ES-2 Ovarian Cancer Cells Present a Genomic Profile Inconsistent with their Reported History

Article Information

Eric J Devor1, 2*, Jace R Lapierre1, David P Bender1, 2

1Department of Obstetrics and Gynecology, University of Iowa Carver College of Medicine, Iowa, USAIA 52242

2University of Iowa Hospitals and Clinics, Holden Comprehensive Cancer Center, Iowa, USA

*Corresponding Author: Eric J Devor, Department of Obstetrics and Gynecology, University of Iowa Carver College of Medicine, 463A MRF, 200 Hawkins Drive, Iowa City, Iowa, USA

Received: 23 November 2021; Accepted: 29 November 2021; Published: 06 December 2021


Eric J Devor, Jace R Lapierre, David P Bender. ES-2 Ovarian Cancer Cells Present a Genomic Profile Inconsistent with their Reported History. Obstetrics and Gynecology Research 4 (2021): 233-238.

View / Download Pdf Share at Facebook


ES-2 ovarian cancer cells have long been reported to have originated from a primary clear cell carcinoma of the ovary presenting in a 47 year-old African American patient. Two recent publications have offered evidence calling both of these characteristics into question. Our objective was to further study this cell line using quantitative real-time PCR (qPCR) and mitochondrial DNA (mtDNA) sequencing in order to confirm or refute these inconsistencies. qPCR assays on two characteristic loci, hepatocyte nuclear factor 1β (NHF-1β) and glutathione peroxidase 3 (GPX3), suggest that ES-2 are unusual clear cell carcinoma cells that appear more like high grade serous carci-noma than clear cell. Further, mtDNA haplotyping places the ancestral origin of the patient’s lineage in the Middle East or Europe and not Africa. These results are consistent with and support the conclu-sions of the two recent publications.


ES-2 cell line, qPCR, mtDNA, Haplotyping

ES-2 cell line articles; qPCR articles; mtDNA articles; Haplotyping articles

Article Details

1. Introduction

The cell line ES-2 (RRID:CVCL_3509) is reported to be derived from a poorly differentiated clear cell carcinoma of the ovary (OVCCC) presenting in a 47-year-old African American patient [1]. In 2014, Kwok and colleagues [2] demonstrated that ES-2 cells displayed immunohistochemical staining for hepatocyte nuclear factor 1β (HNF-1β) and Wilm’s tumor 1 (WT1) in ES-2 cells that was more like that seen in OVCAR3 cells, a high grade serous ovarian carcinoma, than in three well known OVCCC cell lines; OVMANA, OVISE and OVSAYO. In addition, evidence indicating that the ancestry of the ES-2 cell line is 58.8% Southern European, 27.1% Northern European and only 2.6% African has been presented [3]. These data call into question what ES-2 cells really are. Here we present additional data based upon qPCR assays and mitochondrial DNA (mtDNA) haplotyping that suggest ES-2 cells are clearly not typical OVCCC and likely not of purely African ancestry.

2. Methods

2.1 Cell line propagation and validation

ES-2 cells were propagated in McCoy’s 5A media with 10% FBS and 1% pen/strep antibiotic at 37oC and 5% CO2. Validation of cells consisted of purifying genomic DNA (gDNA) using the DNeasy Blood and Tissue Kit (QIAGEN) following manu-facturer’s recommendations and sending aliquots of that gDNA to BioSynthesis (Lewisville, Texas) for CODIS genotyping. Total cellular RNA was purified using the mirVana RNA kit (Thermo Fisher) following manufacturer’s recommendations. Yield and purity of the RNAs was assessed on an Agilent Model 2100 Bioanalyzer and a Trinean DropSense 16. These assessments were carried out in the Genomics Core Facility of the Iowa Institute of Human Genetics (IIHG). RNAs were regarded as acceptable if the RNA integrity number (RIN) [4] was at least 9.0 with minimal non-RNA components present.

Additional human cell lines used in this study are the ovarian cancer cells OVCAR-3 (RRID:CVCL_0465), MDAH-2774 (RRID:CVCL_0420), TOV-112D (RRID:CVCL_3612) and the cervical cancer cell line HeLa (RRID:CVCL_0030). All cells were propagat-ed in optimum media according to source recommen-dations; ATCC for OVCAR-3, TOV-112D and HeLa and AddexBio for MDAH-2774. All cell lines have recently been validated (October, 2021) via STR profiling at BioSynthesis (Lewisville, Texas). Total RNA and gDNA purifications from these cells were carried out as described above. Finally, all cell lines are mycoplasma-free.

2.2 Quantitative real-time PCR

Gene-specific qPCR was carried out by first reverse transcribing purified and QC validated total RNA from ES-2 cells using the SuperScript III RT kit (Thermo Fisher). Resulting cDNAs were amplified in the presence of SYBR Green using the primer sequences presented in Table 1. All amplifications were performed on an Applied Biosystems 5700HT platform in the Genomics Core Facility of the Iowa Institute of Human Genetics (IIHG). Gene-specific expression was then normalized against 18S rRNA (Table 1) to generate ΔCt values. Comparisons of fold changes among ovarian cancer cell lines were carried out using the standard ΔΔCt method [5, 6]. Significance of fold changes was evaluated using a standard t-test with unequal variances [7]. A value of p < 0.05 was regarded as significant.

2.3 Mitochondrial DNA (mtDNA) haplotyping

A 428bp segment of hypervariable region 1 (HVR1) of the human mitochondrial genome was amplified using the primer sequences shown in Table 1. Amplicons were purified using the QIAquick PCR Purification Kit (QIAGEN) according to manu-facturer’s recommendations. Purified amplicons were sequenced on an Applied Biosystems Model 3730xl platform in the Genomics Core Facility of the Iowa Institute of Human Genetics (IIHG). The mtDNA haplotype was determined using a MITOMASTER search in MITOMAP (



Amplicon (bp)

Tm (oC)

CDKN1A (p21)






CCNE1 (cyclin E1)


















18S rRNA












Table 1: PCR and qPCR Primer Sequences Used in This Study.

3. Results

3.1 CODIS validation

We routinely CODIS verify all of the cancer cell lines in our laboratory. A fifteen locus CODIS geno-type profile of our ES-2 cells was compared with profiles provided by ATCC, Cellosuarus and ADDEXBIO, two of which are commercial sources of these cells. No discrepancies were seen. This supports a conclusion that our cells are, in fact, ES-2 cells as reported in several sites.

3.2 Quantitative PCR

Ovarian clear cell carcinomas and cell lines present a genomic and expression profile distinct from other ovarian cancer histologies. Unlike other histologies, OVCCC have a very low TP53 mutation prevalence and elevated expression of both p21 (CDKN1A) and cyclin E1 (CCNE1) [8]. Further, OVCCC has charac-teristically high expression of glutathione peroxidase 3 (GPX3) and hepatocyte nuclear factor -1β (HNF-1β) [8]. We confirmed that the cells we were using contained a TP53 mutation, S241F, as reported else-where. Using OVCAR3 high grade serous ovarian cancer cells and MDAH-2774 and TOV-112D ovarian endometrioid adenocarcinoma cells as refere-nces we found that ES-2 cells expressed both CDKN1A (p21) and CCNE1 (cyclin E1) at similar levels (Table 2). HNF-1β is expressed at much higher levels than in MDAH-2774 and TOV-112D cells but, consistent with a previous report [2], ES-2 HNF-1β was slightly lower than that in OVCAR3 cells (Table 2). Similarly, GPX3 expression was much higher than in MDAH-2774 and TOV-112D cells but enor-mously lower than in OVCAR3 cells (Table 2). This expression profile, particularly for the presumeably diagnostic HNF-1β and GPX3 loci, suggests that ES-2 are not typical OVCCC cells.

3.3 mtDNA haplotyping

The mitochondrial HVR1 sequence from ES-2 is identified in MITOMASTER as belonging to the N1b clade. This clade is historically found in the Middle East and Europe. This is consistent with the ancestry reported for this cell line by Dutil and colleagues [3]. A similar analysis of OCVAR3 cells placed them in the European clade H2a, the MDAH-2774 cells in the European U4a clade and the TOV-112D cells in the European K1a clade. These placements are all consistent with their reported ancestries. As a control, we also sequenced an HVR1 mtDNA amplicon from HeLa cells. This sequence was assigned to the African clade L3b, consistent with their ancestry and the ethnicity of the well-known donor of those cells.
























Table 2: Quantitative PCR Comparison of ES-2 with OVCAR3, MDAH-2774 and TOV-112D ovarian cancer cell lines. Values represent fold changes with positive values being higher expression in ES-2 cells and negative values being lower expression in ES-2 for each cell line pair. All comparisons were statistically significant at p < 0.05.

4. Conclusions

It must be emphasized that the data presented here and the questions raised do not in any way diminish the utility of ES-2 ovarian cancer cells as a resource for in vitro studies. However, they do support a previous caution regarding their use as a model for OVCCC histology [2]. Clearly, studies involving cultured cells should examine those cells beyond simply assessing their CODIS genotypes and reported histologic source. Doubtless, however, the vast majority of cultured cancer cells are what they are supposed to be but it’s best to be sure. Finally, the ancestry estimating that we quote here came from a much larger study involving nearly 1400 cancer cell lines including 65 ovarian and 29 endometrial lines. One striking fact that stood out was the dearth of cancer cell lines from patients with African ancestry.

Indeed, only 82 cell lines in that report (5.9%) displayed an African ancestry greater than 50%. Of the ovarian cancer cells only 3 (4.6%) displayed an African ancestry greater than 50% and none of the endometrial cancer cells. This is a short-fall that should be addressed. It has been pointed out that there is a significantly greater prevalence of more aggressive subtypes of endometrial cancers among non-Hispanic Black (NHB) women as compared with non-Hispanic White (NHW), Hispanic and Asian women [9]. This fact translates into significantly lower survival among NHB women diagnosed with endometrial cancer [9-11]. While factors including socio-economic status and access to healthcare are undoubtedly contributing to differences in prevalence and survival, they cannot account for the differences in histology.

The situation in ovarian cancer is much the same in that NHB women have an overall lower survival rate compared with NHW women [12]. Clearly there are racial/ethnic differences in gynecologic cancer incidence, survival and histologic profiles. Since it is likely that some of the reasons for this will be found in examining various gynecologic cancer cell lines, Dutil and colleagues [3] have performed a valuable service by providing the ancestral composition of many of the cell lines that will be used. It has also served to point out that cell lines established from patients with African ancestry are woefully under-represented or not represented at all.


The authors acknowledge the continued support and encouragement from UIHC gynecologic cancer patients and their families, the Department of Obstetrics and Gynecology, and the Iowa Institute of Human Genetics (IIHG) Genome Core Facility, in particular Garry Hauser and Mary Boes, without whom much of this work would not have been possible. Funding of this research was provided in part by the Iowa cancer patient advocacy group CCKMA, the University of Iowa Department of Obstetrics and Gynecology Research Development Fund and the Holden Comprehensive Cancer Center support of the IIHG Genome Core Facility though NIH/NCI P30CA086862.

Conflict of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data used in this study are available from the authors upon reasonable request.


  1. Lau DHM, Lewis AD, Ehsan MN, Sikic BI. Multifactorial mechanisms associated with broad cross-resistance of ovarian carcinoma cells selected by cyanomorpholino doxo-rubicin. Cancer Research 51 (1991): 5181-5187.
  2. Kwok ALM, Wong OG-W, Wong ESY, et al. Caution over use of ES2 as a model of ovarian clear cell carcinoma. Journal of Clinical Pathology 67 (2014): 921-922.
  3. Dutil J, Chen Z, Montiero AN, et al. An interactive resource to probe genetic diversity and estimated ancestry in cancer cell lines. Cancer Research 79 (2019): 1263-1273.
  4. Schroeder A, Mueller O, Stocker S, et al. The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Molecular Biology 7 (2006): 3.
  5. Livak KJ, Schmittgen TD. Analysis of relative gene expression using real-time quantitative PCR and the 2(-Delta-Delta C(T)) Method. Methods 25 (2001): 402-408.
  6. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative Ct method. Nature Protocols 3 (2008): 1101-1108.
  7. Snedecor GW, Cochran WG. Statistical Methods, 8th Edition. Ames, Iowa: Iowa State University Press (1989).
  8. Tsuchiya A, Sakamoto M, Yasuda J, et al. Expression profiling in ovarian clear cell carcinoma. Identification of hepatocyte nuclear factor-1βas a molecular marker and a possible molecular target for therapy of ovarian clear cell carcinoma. American Journal of Pathology 163 (2003): 2503-2512.
  9. Cote ML, Ruterbusch JJ, Olson SH, et al. The growing burden of endometrial cancer: A major racial disparity affecting black women. Cancer Epidemiology, Biomarkers & Prevention 24 (2015): OF1-OF9.
  10. DiSantis C, Naishadham D, Jemal A. Cancer statistics for African Americans, 2013. CA Cancer J Clin 63 (2013): 151-166.
  11. Farley J, Risinger JJ, Rose GS, et al. Racial disparities in blacks with gynecologic cancers. Cancer 110 (2007): 234-243.
  12. Park HK, Ruterbusch JJ, Cote ML. Recent trends in ovarian cancer incidence and relative survival in the United States by Race/Ethnicity and histologic subtypes. Cancer Epidemiology, Biomarkers & Prevention 26 (2017): 1511-1518.

© 2016-2023, Copyrights Fortune Journals. All Rights Reserved