TP53 oncomorphic mutations predict resistance to platinum‑ and taxane‑based standard chemotherapy in patients diagnosed with advanced serous ovarian carcinoma

Abstract

Individual mutations in the tumor suppressor TP53 alter p53 protein function. Some mutations create a non-functional protein, whereas others confer oncogenic activity, which we term 'oncomorphic'. Since mutations in TP53 occur in nearly all ovarian tumors, the objective of this study was to determine the relationship of oncomorphic TP53 mutations with patient outcomes in advanced serous ovarian cancer patients. Clinical and molecular data from 264 high-grade serous ovarian cancer patients uniformly treated with standard platinum- and taxane-based adjuvant chemotherapy were downloaded from The Cancer Genome Atlas (TCGA) portal. Additionally, patient samples were obtained from the University of Iowa and individual mutations were analyzed in ovarian cancer cell lines. Mutations in the TP53 were annotated and categorized as oncomorphic, loss of function (LOF), or unclassified. Associations between mutation types, chemoresistance, recurrence, and progression-free survival (PFS) were calculated. Oncomorphic TP53 mutations were present in 21.3% of ovarian cancers in the TCGA dataset. Patients with oncomorphic TP53 mutations demonstrated significantly worse PFS, a 60% higher risk of recurrence (HR=1.60, 95% confidence intervals 1.09, 2.33, p=0.015), and higher rates of platinum resistance (χ(2) test p=0.0024) when compared with single nucleotide mutations not categorized as oncomorphic. Furthermore, tumors containing oncomorphic TP53 mutations displayed unique protein expression profiles, and some mutations conferred increased clonogenic capacity in ovarian cancer cell models. Our study reveals that oncomorphic TP53 mutations are associated with worse patient outcome. These data suggest that future studies should take into consideration the functional consequences of TP53 mutations when determining treatment options.

Figure 1

Inclusion criteria to study the effect of TP53 mutation type on serous ovarian cancer patient outcomes. Out of 574 serous ovarian cancer patients included in The Cancer Genome Atlas (TCGA) dataset, patients included in our study had uniform platinum- and taxane-based primary chemotherapy. Of those, we determined the TP53 mutation based on exon sequencing and discriminated each mutation into four groups: known oncomorphic, known loss of function (LOF), unclassified, or wild-type (WT). Oncomorphic mutations were designated based on in vivo or in vitro evidence of an oncogenic phenotype. LOF mutations contain a nonsense or a frameshift mutation. Remaining TP53 mutations are designated as unclassified mutations, and have unknown functions.

Figure 2

Landscape of TP53 mutations in the study population. (A) TP53 gene structure and frequency of TP53 mutations at individual codons in patients included in this study. Denoted codons are oncomorphic alterations. (B) Number and frequency of TP53 mutations in study cohort categorized by functional consequence. (C) Normalized protein expression of p53 in serous ovarian cancer tumors in the three functional categories of TP53 mutations. ^*Kruskal-Wallis tests were performed to assess significance. (D) Baseline p53 expression in a panel of nine ovarian cancer cell lines. TA, transactivation domain; PR, proline rich domain; DBD, DNA-binding domain; Oligo, oligomerization domain; LOF, loss of function; Uncl., unclassified.

Figure 3

Oncomorphic TP53 mutations are associated with worse progression-free survival (PFS) and increased risk of recurrence. (A) Plot of PFS. Log-rank test was used to assess significance among the three TP53 mutational categories. Median PFS is noted in inset. (B) Hazard ratio plot showing clinical factors associated with recurrence. No mac, no macroscopic disease; G, grade; LOF, loss of function.

Figure 4

Five-year overall survival rate in patients with oncomorphic and unclassified TP53 mutations. Plots of the Kaplan-Meier estimated cumulative probabilities of overall survival were constructed. Cox proportional hazards regression was utilized to test for differences in progression-free survival (PFS) between mutation types using a study endpoint of 60 months.

Figure 5

Tumors with oncomorphic TP53 are more resistant to chemotherapy than patients with loss of function (LOF) or unclassified mutations. Rates of platinum resistance among The Cancer Genome Atlas (TCGA) cohort and a validation cohort from the University of Iowa. Number of patients (n) are noted in the plot. (B) The most common TP53 mutations were expressed in a cell line (SKOV3, p53 null) to examine the ability of oncomorphic p53 mutant proteins to cause chemoresistance to cisplatin or taxol. Western blot images are combined from two separate gels (demarked by the gray lines separating lanes from different gels). (C) Clonogenic survival of cells stably expressing various TP53 mutant proteins after 48 h treatment with 1 μM cisplatin (Cis) or 10 nM taxol (Tax). ^*P<0.05 vs. empty vector (EV) with the same treatment.

Figure 6

Tumors with oncomorphic TP53 mutations have elevated expression and activity of proteins involved in tumor growth as compared to tumors with unclassified mutations. (A) Reverse phase protein arrays (RPPAs) were used to determine protein expression in The Cancer Genome Atlas (TCGA) analysis. Normalized protein expression was downloaded and compared using a Wilcoxon rank sum test to identify proteins differentially expressed between the two groups. The full dataset is available in Table III. Comparison of RPPA data among oncomorphic, loss of function (LOF), and unclassified TP53 are available in Table IV. (B) Analysis of the expression of proteins identified through the TCGA analysis in a panel of ovarian cancer cell lines. Western blot image of c-Myc contains gray lines demarking re-arrangement of the image, however it is an image of the same blot. (C) Densitometry analysis of β-catenin trended towards higher expression in the panel of cell lines, tested by Kruskal-Wallis test p=0.07.