identified as a cofactor of p53R245W. Because p53R245W preserves protein structure similar to WT p53, it is not surprising to see Yy1 identified as a p53R245W binding protein. Thus, these data provide support for using this method to identify mutant p53 cotranscriptional factors. Novel factors were also identified as potential cotranscriptional factors for specific p53 mutants, including Klf4 and Ctcf for p53R172H; Ctcf, NFfe2l2, Tcf3, and Rest for p53R245W; and Myc, Runx1, Sp1, Max, and Srf for p53R270H (Table 3). Of note, the transcription factors Egr1 and Gata1 (also novel) were identified to be potentially associated with all three p53 mutants, indicating mutant p53 proteins also share some common interacting transcription factors (13).

Table 3.

Potential transcriptional cofactors associated with mutant p53 identified by overlapping Enrichr, oPOSSUM, and IPA analyses.

Potential transcriptional cofactors associated with mutant p53 identified by overlapping Enrichr, oPOSSUM, and IPA analyses.
Potential transcriptional cofactors associated with mutant p53 identified by overlapping Enrichr, oPOSSUM, and IPA analyses.

To validate the cotranscriptional factors associated with mutant p53, immunoprecipitation experiments were performed using primary osteosarcoma cell lines from different p53 germline mutations. Western blot analyses of 7 osteosarcoma cell lines showed three expressed Egr1 and three expressed Stat3, with one cell line expressing both proteins (Fig. 5D, top). Egr1 was immunoprecipitated with p53R172H in H318 cells and with p53R245W in 14W cells. Furthermore, Stat3 was immunoprecipitated with p53R245W in 14W cells and with p53R270H in 26R cells (Fig. 5D, bottom). Clearly, the association of mutant p53 and cofactors was partially determined by the inherent expression of cofactors in the cell lines studied. In addition, although H222 cells have comparable expression levels of Egr1 to H318 and 14W cells, we did not observe association of p53R172H with Egr1, suggesting other factors also affect the interaction. As mutant p53 coimmunoprecipitated both Egr1 and Stat3 in 14W cells, we performed RNA-seq of 14W cells with and without Trp53 siRNA knockdown to further examine whether mutant p53R245W upregulates target genes mediated by Egr1 and Stat3 (Fig. 5E). Many Egr1 and Stat3 target genes were downregulated in Trp53 siRNA knockdown cells compared with scrambled controls (Supplementary Fig. S4). Moreover, using similar criteria to that used to perform promoter analyses of primary osteosarcomas, we identified predicted Egr1 and Stat3 binding sites upstream of the downmodulated promoters, validating the in vitro coimmunoprecipitation experiments (Fig. 5E).

The generation of a germline Trp53R245W mouse model encompassing a major hotspot mutation in LFS patients allowed comparison of its activities with two other hotspot mutations Trp53R172H and Trp53R270H mice in a similar genetic background. Clearly, p53R245W is a loss-of-function mutation because homozygous Trp53R245W/R245W completely rescued the p53-dependent embryonic lethality of Mdm2-null mice. In addition, irradiated tissues from homozygous Trp53R245W/R245W mice showed no p53-dependent transcriptional activity of two p53 targets examined.

Mutant p53 proteins can inhibit WT p53 activity in cell lines. Because culture conditions affect p53 stability and activity, we provide the first comprehensive and side-by-side comparison of three p53 hotspot mutations in vivo. We examined the roles of IE for WT p53 and found that irradiated thymocytes from all three Trp53Mut/+ mice showed similar IE. Similarly, in a compromised p53 background, all three Trp53neo/Mut mice showed similar survival that was significantly shorter than that of Trp53neo/ mice, indicating decreased WT p53 activity by mutant p53. These data, in particular, show that inhibition of WT p53 activity occurs in vivo and contributes to tumorigenesis.

Expression of the human hotspot p53 mutants p53R248W and p53R273H (the human equivalent of our p53R245W and p53R270H contact mutants) in MCF10A cells leads to increased invasion into the lumen in 3D culture than expression of p53R175H (the human equivalent of p53R172H; ref. 31). A somatic mouse mammary tumor model also demonstrated that the Trp53R245W mutation is more tumorigenic than Trp53R172H (21). These data indicate that these two contact mutants have a stronger GOF than the Trp53R172H structural mutation. Although Trp53R270H/+ mice exhibit similar survival to Trp53+/– and Trp53R172H/+ mice in a 129S/SvJae genetic background (15), our study showed that in a C57BL/6J background, Trp53R245W/+ and Trp53R270H/+ mice survive significantly less than Trp53R172H/+ mice, providing additional support for p53 contact mutations (p53R245W and p53R270H) having a stronger GOF than the structural mutation p53R172H in vivo. Mouse background differences on tumor phenotypes in Trp53 heterozygous mice have been previously observed (32). A detailed analysis of the GOF activities of osteosarcomas, a defining feature of LFS, was performed. Osteosarcomas from all Trp53R172H/+, Trp53R245W/+, and Trp53R270H/+ mice studied lost the WT Trp53 allele, indicating that in this tumor type, the presence of WT p53 prevented tumor development.

Our data further showed that while hotspot mutants retained a similar IE, they differed with regard to GOF with stronger activity in the two contact mutants studied. Mutant p53 proteins had distinct GOF mechanisms in driving tumorigenesis and metastasis in vivo. Furthermore, the IE was observed in irradiated thymocytes and in lymphomas of Trp53neo/Mut mice consistent with the previous observation that IE contributes significantly in myeloid malignancies (28). In contrast, GOF appears to be stronger in different tumor types such as osteosarcomas in our Trp53Mut/+ mice (23, 25), in breast (31, 33–35), and in colorectal cancers (29). Thus, it is important to consider mutation and tumor types when evaluating IE or GOF of p53 missense mutations and in consideration of treatment options. Furthermore, our study identified multiple novel p53-associated cotranscriptional factors, such as Stat3 and Egr1, which might contribute to osteosarcoma metastasis. More investigation of these pathways may lead to new therapeutic strategies.

G. Lozano reports grants from NIH and grants from CPRIT during the conduct of the study. No disclosures were reported by the other authors.

S. Xiong: Conceptualization, data curation, software, formal analysis, validation, investigation, visualization, methodology, writing–original draft. D. Chachad: Data curation, software, formal analysis, methodology, writing–original draft. Y. Zhang: Conceptualization, data curation, formal analysis. J. Gencel-Augusto: Data curation, investigation. M. Sirito: Data curation. V. Pant: Data curation. P. Yang: Resources. C. Sun: Data curation. G. Chau: Data curation. Y. Qi: Data curation, software. X. Su: Data curation, software. E.M. Whitley: Data curation. A.K. El-Naggar: Data curation. G. Lozano: Conceptualization, data curation, formal analysis, supervision, funding acquisition, methodology, writing–original draft, project administration, writing–review and editing.

The authors are grateful to the Genetically Engineered Mouse Facility and the Advanced Technology Genomics Core at MD Anderson Cancer Center (supported by the NIH/NCI through grant P30CA016672). They also thank Amando Guerra, Anushree Agrawal, and Nikita Williams for their technical support, and Amanda Wasylishen and Sydney Moyer for helpful discussion. The authors thank Sunita Patterson for editing this article. This work is supported by Cancer Prevention and Research Institute of Texas grant RP170231 and NIH grant CA82577 to G. Lozano and MD Anderson institutional research grant (2015-00051082-Y1) to S. Xiong.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Moyer
SM
,
Wasylishen
AR
,
Qi
Y
,
Fowlkes
N
,
Su
X
,
Lozano
G
.
p53 drives a transcriptional program that elicits a non-cell-autonomous response and alters cell state in vivo
.
Proc Natl Acad Sci U S A
2020
;
117
:
23663
73
.
2.
Eischen
CM
,
Lozano
G
.
The MDM network and its regulation of p53 activities: a rheostat of cancer risk
.
Hum Mutat
2014
;
35
:
728
37
.
3.
Liu
Y
,
Tavana
O
,
Gu
W
.
p53 modifications: exquisite decorations of the powerful guardian
.
J Mol Cell Biol
2019
;
11
:
564
77
.
4.
Levine
AJ
.
The many faces of p53: something for everyone
.
J Mol Cell Biol
2019
;
11
:
524
30
.
5.
Patch
AM
,
Christie
EL
,
Etemadmoghadam
D
,
Garsed
DW
,
George
J
,
Fereday
S
, et al
.
Whole-genome characterization of chemoresistant ovarian cancer
.
Nature
2015
;
521
:
489
94
.
6.
Bertheau
P
,
Lehmann-Che
J
,
Varna
M
,
Dumay
A
,
Poirot
B
,
Porcher
R
, et al
.
p53 in breast cancer subtypes and new insights into response to chemotherapy
.
Breast
2013
;
22
:
S27
9
.
7.
Malkin
D
.
Li-Fraumeni syndrome
.
Genes Cancer
2011
;
2
:
475
84
.
8.
Kansara
M
,
Teng
MW
,
Smyth
MJ
,
Thomas
DM
.
Translational biology of osteosarcoma
.
Nat Rev Cancer
2014
;
14
:
722
35
.
9.
Zhou
X
,
Hao
Q
,
Lu
H
.
Mutant p53 in cancer therapy-the barrier or the path
.
J Mol Cell Biol
2019
;
11
:
293
305
.
10.
Mello
SS
,
Attardi
LD
.
Not all p53 gain-of-function mutants are created equal
.
Cell Death Differ
2013
;
20
:
855
7
.
11.
Hanel
W
,
Marchenko
N
,
Xu
S
,
Yu
SX
,
Weng
W
,
Moll
U
.
Two hot spot mutant p53 mouse models display differential gain of function in tumorigenesis
.
Cell Death Differ
2013
;
20
:
898
909
.
12.
Gencel-Augusto
J
,
Lozano
G
.
p53 tetramerization: at the center of the dominant-negative effect of mutant p53
.
Genes Dev
2020
;
34
:
1128
46
.
13.
Kim
MP
,
Lozano
G
.
Mutant p53 partners in crime
.
Cell Death Differ
2018
;
25
:
161
8
.
14.
Lang
GA
,
Iwakuma
T
,
Suh
YA
,
Liu
G
,
Rao
VA
,
Parant
JM
, et al
.
Gain of function of a p53 hot spot mutation in a mouse model of Li-Fraumeni syndrome
.
Cell
2004
;
119
:
861
72
.
15.
Olive
KP
,
Tuveson
DA
,
Ruhe
ZC
,
Yin
B
,
Willis
NA
,
Bronson
RT
, et al
.
Mutant p53 gain of function in two mouse models of Li-Fraumeni syndrome
.
Cell
2004
;
119
:
847
60
.
16.
Brosh
R
,
Rotter
V
.
When mutants gain new powers: news from the mutant p53 field
.
Nat Rev Cancer
2009
;
9
:
701
13
.
17.
Oren
M
,
Rotter
V
.
Mutant p53 gain-of-function in cancer
.
Cold Spring Harb Perspect Biol
2010
;
2
:
a001107
.
18.
Xu
J
,
Qian
J
,
Hu
Y
,
Wang
J
,
Zhou
X
,
Chen
H
, et al
.
Heterogeneity of Li-Fraumeni syndrome links to unequal gain-of-function effects of p53 mutations
.
Sci Rep
2014
;
4
:
4223
.
19.
Donehower
LA
,
Lozano
G
.
20 years studying p53 functions in genetically engineered mice
.
Nat Rev Cancer
2009
;
9
:
831
41
.
20.
Song
H
,
Hollstein
M
,
Xu
Y
.
p53 gain-of-function cancer mutants induce genetic instability by inactivating ATM
.
Nat Cell Biol
2007
;
9
:
573
80
.
21.
Zhang
Y
,
Xiong
S
,
Liu
B
,
Pant
V
,
Celii
F
,
Chau
G
, et al
.
Somatic Trp53 mutations differentially drive breast cancer and evolution of metastases
.
Nat Commun
2018
;
9
:
3953
.
22.
Wang
Y
,
Suh
YA
,
Fuller
MY
,
Jackson
JG
,
Xiong
S
,
Terzian
T
, et al
.
Restoring expression of wild-type p53 suppresses tumor growth but does not cause tumor regression in mice with a p53 missense mutation
.
J Clin Invest
2011
;
121
:
893
904
.
23.
Xiong
S
,
Tu
H
,
Kollareddy
M
,
Pant
V
,
Li
Q
,
Zhang
Y
, et al
.
Pla2g16 phospholipase mediates gain-of-function activities of mutant p53
.
Proc Natl Acad Sci U S A
2014
;
111
:
11145
50
.
24.
Xiong
S
,
Van Pelt
CS
,
Elizondo-Fraire
AC
,
Liu
G
,
Lozano
G
.
Synergistic roles of Mdm2 and Mdm4 for p53 inhibition in central nervous system development
.
Proc Natl Acad Sci U S A
2006
;
103
:
3226
31
.
25.
Pourebrahim
R
,
Zhang
Y
,
Liu
B
,
Gao
R
,
Xiong
S
,
Lin
PP
, et al
.
Integrative genome analysis of somatic p53 mutant osteosarcomas identifies Ets2-dependent regulation of small nucleolar RNAs by mutant p53 protein
.
Genes Dev
2017
;
31
:
1847
57
.
26.
Jones
SN
,
Roe
AE
,
Donehower
LA
,
Bradley
A
.
Rescue of embryonic lethality in Mdm2-deficient mice by absence of p53
.
Nature
1995
;
378
:
206
8
.
27.
Montes de Oca Luna
R
,
Wagner
DS
,
Lozano
G
.
Rescue of early embryonic lethality in mdm2-deficient mice by deletion of p53
.
Nature
1995
;
378
:
203
6
.
28.
Boettcher
S
,
Miller
PG
,
Sharma
R
,
McConkey
M
,
Leventhal
M
,
Krivtsov
AV
, et al
.
A dominant-negative effect drives selection of TP53 missense mutations in myeloid malignancies
.
Science
2019
;
365
:
599
604
.
29.
Schulz-Heddergott
R
,
Stark
N
,
Edmunds
SJ
,
Li
J
,
Conradi
LC
,
Bohnenberger
H
, et al
.
Therapeutic Ablation of gain-of-function mutant p53 in colorectal cancer inhibits Stat3-mediated tumor growth and invasion
.
Cancer Cell
2018
;
34
:
298
314
.
30.
Sui
G
,
Affar el
B
,
Shi
Y
,
Brignone
C
,
Wall
NR
,
Yin
P
, et al
.
Yin Yang 1 is a negative regulator of p53
.
Cell
2004
;
117
:
859
72
.
31.
Freed-Pastor
WA
,
Mizuno
H
,
Zhao
X
,
Langerod
A
,
Moon
SH
,
Rodriguez-Barrueco
R
, et al
.
Mutant p53 disrupts mammary tissue architecture via the mevalonate pathway
.
Cell
2012
;
148
:
244
58
.
32.
Koch
JG
,
Gu
X
,
Han
Y
,
El-Naggar
AK
,
Olson
MV
,
Medina
D
, et al
.
Mammary tumor modifiers in BALB/cJ mice heterozygous for p53
.
Mamm Genome
2007
;
18
:
300
9
.
33.
Adorno
M
,
Cordenonsi
M
,
Montagner
M
,
Dupont
S
,
Wong
C
,
Hann
B
, et al
.
A mutant-p53/Smad complex opposes p63 to empower TGFbeta-induced metastasis
.
Cell
2009
;
137
:
87
98
.
34.
Do
PM
,
Varanasi
L
,
Fan
S
,
Li
C
,
Kubacka
I
,
Newman
V
, et al
.
Mutant p53 cooperates with ETS2 to promote etoposide resistance
.
Genes Dev
2012
;
26
:
830
45
.
35.
Stambolsky
P
,
Tabach
Y
,
Fontemaggi
G
,
Weisz
L
,
Maor-Aloni
R
,
Siegfried
Z
, et al
.
Modulation of the vitamin D3 response by cancer-associated mutant p53
.
Cancer Cell
2010
;
17
:
273
85
.

Supplementary data