Remember me
Research ArticleCell biologyOncology
Open Access | 
10.1172/jci.insight.186409
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Devenport, J. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Tran, T. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Harris, B. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Fingerman, D. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by DeWeerd, R. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Elkhidir, L. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by LaVigne, D. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Fuh, K. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by
Sun, L.
in:
JCI
|
PubMed
|
Google Scholar
|
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Bednarski, J. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by
Drapkin, R.
in:
JCI
|
PubMed
|
Google Scholar
|
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Mullen, M. in: JCI | PubMed | Google Scholar
1Department of Pediatrics,
2Cancer Biology Graduate Program, and
3Molecular Genetics and Genomics Graduate Program, Washington University School of Medicine, St. Louis, Missouri, USA.
4Department of Obstetrics, Gynecology, and Reproductive Sciences, UCSF, San Francisco, California, USA.
5Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA.
6Penn Ovarian Cancer Research Center, Department of Obstetrics and Gynecology, and
7Basser Center for BRCA, Abramson Cancer Center, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA.
8Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Siteman Cancer Center, and
9Center for Genome Integrity, Siteman Cancer Center, Washington University School of Medicine, St. Louis, Missouri, USA.
Address correspondence to: Abby Green, 425 South Euclid Ave., St. Louis, Missouri, 63105, USA. Phone: 314.273.3935; Email: abby.green@wustl.edu.
Find articles by Green, A. in: JCI | PubMed | Google Scholar
Published March 10, 2025 - More info
Published in Volume 10, Issue 5 on March 10, 2025
AbstractHigh-grade serous ovarian cancer (HGSOC) is the most prevalent and aggressive histological subtype of ovarian cancer and often presents with metastatic disease. The drivers of metastasis in HGSOC remain enigmatic. APOBEC3A (A3A), an enzyme that generates mutations across various cancers, has been proposed as a mediator of tumor heterogeneity and disease progression. However, the role of A3A in HGSOC has not been explored. We observed an association between high levels of APOBEC3-mediated mutagenesis and poor overall survival in primary HGSOC. We experimentally addressed this correlation by modeling A3A expression in HGSOC, and this resulted in increased metastatic behavior of HGSOC cells in culture and distant metastatic spread in vivo, which was dependent on catalytic activity of A3A. A3A activity in both primary and cultured HGSOC cells yielded consistent alterations in expression of epithelial-mesenchymal transition (EMT) genes resulting in hybrid EMT and mesenchymal signatures, providing a mechanism for their increased metastatic potential. Inhibition of key EMT factors TWIST1 and IL-6 resulted in mitigation of A3A-dependent metastatic phenotypes. Our findings define the prevalence of A3A mutagenesis in HGSOC and implicate A3A as a driver of HGSOC metastasis via EMT, underscoring its clinical relevance as a potential prognostic biomarker. Our study lays the groundwork for the development of targeted therapies aimed at mitigating the deleterious effect of A3A-driven EMT in HGSOC.
IntroductionOvarian cancer is the fifth most prevalent cancer among women and is the leading cause of death from gynecologic malignancies worldwide (1). High-grade serous ovarian cancer (HGSOC) represents the most frequent and aggressive histological subtype of ovarian cancer, accounting for approximately 70%–80% of ovarian cancer-related deaths (2). HGSOC is often diagnosed at a late stage and exhibits extensive intra- and intertumor heterogeneity, resulting in significant challenges in clinical management (3). Intratumor heterogeneity in HGSOC manifests as diverse subclones with distinct genomic alterations and gene expression profiles within individual tumors. This clonal diversity is associated with treatment resistance and disease progression (2, 4, 5). Despite aggressive therapeutic interventions, most patients with HGSOC experience relapse and diminished survival, highlighting the critical need for a deeper understanding of the molecular drivers of HGSOC disease progression.
Cancer genome sequencing has established that APOBEC3 (apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like 3) enzymes play a significant role in promoting widespread mutagenesis across various tumor types (4, 6–12). The APOBEC3 family of cytidine deaminases (APOBEC3A–APOBEC3H [A3A–A3H]) induce mutagenesis through deamination of cytidine to uracil in single-stranded DNA (ssDNA). APOBEC3 enzymes normally act as virus restriction factors (13). However, aberrant deaminase activity by APOBEC3 enzymes results in damage and mutagenesis to the cellular genome (14–16). Mutations resulting from the enzymatic activity of 2 APOBEC3 family members, A3A and A3B, leave distinct mutational patterns, defined as single-base substitution (SBS) signatures 2 and 13 in the Catalogue of Somatic Mutations in Cancer (COSMIC) database (17). Analyses of tumor genomes have identified APOBEC3 mutagenesis in various cancer types, including breast, bladder, and lung cancer, where is it associated with the accumulation of somatic mutations and the development of subclonal diversity (6, 8, 9, 11, 18–20). Moreover, APOBEC3-mediated mutagenesis has been linked to tumor progression, metastasis, and therapy resistance in breast and lung cancer (19, 21–23).
Detection of APOBEC3 mutational signatures is well established in clear-cell ovarian carcinoma (24–28); however, prior studies have demonstrated a perplexing range in the contribution of APOBEC3 mutagenesis to overall tumor mutation burden in HGSOC. In a study of more than 100 patients with HGSOC, APOBEC3 mutagenesis was found to contribute to approximately 3% of the overall mutational burden in tumor genomes (28). A smaller study assessing the mutational profiles of various gynecologic cancers found that APOBEC3 mutagenesis contributed to a vast range of 0%–70% of the overall mutational burden in 4 patients with HGSOC (25). Thus, the frequency of APOBEC3 mutagenesis in HGSOC remains unclear.
Activity of both A3A and A3B contribute to tumor mutational burden, although several studies have identified A3A as the predominant driver of SBS2 and SBS13 in tumor genomes (8, 29, 30). A germline deletion of A3B, which causes A3A transcript terminating in the A3B 3′UTR, results in increased A3A mutagenesis (31, 32). Epidemiologic studies have demonstrated that this polymorphism is associated with a higher risk of developing tumors, including ovarian cancer (33, 34). The A3B deletion polymorphism is also associated with increased APOBEC3 mutational signatures, indicating elevated A3A activity in tumors (32, 35). However, the contribution of A3A activity to genomic heterogeneity and the biological consequences for HGSOC disease progression are unknown.
In addition to genomic heterogeneity, metastatic progression of HGSOC is driven by epithelial-mesenchymal transition (EMT), a dynamic biological process through which epithelial cells undergo a series of molecular changes that shift the cells toward a mesenchymal phenotype (36). During EMT, cells lose epithelial characteristics, such as cell-cell adhesion and apical-basal polarity, and gain mesenchymal traits such as motility, invasiveness, and resistance to apoptosis (36). EMT is vital during embryonic development, wound healing, and tissue generation but can be hijacked by cancer cells, facilitating metastasis and disease progression (37, 38). In HGSOC, EMT plasticity generates phenotypically diverse cancer cell populations that exhibit epithelial and mesenchymal characteristics simultaneously, often termed hybrid EMT (hEMT) (39). In a large-scale analysis of EMT phenotypes in cancer, an enrichment of the APOBEC3-mediated SBS2 and SBS13 was observed in tumors exhibiting hEMT expression profiles (38).
In this study, we examined the prevalence of APOBEC3 mutational signatures in HGSOC genomes and investigated the consequences of A3A mutagenesis on patient outcomes. Through genome sequencing of HGSOC, we found an enrichment of A3A mutagenesis in metastatic tumors. Elevated levels of A3A mutagenesis significantly correlated with reduced patient survival. By modeling expression of A3A in HGSOC we demonstrated that A3A activity promotes prometastatic phenotypes in cultured cells and distant metastatic spread in murine models. Furthermore, we found that A3A activity led to expression of hEMT and mesenchymal genes in HGSOC cells and metastatic patient tumors, providing a mechanistic explanation for accelerated metastasis. Concordant with this, we identified elevated expression and secretion of IL-6 in HGSOC cells exposed to A3A activity. We found that EMT-associated phenotypes in HGSOC cells exposed to A3A was dependent on IL-6, as evidenced by a reversal of EMT-associated phenotypes upon IL-6 blockade. Our study demonstrates the effect of A3A activity on HGSOC progression via EMT, which may be applicable to other tumors, and identifies potential therapeutic vulnerabilities in tumors with high A3A activity.
ResultsAPOBEC3 mutagenesis is enriched in metastatic high-grade serious ovarian cancer. To determine the prevalence of APOBEC3 activity in HGSOC, we assessed genome sequencing from 3 previously published HGSOC datasets: Pancancer Analysis of Whole Genomes (PCAWG) (40), The Cancer Genome Atlas (TCGA), and Dana Farber Cancer Institute/University of Pennsylvania cohort (DFCI/Penn) (41). For each genome, we assessed the contribution of SBS signatures defined by COSMIC. When averaged across each dataset, we found that the APOBEC3 mutational signatures, SBS2 and SBS13, comprised 4.5%–6.5% of the mutational burden in HGSOC (Figure 1A and Supplemental Figure 1A; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.186409DS1). APOBEC3 mutational signatures did not correlate with specific genetic (i.e., homologous recombination deficiency, cyclin E amplification) or molecular subtypes (42) but were evident in slightly less than half of all HGSOC genomes analyzed (Supplemental Figure 2).
Figure 1APOBEC3 activity correlates with poor survival in patients with HGSOC. (A) Whole genome sequencing was assessed to determine the mutational processes occurring within tumor genomes from PCAWG Ovarian Cancer (PCAWG), TCGA Ovarian Cancer (TCGA), and Dana Farber Cancer Institute/University of Pennsylvania (DFCI/PENN) patient cohorts. Relative contribution of APOBEC3 mutational signatures (SBS2 and SBS13) is shown as a fraction of total mutational burden (PCAWG, 5.9%; TCGA, 4.9%; DFCI/PENN, 7.3%). (B–E) Whole exome sequencing from a patient cohort at Washington University in St. Louis (WUSTL), which includes matched metastatic (M) and primary (P) samples, was assessed for relative contribution of SBS2 and SBS13 (B). For patients with metastatic sites that demonstrated > 4% relative contribution (upper quartile) of SBS2 and SBS13 (patient nos. 4, 7, 14, 15, 17, 18), paired P and M SBS2 and SBS13 fractions are shown. (C) Survival of patients with tumors designated as APOBEC3 high (>4%) versus low (<4%). (D) Patients were grouped by long-term (>5 years) and short-term (<3.5 years) survival. The relative contribution of SBS2+13 in metastatic sites relative to matched primary is shown for individual patients as a heatmap. (E) Analysis of relative contribution of SBS2 and SBS13 shows increased SBS2+13 in primary and metastatic sites for individual patients is shown. Blue dots are from patients with short-term survival, and purple dots represent patients with long-term survival. Statistical significance was determined by uncorrected Fisher’s test, P = 0.02. (F) Genomic assessment of multisite biopsies from patients in the DFCI/PENN cohort. Seven patients with >4% SBS2 and SBS13 relative contribution are shown. Relative contribution of SBS2 and SBS13 shown for each biopsy site ranging from precursor lesions (p53 signatures, serous tubal intraepithelial carcinoma [STIC]) to metastatic sites. Three patients had biopsies of only STIC lesions (64, 65) (patient genome 304). Unlabeled circles and arrows between lesions indicate inferred subclonal hierarchy, adapted from Labidi-Galy, et al. (41), and predicted progression of disease.
In a previous pan-cancer study, APOBEC3 mutagenesis was determined not to be significantly different between primary tumors and metastatic sites (28). To query whether differences in APOBEC3 activity exist between primary and metastatic sites in HGSOC, we took advantage of a dataset from Washington University in St. Louis (WUSTL; St. Louis, Missouri, USA), which includes paired primary and metastatic whole exome sequencing (WES) from 35 patients with FIGO stage III–IV HGSOC prior to treatment (43). When pooling all primary and metastatic samples, we observed that SBS2 and SBS13 were increased in metastatic sites by 6-fold relative to primary sites (Supplemental Figure 1B). We then looked at the individual patients from the WUSTL cohort with the highest contribution of APOBEC3 signature mutations to tumor mutational burden. Six of 35 patients (5.8%) exhibited greater than 4% contribution of APOBEC3 signature mutations in metastatic tumor genomes (Figure 1B and Supplemental Figure 1C). All 6 patients had an increase in the contribution of APOBEC3 signature mutations in metastatic sites compared with primary tumors (Figure 1B). To delineate which APOBEC3 enzyme was most likely responsible for this mutagenesis, we analyzed the extended sequence context in which cytosines were mutated within cancer genomes. Prior studies have demonstrated that A3A mutagenesis occurs within a YTCA sequence context (Y=T or C), whereas A3B acts more frequently at RTCA sequences (R=A or G) (29). We found that mutated cytosines within SBS2/13 signatures in HGSOC genomes more commonly occurred in a YTCA context, consistent with A3A activity (Supplemental Figure 1D).
Given the observed association between APOBEC3 mutagenesis and tumor metastasis, we investigated how APOBEC3 SBS signatures correlated with patient survival. We divided the WUSTL patients by those with high (>4%) or low (<4%) contribution of APOBEC3 signature mutations in metastatic tumors. We found that patients with high APOBEC3 mutagenesis had an average survival of 816.7 days while patients with low APOBEC3 mutagenesis had an average survival of 1,655.3 days, correlating a high burden of APOBEC3 SBS signatures with decreased overall survival (Figure 1C). We next analyzed patients in the WUSTL cohort by categorizing them as short-term (<3.5 years) or long-term (>5 years) survivors (43). We identified that APOBEC3 SBS signatures were more abundant in tumor genomes from short-term compared with long-term survivors (Figure 1D). Upon examination of primary-metastatic pairs, we found that APOBEC3 SBS signatures were enriched in the metastatic sites of short-term survivors compared with long-term survivors (Figure 1E). These data indicate that APOBEC3 mutagenesis is enriched in metastatic HGSOC and correlated with decreased patient survival.
As APOBEC3-mediated mutagenesis increased from primary to metastatic tumors, we sought to determine the kinetics of APOBEC3 activity throughout HGSOC evolution. Within the DFCI/Penn cohort, we assessed multisite biopsies from patients ranging from normal fallopian tube tissue, TP53-mutant single-cell epithelial layer (p53 signature), serous tubal intraepithelial carcinoma (STIC), and primary and metastatic HGSOC lesions (41). In addition, 3 patients with only STIC lesions were analyzed for mutational signatures relative to normal fallopian tube tissue. Of 9 patients in the cohort, 7 had measurable APOBEC3 mutational signatures in at least 1 tumor site (Figure 1F). Interestingly, all 7 patients in whom APOBEC3 mutational signatures were evident had measurable APOBEC3 mutational signatures early in tumor development at the STIC lesion stage, with contributions ranging from 7% to 21%. Four of the DFCI/Penn patients had sequenced biopsies of samples beyond the STIC lesion stage, and 3 of 4 exhibited persistence or increase of APOBEC3 mutagenesis as tumors evolved (9.5%–29% contribution; Figure 1F), consistent with the increased APOBEC3 activity observed in the WUSTL cohort metastatic lesions. These data indicate that APOBEC3 mutagenesis may arise early in HGSOC development but can accelerate or accumulate throughout tumor evolution. In a minority of tumors, APOBEC3-mediated mutagenesis appears not to expand in metastatic lesions. Together, findings from primary tumor genomes demonstrate relatively frequent A3A mutagenesis in HGSOC that is enriched in metastatic sites and associated with poor survival, suggesting that A3A activity enables tumor progression.
Modeling episodic A3A expression in HGSOC cells. To experimentally determine how APOBEC3 mutagenesis affects ovarian cancer progression, we developed cellular models of APOBEC3 expression in HGSOC. We utilized 2 HGSOC cell lines to model episodic A3A activity. A3A is an IFN-stimulated gene (44), and we selected OVCAR3 and OVCAR4, which do not express A3A even when stimulated with IFN (Supplemental Figure 3). Importantly, both cell lines have TP53-inactivating mutations, consistent with the designation of HGSOC. We introduced a doxycycline-inducible (dox-inducible) A3A transgene by lentiviral integration into both cell lines (OVCAR3-A3A and OVCAR4-A3A), enabling controllable A3A expression and activity (Figure 2A and Supplemental Figure 4). A3A mutagenesis has been shown to occur in intermittent bursts over time (45, 46). To mimic intermittent A3A expression in cancer, we treated cells once weekly with low-dose (0.5 mg/mL) dox followed by dox washout 72 hours later (Figure 2A). This approach resulted in A3A expression after 1 day of treatment that gradually led to undetectable levels by day 5 of treatment (Figure 2A). We repeated this treatment course every 7 days for 8 weeks, after which we initiated experimental investigations. This treatment schema provided a system to address how the consequences of historic, intermittent A3A expression, rather than ongoing A3A mutagenesis, affects tumor cell phenotype. To account for the previously established stochastic nature of A3A activity (7), we generated 3 independent replicates of each cell line (A3A V1, A3A V2, A3A V3) (Figure 2B). All A3A replicates were derived from nontreated (NT) control cells, which underwent viral integration of the transgene but were never exposed to dox or A3A expression (Figure 2B). A prior report suggested that A3A expression, but not activity, influenced survival and metastatic spread of pancreatic cancer (47); thus, we generated OVCAR3 and OVCAR4 cells with catalytically inactive A3A transgenes (A3A-C016S) and subjected them to the same treatment schema as in Figure 2A to enable assessment of deaminase activity.
Figure 2Episodic A3A expression promotes HGSOC cell survival, migration, and invasion. (A) OVCAR3 and OVCAR4 cells engineered to stably express a dox-inducible A3A transgene were treated with dox on day 0 and washed out on day 3. Treatment schema was repeated for 8 weeks. Immunoblot of HA-tagged A3A from cell lysates harvested on sequential days throughout 1 week following dox treatment. Ku86 and H3 are loading controls. Bands are quantified relative to loading control and normalized to day 1 lane. (B) Three biological replicates of A3A cell lines (A3A V1–V3) and a cell line induced to express a catalytic mutant of A3A (C106S) were independently derived from the parental cell lines (NT). NT cells were cultured in parallel for 8 weeks. (C) Cell survival under stress was assessed using colony formation assays. Cells were seeded at ultra-low cell densities and resulting colonies were then stained, imaged, and quantified. (D) Wound healing assays were performed to assess the migratory phenotype of OVCAR3 and OVCAR4 NT, A3A V1–V3, and C106S cells. The wound was imaged using a 4× objective at 0 hours and 48 hours. Wound area at 48 hours relative to 0 hours is shown in bar graph. (E) Spheroids of each cell line were generated and placed onto a Matrigel-containing pseudo–basement membrane. Spheroids were imaged with a 10× objective at 0 hours, 24 hours, and 4 days. Area of the spheroid at 24 hours or 4 days relative to 0 hours is shown. Invasion area is outlined in red. For data in C–E, representative images are shown and quantification is depicted in bar graphs below. Significance was determined by 1-way ANOVA with Dunnett’s correction for multiple comparison, ****P ≤ 0.0001, ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05. Data are shown as mean ± SD for n ≥ 3 replicate experiments. Representative images are shown.
Episodic A3A deaminase activity promotes HGSOC cell survival, migration, and invasion. Since APOBEC3 mutagenesis in patient tumors was correlated with decreased survival (Figure 1, C–E), we hypothesized that A3A would promote phenotypic changes consistent with increased metastatic potential. We examined 3 key steps in metastasis — cell survival, migration, and invasion — in A3A-exposed cells. We assessed the ability of HGSOC cells to survive under stress by seeding at ultra-low densities. We found that OVCAR3-A3A and OVCAR4-A3A cells formed significantly more colonies than control (NT) cells (Figure 2C). Increased colony formation was not due to an altered rate of proliferation, since A3A and NT HGSOC cells grew at similar rates (Supplemental Figure 5). In addition, using wound-healing assays, we found that OVCAR3-A3A and OVCAR4-A3A cells migrated more rapidly into a wound than NT controls, indicating enhanced migratory phenotypes (Figure 2D). We analyzed whether A3B expression elicited the same effect and found no significant differences in wound closure after intermittent A3B expression in OVCAR3 cells relative to NT controls (Supplemental Figure 6).
Finally, we assessed how episodic A3A expression affects the invasive potential of HGSOC cells. We generated spheroids of each cell line and placed them in a pseudo–basement membrane (0.5 mg/mL Matrigel). By serial imaging, we found that A3A-exposed OVCAR3 and OVCAR4 cells demonstrated a greater area of invasion into a pseudo–basement membrane than NT controls (Figure 2E). Notably, we found that all OVCAR3-A3A and OVCAR4-A3A cell lines (V1–V3) exhibited similarly altered phenotypes. Importantly, cells exposed to intermittent A3A-C106S expression did not exhibit increased survival, migration, or invasion (Figure 2, C–E), indicating that catalytic activity of A3A is required for this phenotype. Together these data show that episodic A3A activity promotes protumor, metastatic phenotypes in culture indicated by increased survival, migration, and invasion.
A3A accelerates distant HGSOC metastasis in vivo. We reasoned that the phenotypic changes observed in HGSOC cells in culture would affect tumor metastases in vivo. As ovarian cancer progresses in patient tumors, clusters of cancer cells detach from the primary tumor into the peritoneal cavity and seed both local and distant sites (48). Thus, we designed in vivo experiments to mimic ovarian cancer progression in patients by engrafting OVCAR3 or OVCAR4 cell clusters suspended in 0.5 mg/mL Matrigel into the peritoneal cavity of immunodeficient mice. This protocol enabled the formation of tumor spheroids immediately after injection, replicating the behavior of tumor clusters that have migrated from the primary tumor site in patients (48, 49). After delivery of the OVCAR3 or OVCAR4 cells, we monitored tumor burden via bioluminescent imaging (BLI) and found stable engraftment of all cell lines (NT and A3A V1–V3) within 2–3 weeks of injection (Figure 3A). Following stable engraftment of tumor in the peritoneal cavity, we monitored tumor burden through serial BLI for 165 days and found no differences in overall tumor burden between A3A-exposed and NT control xenografts (Figure 3A). Both BLI and postmortem measurement of tumor weights demonstrated a high burden of disease within the peritoneal cavity, consistent with multifocal seeding from i.p. tumor injection (Figure 3A and Supplemental Figure 7).
Figure 3A3A promotes distant HGSOC metastasis in vivo. (A) Representative images of total emission in photons/sec of BLI from day 130 after injection. BLI over time plotted as total emission in photons/sec for each time point; data are shown as median ± SD. (B) Lungs were harvested from the mice at death or 165 days after injection, and tumor burden was determined by macroscopic count of tumor nodules. Representative images are shown; blue arrows indicate macroscopic tumor sites. Significance determined by 1-way ANOVA with Dunnett’s multiple comparison correction. Data are shown as mean ± SD for duplicate experiments (n = 3–5 mice fo
Comments (0)