Background
Among all the gynecological cancers, ovarian cancer is considered the cancer with the worst prognosis and the highest mortality rate [
1]. Most patients are diagnosed with advanced-stage tumors [
2] because the manifestation of nonspecific symptoms makes their detection and diagnosis difficult. The late detection of the disease is directly related to its high mortality rate. The diagnosis of ovarian cancer typically involves Ca125 and transvaginal ultrasound; these methods have some limitations, and the need for more specific and sensitive diagnostic tools remains [
3]. New molecules are being studied to determine their relevance for the development of ovarian cancer, improve diagnosis, and treat the disease to reduce patient mortality.
In cancer, there are a wide variety of altered transcriptional regulators that control processes to promote cancer progression and metastasis [
4]. Similar to many other cancers, ovarian cancer is closely associated with genetic disorders [
5]. Therefore, the set of characteristics that give rise to the tumor phenotype, including those that affect the clinical course and the response to therapy, are controlled by deregulated transcriptional programs that operate on tumor cells [
6]. Knowing more about these transcriptional regulators in ovarian cancer could generate useful tools to diagnose, treat, or understand tumor biology.
PHF20L1 (
PHD finger protein 20-like 1) is a transcriptional regulator that is present at low levels in almost all cell types [
7,
8], where it binds to key molecules with which it exerts transcriptional repression [
9]. Approximately 21 splice variants are known, of which three are expressed [
7]. However, to date, its function has not been clearly described. New fucosylated proteins were previously reported as potential biomarkers in ovarian cancer by our laboratory [
10]. In addition to other proteins, PHD finger 20-like protein 1 was found in the list of fucosylated proteins. Studies carried out in breast cancer demonstrated an increase in PHF20L1 expression associated with a poor prognosis. In ovarian cancer, alterations in the gene copy number of PHF20L1 were previously found by another research group [
11]. However, the expression of the protein in ovarian cancer has not yet been analyzed [
12]. It is also unknown whether the tumor microenvironment, namely, malignant ascites from ovarian cancer patients [
13,
14], modulate the expression of this protein. Thus, in the present work, we focus on the relevance of PHF20L1 expression in tumor tissue sections and ovarian cancer cell lines and the correlations between its expression level and the clinical data of ovarian cancer patients. Furthermore, we determined whether ascites could modify PHF20L1 protein expression levels in OVCAR-3 and SKOV-3 cells.
The results showed increased PHF20L1 expression in ovarian cancer tissue compared with healthy ovaries; overexpression showed a negative trend with patient progression-free survival (PFS) and overall survival (OS). When the cell lines were analyzed, we found that the majority of the ascites tested (8/10) stimulated overexpression of the protein. Based on these results, we conclude that PHF20L1 is a protein that can be modulated by ascites components and suggest that its function and expression are important in cancer progression given that its overexpression in tumor tissues was associated with a worse prognosis. However, the conclusions of this work are limited due to the need to analyze a larger number of tissue samples from ovarian cancer patients and to study the association between PHF20L1 expression and the response to treatment, which could provide additional information about the role of the protein in tumor progression.
Material and methods
Biological samples
Paraffin-embedded tissues were used for immunohistochemistry assays. In total, 33 slides with paraffin-embedded sections of tumor tissue were employed. Of these slides, 29 corresponded to tumor tissue sections from patients with different histological subtypes of epithelial ovarian cancer (Table
1), whereas two slides corresponded to cervical cancer and breast cancer. These two last slides were used as a comparison to other cancer tissues. Additionally, two negative controls consisting of cancer-free ovarian tissue with normal appearance obtained from adjacent areas to the tumor were included.
Table 1
Clinical and pathological characteristics of ovarian cancer patients whose tumor tissue sections were employed in the immunohistochemistry analysis
Age (years) |
Mean ± SD | 49 ± 10.65 |
Median (Range) | 50 (28–67) |
CA 125 (U/mL) |
Mean ± SD | 1097.18 ± 1515.6 |
Median (range) | 365 (11.4–5958) |
Clinical stage% (n/N) |
IA | 6.9 (2/29) |
IC | 3.45 (1/29) |
IIIA | 6.9 (2/29) |
IIIB | 3.45 (1/29) |
IIIC | 58.6 (17/29) |
IVA | 13.8 (4/29) |
IVB | 6.9 (2/29) |
Histology% (n/N) |
HGSP | 55.17 (16/29) |
LGSP | 6.9 (2/29) |
Endometrioid | 20.68 (6/29) |
Clear cells | 13.8 (4/29) |
Mucinous | 3.45 (1/29) |
Paraffin-embedded tissues were obtained from patients diagnosed with epithelial ovarian cancer (EOC) and cervical and breast cancer, as appropriate, at the Instituto Nacional de Cancerología (INCan) under the approval of the Research and Ethics Committees (012/018/GII) (CB764/12).
Ascites were obtained from patients diagnosed with epithelial ovarian cancer (EOC) at the Instituto Nacional de Cancerología (INCan) under the approval of the Research and Ethics Committees (009/029/GOI) (CB/549/09). All the samples were used based on the principles of the Declaration of Helsinki regarding the ethical principles of medical research involving human samples, and patients signed the corresponding informed consent form. Ascites were collected by medical personnel. Ascites samples contaminated with blood were excluded from the study. Approximately 100 mL was centrifuged at 1000 rpm for 10 min to recover the cell-free supernatant and the cell pellet. Only cell-free ascites (supernatants) were further employed for experiments and were stored at − 70 °C until use. Before use, ascites were defrosted and warmed at room temperature. The clinical and pathological characteristics of the epithelial ovarian cancer patients whose ascites were used in this project are described in Table
2.
Table 2
Clinical and pathological characteristics of ovarian cancer patients whose ascites were employed in western blot and immunofluorescence analyses
Age (years) |
Mean ± SD | 57 ± 9.9 |
Median (range) | 55 (43–73) |
CA 125 (U/mL) | |
Mean ± SD | 2781.71 ± 2760.741 |
Median (range) | 2225.55 (570.5–13,730.2) |
Clinical stage |
IIIC | 40 (4/10) |
IVA | 10 (1/10) |
IVB | 50 (5/10) |
Histology |
HGSP | 90 (9/10) |
Endometrioid | 10 (1/10) |
The reference cell line SKOV-3 (ATCC HTB-77, purchased in 2010) was used for in vitro experiments. Additionally, OVCAR-3 (ATCC HTB-161, purchased in 2010) cell line was used for comparative purposes in the development of this work. The SKOV-3 model has been used previously in our laboratory as described [
15]. Briefly, SKOV-3 cells were cultured in McCoy’s 5A (Corning, 10-050-CVR) culture medium (15 mL) supplemented with 10% fetal bovine serum (Corning, 35-010-CV) and 1% penicillin/streptomycin (PAA, P11-010) at 37 °C and 5% CO
2. OVCAR-3 cells were grown in RPMI 1640 culture medium (Corning, 10-004-CM) supplemented with 10% fetal bovine serum and bovine insulin (Millipore, I0516). These cultures were grown until cells reached 75% confluence. Then, the medium was discarded before the subsequent addition of ascites (15 mL). Cells maintained in culture medium were used as controls. The cells were maintained under ascites stimulus at 37 °C and 5% CO
2 for the indicated periods. SKOV-3 and OVCAR-3 cells were maintained under ascites stimulus at 37 °C and 5% CO
2 for 24 or 48 h.
SKOV-3 cell line was typified in 2019 by STR-type genetic markers amplified at the National Institute for Genomic Medicine (INMEGEN). Experiments to test for the presence of contaminating mycoplasma in cultures are routinely performed.
Immunohistochemistry
Paraffin-embedded tissues were used to perform PHF20L1 immunohistochemistry. Briefly, tissues were dewaxed overnight at 50 °C. The specimens were hydrated in xylol (3 × 5 min) and decreasing ethanol solutions (100, 95, 90, 85, 80%; 30 s each). Antigenic recovery was carried out using 0.1 N citrate buffer at 121 °C for 20 min. Slides were rinsed 3 × with 1 × phosphate-buffered saline (PBS) for 5 min each. Endogenous peroxidase was blocked by incubating the slide in hydrogen peroxide solution for 1 h at room temperature and washed again with 1 × PBS (5 × 5 min). PBS-10% milk was added for 1 h at room temperature to block nonspecific binding. Then, the primary anti-PHF20L1 antibody (Sigma-Aldrich HPA028417) (1:100 dilution, 1 × PBS-SFB 10%) was added with overnight incubation at 4 °C. The next morning, the slides were rinsed 5 × with 1 × PBS, and the secondary horseradish peroxidase-conjugated anti-rabbit IgG antibody (HRP) was placed in a 1:100 dilution and incubated for 2 h at room temperature. Corresponding washes were performed, and diaminobenzidine substrate (Invitrogen, Catalog number 002020) was added to visualize the reaction followed by incubation for 5 min. Finally, hematoxylin counterstaining was performed, and the slides were fixed for observation under a light microscope (Nikon, E100).
Immunohistochemistry evaluation of PHF20L1 positivity (immunoreactive score)
The QuPath program (
https://github.com/qupath) was used to quantify the number of positive cells and the signal intensity. The determination of the positive signal was assigned by the software, and three levels of signal intensity are indicated: slightly positive (1 +), moderately positive (2 +), and strongly positive (3 +). Areas without staining were classified as negative (Neg).
Additionally, to generate the survival curves, the immunoreactive scores were calculated for each slide [
16]. For this analysis, values were assigned to the percentage of positive cells (A) and the intensity of the signal (B); the result of the A value multiplied by the B value, which ranges from 0 to 12, allows us to classify the expression as negative, weak, moderate or intense.
Treatment of SKOV-3 cells with ovarian cancer ascites and cellular protein extract
After the treatment of SKOV-3 cells (24 or 48 h) with ten different ascites samples and the control condition (culture medium), cell cultures were washed with sterile 1 × PBS and were subsequently recovered using trypsin solution (Corning, 25-051-CI). Cell lysis was carried out using RIPA buffer (5 mM Tris–HCl, 2 mM EDTA, 50 mM NaCl, and 1% Nonidet P-40) with a cocktail of protease and phosphatase inhibitors (1 mg/ml aprotinin and leupeptin and 1 mM PMSF, NaF, and Na3VO4). Finally, the samples were quantified using a DC Protein Assay Kit (Bio-Rad, 500-0114).
SDS-PAGE and Western blotting
The total protein extracts (60 µg) were visualized by 10% SDS-PAGE. The gel was transferred to a nitrocellulose membrane. The nitrocellulose membrane was blocked with 5% milk for 1 h at room temperature. Then, the membrane was washed 3× with TBS-Tween 0.1% for 15 min each time and incubated with the antibody against PHF20L1 (Sigma-Aldrich HPA028417) (1:2000 dilution in TBS-T 0.1% with milk 1%) overnight at 4 °C. Then, the membrane was washed 3× with TBS-T (containing 0.1% Tween) and incubated with HRP goat anti-rabbit (IgG) secondary antibody (1:10,000 dilution in 0.1% TBS-T with 1% milk) for 1 h at room temperature. Again, the membrane was washed 3× with TBS-T. Finally, the cells were incubated with Super Signal West Femto (Thermo Fisher Scientific, 34094) for visualization.
As a loading control for western blot to normalize the levels of protein detected, a GAPDH antibody was employed (1:15,000 dilution in 0.1% TBS-T with 1% milk) (Genetex GTX100118).
Immunofluorescence
SKOV-3 and OVCAR-3 cells adhered to coverslips were subjected to different treatment conditions. Upon completion, the cells were fixed with 4% paraformaldehyde for 1 h at 37 °C. Cells were rinsed 3× with filtered 1× PBS and permeabilized with a 0.2% solution of Triton X-100 in 1× PBS for 15 min at room temperature with stirring. Cells were rinsed thrice with filtered 1× PBS and blocked with 10% FBS for 1 h at 37 °C. The primary antibody (Sigma-Aldrich HPA028417) was added at a 1:100 dilution in 1× PBS and incubated overnight at 4 °C in a humid chamber. On the next day, cells were rinsed 3× with 1× PBS. TRITC goat anti-rabbit (IgG) secondary antibody (ab50598, Abcam) was added to OVCAR-3 cells and incubated for 1 h at 37 °C under protection from light. For the SKOV-3 assays, a different secondary antibody was employed: Alexa Fluor 647 donkey anti-rabbit (IgG) secondary antibody (1:100 dilution) (711-605-152, Jackson InmunoResearch Laboratories), this with the objective to improve the signal of the protein. Vectashield with DAPI (Cat. No. H-1200) was added, and the coverslip was mounted on a slide. The assembled samples were stored at − 20 °C protected from light until observation under a Carl Zeiss LMS 700 confocal microscope. ZEN 2012 software was used to analyze the samples (Carl Zeiss).
Statistical analysis
Densitometric analysis was performed using ImageJ software. Statistical analyses were performed using GraphPad Prism 7 software. For western blot comparisons, Student’s t-test was performed. The Spearman’s correlation test was performed to assess the correlation between PHF20L1 expression and the clinical data of the patients, and the value of ρ was reported in the corresponding tables. Student’s t-test comparatively evaluated the control and problem conditions. Comparisons between groups were performed using ANOVA. The average and the corresponding standard deviations of in vitro assays performed in triplicate are included in the graphs. Significant differences are indicated as follows: *ρ < 0.01, **ρ < 0.001, ***ρ < 0.0001.
Patient survival was analyzed using the Kaplan–Meier method. Overall survival (OS), which is defined as the length from time of diagnosis to death or the last follow-up, and progression-free survival (PFS), which is defined as the time length from a random assignment in a clinical trial to a diagnosis of progressive disease or relapse, of the 29 patients with EOC were determined. Survival data of patients alive without progression or those who died due to other diseases were censored.
Expression values (arbitrary units obtained from the densitometric analysis) obtained from western blotting data were used to categorize patients into two groups: low or high expression of any of the isoforms. In each cohort, each patient was further classified as follows:
For the “a” isoform:
Low: value from 0.3 to 0.45.
High: Value from 0.57 to 0.76.
For the "b" isoform:
Low: value from 0.72 to 0.8
High: value from 0.93 to 1.36.
For the "c" isoform:
Low: value from 0.05 to 0.15.
High: value from 0.22 to 0.52.
Discussion
EOC is the deadliest gynecological cancer. A prerequisite for reducing the number of deaths involves the availability of adequate diagnostic and monitoring tools that allow physicians to make accurate and timely decisions. In this work, PHF20L1 protein was analyzed because it was previously identified as a fucosylated protein in ovarian cancer cell lines with an aggressive phenotype [
10] to evaluate its potential as a key component in the pathogenesis and/or as a possible biomarker of ovarian cancer. Studies have assessed the expression of this protein in breast cancer tissues [
9,
12], but no study has assessed PHF20L1 expression in EOC prior to ours [
10].
Immunohistochemical detection of PHF20L1 in 29 tumor tissues from different patients with epithelial ovarian cancer showed increased expression of the protein (Fig.
1; Additional file
2: Fig. S2) compared to control tissues (cancer-free tissues) where the expression was very low (the immunoreactive score values were considered negative and low for these two tissues) (H1 and H2, Additional file
6: Table S1). PHF20L1 in the samples in our study was also compared with that reported in the Human Protein Atlas database, which indicated that the expression was confined to the stroma and follicular cells in a healthy ovary [
8]. It is important to mention that we found discrete expression in the stroma, but the expression was less evident than that noted in the remainder of the tissue sample. These results strongly suggest an increase in PHF20L1 protein expression in EOC tissue (Additional file
7: Fig. S6).
The interaction of SOX-2 with PHF20L1 has already been demonstrated by others [
23]; SOX-2 and PHF8, another PHD finger protein, have been analyzed in ovarian cancer tumor tissues [
22], showing that SOX-2 was overexpressed in mucinous carcinoma. In contrast, PHF8 was found to be overexpressed in clear cell carcinoma compared to the other histotypes. Both genes, SOX-2 and PHF8, were also detected in the cells of the epithelial surface of a normal ovary [
22]. However, as our results show, the overexpression of PHF20L1 was not associated with any particular histotype.
Our results suggest that patients with higher PHF20L1 expression exhibit a shorter progression-free survival period compared with patients with lower protein expression, where the progression-free survival period was longer. In the OS analysis, we noted a trend for patients with higher protein expression to have a shorter survival time. However, the differences were not significant. This was also the case when we analyzed the isoforms by western blotting, even though no significant differences were found in any of the isoforms. Another aspect to consider is the sample size or that different clinical factors could influence the behavior of the correlation with these survival parameters. Thus, it would be necessary to increase the number of samples representing different clinical stages. We propose that if the sample size is increased, true significance would be observed. Nevertheless, based on our results, we strongly suggest an association between the overexpression of PHF20L1 and poor prognosis in ovarian cancer patients.
Protein isoform expression was analyzed in in vitro assays using SKOV-3 cells stimulated with ascites from patients with EOC (Fig.
4). Ascites from EOC patients exhibited modulation of PHF20L1 isoform expression, both at 24 and 48 h of ascites stimulation. According to previous work in our laboratory, cells stimulated with ascites acquire characteristics of a more aggressive phenotype compared with cells not in contact with this fluid [
15], and this finding indirectly supports the association between PHF20L1 and this aggressive phenotype. Regarding the modulation of PHF20L1 expression, in a recently published article, it was suggested that the MYC genes and genes that respond to hypoxia, such as HIF1α, could regulate the expression of PHF20L1 [
9]. MYC is a gene frequently involved in cancer [
24], and it is one of the four genes (in addition to Oct4, Sox2, and Klf4) that can reprogram fibroblasts to become pluripotent stem cells [
25‐
27]. MYC expression is regulated by the activation of different signaling pathways, including the WNT pathway, through receptor tyrosine kinase and TGF-β [
26]. As TGF-β is one of the main components of ascites [
28], it is possible to hypothesize that, through MYC regulation, PHF20L1 protein could be overexpressed upon ascites stimulation. However, in tumor tissue, where we also observed high expression of PHF20L1, its regulation could be associated with both TGF-β, which is overexpressed in ovarian cancer tumor tissue [
29], and with other components given that signaling via WNT or receptor tyrosine kinase would make more sense [
30]. It will be necessary to perform additional experiments under the induction of TGF-β, for example, or with an MYC knockdown animal model to corroborate whether MYC absence would affect the expression of PHF20L1.
On the other hand, ascites fluid stimulates the expression of all the isoforms of PHF20L1 since very little expression is found in medium SKOV-3 cells, all the isoforms were overexpressed at 48 h, and all of them were expressed from 24 h, apart from isoform c, which is almost not expressed some of the ascites (Fig.
4). Studies performed to date by other groups do not highlight any particular isoform’s relevance, except for “c” isoform, which was reported to interact with the DNMT1 protein [
20].
However, based on our results, we suggest that PHF20L1 could eventually represent an EOC biomarker. To achieve this goal, it will be necessary to increase the number and clinical stages of EOC samples to analyze the expression level of PHF20L1; furthermore, correlation analyses with the expression of additional markers, such as CA125 or Ki67, in tumor tissues should be performed given that these molecules may indicate recurrence or response to therapy (CA125) or a high rate of cell proliferation (Ki67).
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.