Sensitivity of ovarian cancer cells to acetaminophen reveals biological pathways that affect patient survival

Introduction

Ovarian cancer (OVCA) is the most lethal gynecological cancer, with an estimated 21,980 new cases and 14,270 deaths in 2014 (1). Unfortunately, 75% of cases are diagnosed at an advanced (III–IV) stage due to the lack of symptomatic early-stage disease or effective screening tests. Although ≤70% of these women demonstrate a complete response to primary treatment with cytoreductive surgery and adjuvant platinum/taxane-based chemotherapy, the majority will eventually develop recurrent disease that is associated with poor survival (2,3). Such patients frequently receive multiple lines of systemic cytotoxic therapy that may be associated with significant toxicity and detriment to the quality of life. The search for additional tolerable but efficacious drugs against OVCA has progressed slowly.

Acetaminophen is a commonly used analgesic with minimal side effects that may be active against OVCA cells (4). Acetaminophen was first synthesized in 1877 by von Mehring at Johns Hopkins University, but it was not extensively used until 1949; it was briefly recalled in 1951 due to reports of blood dyscrasias, but has been marketed since 1955 as an over-the-counter drug (5). Acetaminophen is currently a commonly used analgesic and antipyretic. Adverse events associated with acetaminophen are rare when used at therapeutic dosage; however, overdose may cause fatal hepatotoxicity. Acetaminophen is generally categorized with aspirin and aspirin-like drugs, which are known to act by inhibiting prostaglandin synthesis from arachidonic acid or other fatty acid precursors (6). While the exact mechanism of action for acetaminophen remains controversial, the prevailing hypothesis is that acetaminophen inhibits the synthesis of prostaglandins by competing at the active site of the cyclooxygenase enzyme with arachidonic acid (7).

The epidemiological reports on the effect of acetaminophen on OVCA risk and behavior are conflicting. Moysich et al (8) reported that regular acetaminophen use reduced the risk of OVCA [odds ratio (OR) = 0.56; 95% confidence interval (CI): 0.34–0.86]. Rodriguez-Burford et al (9) were able to demonstrate a decrease in viable OVCA cells with acetaminophen treatment compared with controls. Cramer et al (10) reported a statistically significant inverse association between the use of paracetamol, which is the European name for acetaminophen, and OVCA risk. A more recent Danish case-control study found paracetamol use to be associated with a reduced OR for developing OVCA (OR=0.82, 95% CI: 0.74–0.92; P<0.001) (11). Others, however, have failed to demonstrate such an association (12,13). These inconsistent findings prompted us to investigate: i) The effect of acetaminophen on OVCA cell lines; ii) the molecular signaling pathways associated with acetaminophen response; and iii) in an effort to evaluate the clinical significance of our findings, the associations of these pathways with overall patient survival from OVCA.

Materials and methods

Cell line processing

A total of 16 OVCA cell lines were subjected to treatment with acetaminophen in increasing concentrations (0–10 mM), and the data were used to determine half maximal inhibitory concentration (IC50) values. Prior to any treatment, the 16 cell lines also underwent Affymetrix-based analysis of baseline genome-wide expression. The Pearson's correlation test was used to identify genes that were associated with acetaminophen sensitivity, which were then analyzed for representation of molecular pathways. The genes in the top 2 pathways underwent principal component analysis (PCA) and were assigned a first principal component (PC1) score, which functions as a summary measure of pathway expression. A total of 4 clinico-genomic datasets were then queried using the PC1 scores to assess the associations between acetaminophen response pathways and patient survival from OVCA.

Cell culture

The OVCA cell lines OV90, OVCAR3, SK-OV-3 and TOV112D were obtained from the American Type Culture Collection (Manassas, VA, USA); A2780S was obtained from the European Collection of Cell Cultures (Salisbury, UK); FUOV1, IMCC3, IMCC5, IGR-OV1, MCAS, OVCA420, TOV-21-G, OVCAR2, OVCAR5, PEO1 and PEO4 were kind gifts from Dr Patricia Kruk (Department of Pathology, College of Medicine, University of South Florida, Tampa, FL, USA) and Dr Susan Murphy (Department of Obstetrics and Gynecology/Division of Gynecologic Oncology, Duke University, Durham, NC, USA). The cell lines were maintained in RPMI-1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS; Fisher Scientific, Pittsburgh, PA, USA), 1% sodium pyruvate, 1% penicillin/streptomycin (Cellgro, Manassas, VA, USA), and 1% nonessential amino acids (HyClone, Hudson, NH, USA). The cells were genotyped by short tandem repeat profiling to confirm the tissue of origin, and mycoplasma testing was performed every 6 months, in accordance with the manufacturer's protocol (Lonza, Rockland, ME, USA).

RNA extraction and microarray expression analysis

OVCA cell RNA was extracted at baseline using the RNeasy mini-kit following the manufacturer's recommendations (Qiagen, Valencia, CA, USA). The RNA quality was evaluated using an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA, USA). The targets for Affymetrix microarray analysis were prepared according to the manufacturer's instructions and hybridized to customized Human Affymetrix HuRSTA gene chips (HuRSTA-2a520709), which include 60,607 probe sets and representation of 19,308 genes (Gene Expression Omnibus accession no. GSE34615).

CellTiter-Blue cell viability assays

Drug activity was evaluated using a high-throughput CellTiter-Blue cell viability assay. Cells (2.5×103 per well) were plated in 384-well plates using complete media with 10% FBS and allowed to adhere overnight. After cell adherence, increasing concentrations of acetaminophen were added to appropriate wells using an automated pipetting station. Four replicate wells were used for each drug concentration and for vehicle controls. Drug dilutions initially consisted of serial dilutions from a maximum concentration of 100 mM. The cells were incubated with the drug for 72 h, and 5 µl of CellTiter-Blue reagent (Promega Corp., Madison, WI, USA) were added to each well. Fluorescence was read at 579-nm excitation/584-nm emission using a Synergy 4 microplate reader (Bio-Tek Instruments, Inc., Winooski, VT, USA). IC50 values were determined using a sigmoidal equilibrium model fit (XLfit 5.2; ID Business Solutions Ltd., Alameda, CA, USA). The IC50 was defined as the concentration of drug required for a 50% reduction in cell growth/viability.

Statistical analysis

Expression data from the OVCA cell lines were subjected to background correction and normalization using the Robust Multichip Average algorithm in the Affymetrix Expression Console (http://www.affymetrix.com) and expressed as log2 values. Individual gene expression and IC50 results were evaluated using Pearson's correlation test. Probe sets demonstrating P<0.01 were considered to be significantly correlated with the IC50 results and were analyzed using GeneGo MetaCore™ for pathway analysis (http://www.genego.com/metacore.php). Pathways demonstrating P<0.05 (with a 30% false discovery rate) were considered significant, based on the GeneGo MetaCore™ statistical test for significance.

Clinical significance of pathways

Pathways found to be associated with acetaminophen sensitivity were tested for associations with overall survival using PCA modeling, as previously described (14,15). Survival associations were evaluated by the long-rank test using median PCA value as a cut-off in 4 publically available clinico-genomic datasets from 820 women with OVCA for whom gene expression and overall survival data were available, including i) the Australian dataset (n=218, GSE9891) (16), ii) the MD Anderson dataset (n=53, GSE18520), iii) the Cancer Genome Atlas (n=492), and iv) the Total Cancer Care (TCC®) dataset (n=57).

Results

Genes and pathways associated with acetaminophen sensitivity in OVCA cell lines

Acetaminophen exhibited antiproliferative activity against all tested OVCA cell lines, with IC50 values ranging from 63.2 to 403 µM (Fig. 1). The Pearson's correlation test using acetaminophen IC50 and OVCA cell line genome-wide expression data revealed the expression of 727 unique genes (P<0.01) representing 13 molecular signaling pathways (P<0.01) identified by GeneGo MetaCore™ analysis (Table I).

Figure 1. Line graphs showing acetaminophen dose-response curves for 16 ovarian cancer (OVCA) cell lines (percent survival vs. log acetaminophen dose). SEM, standard error of the mean.

Table I. Biological pathways associated with acetaminophen sensitivity in 16 OVCA cell lines.

Table I.

Biological pathways associated with acetaminophen sensitivity in 16 OVCA cell lines.

Pathway name	P-value	FDR	Pathway objects
Protein folding and maturation/angiotensin system maturation/human version	1.13E-05	6.17E-03	8/43
Cytoskeleton remodeling/neurofilaments	3.76E-04	6.86E-02	5/25
Signal transduction/JNK pathway	6.55E-04	8.96E-02	6/42
Immune response/MIF - the neuroendocrine-macrophage connector	1.07E-03	1.18E-01	6/46
Androstenedione and testosterone biosynthesis and metabolism p.2	1.86E-03	1.29E-01	5/35
Cytoskeleton remodeling/keratin filaments	2.12E-03	1.29E-01	5/36
Effect of low doses of arsenite on glucose-stimulated insulin secretion in pancreatic cells	2.12E-03	1.29E-01	5/36
Development/angiotensin signaling via β-arrestin	3.53E-03	1.86E-01	4/25
Androstenedione and testosterone biosynthesis and metabolism p.3	3.80E-03	1.86E-01	5/41
Neurophysiological process/dopamine D2 receptor transactivation of PDGFR in CNS	4.09E-03	1.86E-01	4/26
Cell adhesion/gap junctions	6.91E-03	2.70E-01	4/30
Regulation of lipid metabolism/FXR-dependent negative feedback regulation of bile acid concentration	7.78E-03	2.84E-01	4/31
Cell cycle/role of Nek in cell cycle regulation	8.72E-03	2.98E-01	4/32

[i] OVCA, ovarian cancer; FDR, false discovery rate; JNK; c-Jun N-terminal kinase; MIF, macrophage migration inhibitory factor; p.2, pathway 2; p.3, pathway 3; PDGFR, platelet-derived growth factor receptor; CNS, central nervous system; FXR, farnesoid X receptor; Nek, never-in-mitosis A-related kinase.

Pathways associated with acetaminophen sensitivity affect overall survival

PCA modeling of the 13 pathways associated with in vitro acetaminophen sensitivity in 4 publically available clinico-genomic datasets identified 4 pathways with expression that may affect overall survival from OVCA (P<0.05): Development/angiotensin signaling via β-arrestin (P=0.04, Australian dataset) (Fig. 2A), protein folding and maturation/angiotensin system maturation (P=0.02, Australian dataset) (Fig. 2B), signal transduction/c-Jun N-terminal kinase (JNK) pathway (P=0.03, Australian dataset) (Fig. 2C), and androstenedione and testosterone biosynthesis and metabolism (P=0.02, TCC dataset) (Fig. 2D).

Figure 2. (A) Expression of the development/angiotensin signaling via β-arrestin; (B) signal transduction/JNK pathway; (C) protein folding and maturation - angiotensin system maturation; and (D) androstenedione and testosterone biosynthesis and metabolism pathways is associated with overall survival from OVCA. Kaplan-Meier curves depict the association between the pathway principal component analysis (PCA) score (using median PCA threshold) and overall survival in OVCA datasets. OVCA, ovarian cancer; JNK, c-Jun N-terminal kinase.

Discussion

There are conflicting epidemiological and in vitro data on the effects of acetaminophen on OVCA. In this study, we demonstrated that acetaminophen exerts a measurable effect on OVCA cells and identified possible molecular determinants of OVCA sensitivity to acetaminophen. In addition, we demonstrated that overall patient survival is associated with the expression of 4 acetaminophen response molecular signaling pathways. This finding suggests a clinical relevance of the identified molecular signaling pathways in OVCA.

Acetaminophen hepatotoxicity is known to be mediated through the JNK pathway (17). Antineoplastic properties have been demonstrated in vitro, where acetaminophen has been shown to enhance the apoptotic effect of cisplatin and paclitaxel in SKOV3 human ovarian carcinoma cells through modulation of intracellular glutathione concentration (4). The epidemiological data vary widely regarding the protective effect of regular acetaminophen use against OVCA development (8–13).

Our study identified an antiproliferative effect of acetaminophen and 4 clinically relevant molecular signaling pathways that are associated with acetaminophen sensitivity in vitro. The Signal transduction/JNK, development/angiotensin signaling via β-arrestin, protein folding and maturation/angiotensin system maturation, and androstenedione and testosterone biosynthesis and metabolism pathways were significantly associated with in vitro sensitivity of OVCA cell lines to acetaminophen, and clinically relevant given their association with overall survival.

JNKs are serine/threonine kinases that belong to the mitogen-activated protein kinase (MAPK) family (18). Phosphorylation of JNK by MAPK kinase 4/7 activates JNK, which may then regulate a number of metabolic and survival pathways, as well as mediate cell death (18,19). Sustained JNK activation imparts an increase in apoptosis and necrotic cell death pathways, while transient JNK activation protects against cell death (20–22). However, the presence of JNK is required for development of gastric cancer, suggesting its role in carcinogenesis (23). The JNK pathway has previously been implicated in acetaminophen-induced hepatotoxicity (17). JNK expression has been shown in a substantial number of OVCA patients and its expression was correlated to progression-free survival, while inhibition of JNK resulted in decreased tumor growth in vivo (24). While sustained activation of JNK leads to an increase in apoptosis of OVCA cells, JNK has been shown to be important in the mechanism of action of acetaminophen and has also been shown to affect OVCA survival and response to therapy (25).

We identified a potential novel implication of the development/angiotensin signaling via β-arrestin, protein folding and maturation/angiotensin system maturation, and androstenedione and testosterone biosynthesis and metabolism pathways. These three pathways have not been previously implicated in OVCA or the mechanism of action of acetaminophen; further investigation is required on the interplay of these biological processes in ovarian carcinogenesis.

In summary, we demonstrated that acetaminophen exerts antiproliferative effects on OVCA cells in vitro. By using an integrated methodology, we identified 4 molecular signaling pathways that appear to be clinically relevant for OVCA. One of these pathways (signal transduction/JNK pathway) was previously found to be involved in the mechanism of action of acetaminophen and to be predictive of decreased survival from OVCA.

Acknowledgements

We would like to thank Rasa Hamilton (H. Lee Moffitt Cancer Center and Research Institute) for editorial assistance. We would like to thank The Foundation of Women's Cancer and Moffitt's Total Cancer Care® Protocol (TCC). TCC is enabled, in part, by the generous support of the DeBartolo Family. Our study also received valuable assistance from the Translation Research Core and Cancer Informatics Core at H. Lee Moffitt Cancer Center and Research Institute; a National Cancer Institute-designated Comprehensive Cancer Center, supported under NIH grant P30 CA76292.

References