Background
Endometrial cancer (EC) is the most common gynecological malignancy in the USA [
1] and has an increasing incidence and mortality rate predicted globally [
2]. It often arises from the inner lining of the uterine corpus, the endometrium [
3]. Being an estrogen-responsive tissue, unopposed estrogen exposure to postmenopausal endometrium and the increasing amount of circulating estrogen from abdominal fat of obese women inflict EC. Excess body weight and a sedentary lifestyle impose a greater disease risk [
4]. The most established treatment procedures for women presented with EC are surgery, radiation, hormonal therapy, and/or chemotherapy, depending on disease progression. Immunotherapy and targeted therapy also showed promising outcomes in managing disease in specific situations [
5]. The 5-year survival rate of EC with localized tumors is 82% though it is significantly reduced for distant tumors [
6]. In general, the survival rate for common cancers has improved in past decades, except for a few cancers, including EC, which begs attention for a better therapeutic strategy [
1].
The Cancer Genome Atlas (TCGA) published comprehensive genomic and transcriptomic analyses of EC, identifying four molecular subgroups:
POLE ultramutated, microsatellite instability hypermutated, copy-number low, and copy-number high [
7]. Except for the copy-number high, all three subgroups attribute endometrioid phenotype and a high percentage of genetic alterations in the tumor suppressor gene
PTEN. In contrast, the copy number high group is characterized by serous histology with a high frequency of
TP53 gene mutation. Thus the loss of expression of PTEN served as a validated biomarker for the endometrioid subtype- the most common form of EC [
8]. In addition,
PTEN alterations are associated with favorable survival outcomes for EC patients [
9], indicating the PTEN pathway as a potential molecular candidate for the targeted therapy in managing the disease.
Canonical PTEN negatively regulates the PI3K/AKT/mTOR signaling pathway by dephosphorylating the upstream kinase PIP3 through lipid phosphatase activity. Therefore, loss of the PTEN function leads to overactivation of the PI3K/AKT pathway and stimulation of cell proliferation and tumorigenesis [
10]. On the other hand, the noncanonical nuclear functions of PTEN in maintaining genome stability and chromatin organization suggest its role as caretaker of the genome [
11,
12].
Targeting PI3K/AKT/mTOR signaling axis in PTEN-negative EC has limited success in preclinical [
13] and clinical studies [
14], as PI3K inhibitors are often attributed to drug-related cytotoxicity and feedback upregulation of compensatory mechanisms [
15]. Alternatively, targeting either DNA repair function alone [
16] or combining PI3K signaling [
17] might strengthen personalized therapeutic strategies in EC patients with
PTEN’s loss-of-function mutations. In a recent study, we demonstrated that the loss of nuclear PTEN in EC accumulates DNA damage and increases sensitivity to the PARP inhibitor, olaparib [
18].
Since successful targeting of DNA repair functions in EC requires comprehending the complex interaction of PTEN mutations with DNA damage repair (DDR) processes, the current study aims to understand the DNA repair mechanism in PTEN-negative EC. Here we report a functional association of nucleotide excision repair (NER) in EC through noncanonical functions of PTEN.
NER removes many types of DNA lesions from genomic DNA, including UV-induced DNA photolesions. The tight wrapping of genomic DNA around histones impedes DNA repair proteins from accessing DNA lesions buried in nucleosomal DNA. Damage-specific DNA binding protein 2 (DDB2) and Xeroderma pigmentosum complementation group C (XPC) protein complexes detect DNA lesions in nucleosomal DNA in mammalian global genome nucleotide excision repair (GG-NER), which scans the entire genome for damage and initiates repair. DNA damage recognition is the initial critical step influencing the overall efficiency of DNA repair [
19].
Methods
Analysis of public databases
Plasmids and sub-cloning
The pcDNA3-HA-PTEN plasmid was purchased from Addgene (catalog no. 78776). The full-length cDNA of PTEN was subcloned into pEGFP-C1 (Clontech, catalog no. 6084-1) by digestion with Kpnl HF (New England Biolabs, catalog no. R3142S), followed by a ligation reaction using T4 DNA ligase (New England Biolabs, catalog no. M0202S) with the vector pEGFP-C1 digested with the same restriction enzyme. An alkaline phosphatase treatment was performed using a Quick CIP kit (New England Biolabs, catalog no. M0525S) just before the ligation reaction to prevent re-legation of vector and insert. The orientation of the PTEN cDNA insert into the pEGFP-C1 was checked by restriction digestion with Nhel (New England Biolabs, catalog no. R0131), and the clones were confirmed by Sanger sequencing.
Cell culture, transfection, and establishment of stable cell lines
The EC AN3CA cell lines were purchased from American Type Culture Collection (ATCC, catalog no. HTB-111). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Himedia, catalog no. 11330032) supplemented with 10% fetal bovine serum (FBS) (Thermo Fisher Scientific, catalog no. 10270106) and 1% antibiotic-antimycotic (Thermo Fisher Scientific, catalog no. 15240062) at 37 °C in a humidified environment containing 5% CO2. MCF 10A cell line was cultured in DMEM/F-12 (Gibco, catalog no. 11330032) supplemented with 5% horse serum ( Gibco, catalog no. 26050088), 20 ng/ml hEGF (Thermo Fisher Scientfic, catalog no. PHG0311), 0.5 µg/ml Hydrocortisone (Sigma-Aldrich, catalog no. H0888), 100 ng/ml cholera toxin (Sigma-Aldrich catalog no, C8052), 10 µg/ml insulin and 1% antibiotic-antimycotic (Thermo Fisher Scientific, catalog no. 15,240,062) at 37 °C in a humidified environment containing 5% CO2.
AN3CA cells grown to 60% confluency in tissue culture dishes were transfected with either empty vector pEGFP-C1 or pEGFP-C1-PTEN-FL plasmids using the Lipofectamine 3000 (Thermo Fisher Scientific, catalog no. L3000001) transfection reagent as described by the manufacturer.
After transfection, the cells were serially diluted into a 96-well plate, so the last well contained only a single cell. The transfected clones were then selected by growing the cells in the growth medium containing antibiotic geneticin (Thermo Fisher Scientific, catalog no. 10131035). The 96-well plates were screened for GFP-expressing single-cell colonies. AN3CA cell lines stably expressing empty vector pEGFP-C1 (Vector) and pEGFP-C1-PTEN-FL (PTEN-FL) were derived from the colonies displaying the highest GFP expression levels under a fluorescence microscope.
UV irradiation to live cells
Cells were irradiated with varying UV doses (J/m2/sec) using a UV Crosslinker (VWR) equipped with five overhead 8 W UV lamps producing 254 nm wavelength (UVC).
Determination of LD50 and cell death analysis
Vector and PTEN-FL cells were plated on a 96-well plate at a density of 1 × 104 cells/well. The cells were then irradiated with different UV doses (0 J/m2, 2.5 J/m2, 5 J/m2, 10 J/m2, 25 J/m2, 50 J/m2, 100 J/m2, 250 J/m2, 500 J/m2, and 1000 J/m2), followed by further incubation at 37 °C for 4 h. 10 µl of MTT solution (5 mg/ml) (Sigma-Aldrich, catalog no. M5655) was added to 100 µl of media in each well incubated for 3 h at 37 °C. The purplish formazan crystals were dissolved in the dark with 100 µl of solubilizing buffer, DMSO (Sigma-Aldrich, catalog no. 41639). After the crystals completely dissolved, plates were swirled gently to make a uniform color, and absorbance was measured at 570 nm using a microtiter plate reader (Thermo Fisher Scientific). LD50 values were calculated by fitting the data into an inhibitory dose-response curve equation in GraphPad Prism software.
For analyzing the apoptosis, Vector and PTEN-FL cells seeded on 6-well plates were irradiated with a UV dose of 5 J/m2. After 4 h post-irradiation at 37 °C, cells were trypsinized (Thermo Fisher Scientific, catalog no. 25200072) and resuspended in 1 ml ice-cold PBS (Thermo Fisher Scientific, catalog no. 14190235). Annexin binding buffer and annexin V-APC (Thermo Fisher Scientific, catalog no. R37176) were then added to the cell pellets and incubated for 15 min at dark. 5 µl of propidium iodide (PI) (1 mg/ml) (Sigma-Aldrich, catalog no. P4170) was then added and incubated for 5 min at dark. The percentages of apoptotic cells were analyzed by flow cytometry.
Flow cytometric analyses of cell cycle and unscheduled DNA synthesis (UDS)
The NER activity was measured by the amount of repair synthesis detected after UVC-induced damage [
25] in the Vector and PTEN-FL cell lines. Cells were irradiated with a UVC dose of 5 J/m
2. Immediately after UV exposure, cells were incubated with 5 µM click-chemistry compatible EdU (Thermo Fisher Scientific, catalog no. A10044) for 4 h in serum-free media. Cells were then fixed with fix buffer A (300 mM sucrose, 2% formaldehyde, 0.5% triton X-100), and blocked with 10% FBS in PBS, followed by azide coupling reaction for 1 h at dark to conjugate fluorophore Alexa Fluor 647 (Invitrogen, catalog no. A10277). After the reaction, cells were stained with PI in the dark for 20 min at room temperature. Flow cytometry analyses were performed to determine the UDS-positive cells in different cell cycle phases.
Subcellular fractionation and western blot analysis
Subcellular fractionation was performed as described previously [
26]. Vector and PTEN-FL cells grown to 80% confluency were trypsinized and washed with 1 ml PBS. 250 µl from the cell suspension was centrifuged and the pellets were lysed with 2% SDS-lysis buffer (2% SDS, 50 mM Tris-HCl (pH 7.4), 10 mM EDTA) followed by rigorous resuspension. This part of cell suspension was taken as whole-cell extract. The remaining cell suspension were centrifuged and lysed with cytoskeletal (CSK) buffer (300 mM sucrose, 100 mM NaCl, 3 mM MgCl
2, 0.5% Triton-X-100, 1 mM EGTA, 10 mM PIPES) added with protease inhibitor (Roche, catalog no. 04693124001). The cell suspension was centrifuged and the supernatant was taken as soluble part of the cell, and the pellet was again lysed with 2% SDS-lysis buffer, and taken as the chromatin part of the cell suspension. The whole cell lysate and chromatin samples were then boiled and sonicated, followed by quantification and western blotting of the protein samples.
For western blotting, Vector, PTEN-FL and MCF 10A cells were lysed using RIPA buffer, and the total protein was extracted. BCA protein assay kit (Thermo Fisher Scientific, catalog no. 23227) was used to detect total protein concentration. 10 to 25 µg of total protein was loaded for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes. The membranes were cut after the transfer before probing with antibodies according to the molecular weights. PVDF membranes were blocked in 1X TBST solution containing 5% non-fat dried milk or 5% BSA for 1 h at room temperature and incubated with diluted primary antibodies overnight at 4 °C. 1X TBST solutions were used to wash the membranes three times, 5 min each. PVDF membranes were then incubated with secondary antibodies raised against mouse (Jackson ImmunoResearch Laboratories, catalog no. 115-035-062) or rabbit (Jackson ImmunoResearch Laboratories, catalog no. 111-035-045) conjugated with horseradish peroxidase (HRP) for 1 h at room temperature. After washing the membranes three times with 1X TBST, the chemiluminescence detection was carried out using Clarity Max Western ECL Substrate (Bio-Rad, catalog no. 1705062). Using the ImageJ gel analysis tool, the band intensities were measured and values were represented as adjusted densities relative to their respective loading controls.
The primary antibodies used for the analyses were anti-GFP (Cell Signaling Technology, catalog no. 2956), anti-PTEN (Cell Signaling Technology, catalog no. 9188), anti-phospho-serine 473-Akt (Cell Signaling Technology, catalog no. 4051), anti-Akt (Cell Signaling Technology, catalog no. 9272), anti-DDB2 (Novus Biologicals, catalog no. NBP275718), anti-DDB1 (Abcam, catalog no. AB109027), anti-XPC (Novus Biologicals, catalog no. NB100-477), anti-XPB (Novus Biologicals, catalog no. NB100-61059), anti-Caspase-3 (Cell Signaling Technology, catalog no. 9668), anti-β-actin (Sigma-Aldrich, catalog no. A5441), anti-GAPDH (Sigma-Aldrich, catalog no. G9545), anti-H3 (Cell Signaling Technology, catalog no. 4499).
Immunocytochemistry
Cells seeded on sterile 18 mm coverslips in a 12-well plate were subjected to indirect co-immunofluorescence labelling at 50% confluency. The cells were fixed with the buffer containing 4% paraformaldehyde for 15 min and permeabilized with 0.5% of Triton X-100 for 10 min. The cells were then blocked with 5% BSA in PBST for 1 h at room temperature, followed by anti-GFP (Cell Signaling Technology, catalog no. 2956) and anti-PTEN antibody (Cell Signaling Technology, catalog no. 9188) incubation for overnight at 4 °C. After multiple washing, the cells were labelled with Alexa Fluor 488 (Invitrogen, catalog no. A11034) and 633 (Invitrogen, catalog no. A21052) conjugated secondary antibodies for 1 h at room temperature, followed by DAPI staining of the nuclei (Sigma-Aldrich, catalog no. D9542) for 15 min at room temperature in dark. Coverslips were mounted onto glass slides, and confocal microscopy was performed for image acquisition. Image analyses were conducted using ImageJ software.
RNA isolation and RT-qPCR analysis
Total RNA was extracted from the cells using TRIzol reagent (Invitrogen, catalog no. 15596026). The concentration and purity of total RNA were measured using NanoDrop (Thermo Fisher Scientific). Reverse transcription reactions were performed using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, catalog no. 4368814). The cDNA was amplified by gene specific primers using the SYBR Green PCR Master Mix (Thermo Fisher Scientific, catalog no. 4367659) for RT-qPCR in the StepOnePlus real-time PCR system (Applied Biosystems). Primers for DDB2 gene were:
Forward: 5’ GAGCTTCCTGGCCATCTGT 3’.
Reverse: 5’ GGGCAGCCTTTTGTAATATCC 3’.
Primers for DDB1 gene were:
Forward: 5’ ATTGCGGTCATGGAGCTTT 3’.
Reverse: 5’ CAGGATGCAGGCATTGTACTT 3’.
Primers for housekeeping gene GAPDH were:
Forward: 5’ CACCAGGGCTGCTTTTAACTCTGGTA 3’.
Reverse: 5’ CCTTGACGGTGCCATGGAATTTGC 3’.
Chromatin immunoprecipitation
The cells were subjected to ChIP analysis as previously described [
27]. Briefly, AN3CA cells were irradiated with 5 J/m
2 of UV dose. After trypsinization, 1 × 10
6 cells per assay were crosslinked with methanol free 1% formaldehyde (Thermo Fisher Scientific, catalog no. 28908) and sonicated to generate chromatin fragments. Protein-DNA crosslinks were precipitated using antibody against RNA polymerase II carboxy-terminal domain phospho-serine 5 (Active Motif, catalog no. 39233). Primers were designed to measure the enrichment of chromatin fragments by RT-qPCR. SYBR Green reporter intensity was measured relative to the standard curve of input chromatin prepared from Vector and PTEN-FL cells. IgG (Vector Laboratories, catalog no. I-1000-5) was used as the negative control. Primers for
DDB2 promoter were:
Forward: 5’ TGAGCGACAGAGCCAGACC 3’.
Reverse: 5’ CCGAGCTAAGCCAACTTCC 3’.
Statistical analysis
Quantitative data were described as mean ± standard deviation (SD) values of the mean or median values with 95% confidence interval (CI). Representative data from three independent experiments were shown for immunocytochemistry, western blot, and flow cytometry images. Two-group comparisons were conducted by the paired Student’s t-test. Multi-group comparisons were conducted by two-way ANOVA and Tukey’s multiple comparison tests. The specific statistical tests applied to the experiments were mentioned in the respective figure legend. Statistical analyses were performed using GraphPad Prism 9 (GraphPad Software).
Discussion
In this study, we have presented a mechanistic insight into an observation that DDB2 and PTEN expression negatively correlates in EC. This phenomenon is not normal. GTEx data on the disease-free uterine tissues suggests very comparable expression patterns for
DDB2 and
PTEN transcripts [
24]. However, the expression pattern varies along the female reproductive tract due to the tissue specificity. Interestingly, irrespective of sunlight exposure, the normal skin tissues and cultured fibroblast cells show gene expression correlation patterns of
DDB2 and
PTEN similar to EC (Fig.
S1A). The possibility could be that the higher tissue-specific
DDB2 expression is also associated with cell survival functions. Indeed, emerging evidence suggests that DDB2 could regulate chromatin states, gene transcription, cell cycle progression, and protein degradation [
40]. However, its current role in carcinogenesis is debatable as it could act as anti-oncogenic and pro-oncogenic in the context of cancer hallmarks [
40,
41].
We showed that the lack of functional PTEN contributed to the transactivation of DDB2, which, in turn, protects the genome of EC cells through augmented NER activity. This phenomenon could partly be a reason for the favorable survival of EC patients attributed to high
DDB2 expression that we observed (Fig.
S1B) in the Human Protein Atlas portal [
42] for TCGA data [
23]. Nonetheless, the inadequate evidence on endogenous DNA damages that trigger tissue-specific NER makes it challenging to establish a direct relation between NER and EC.
We observed that neither the absence nor the re-expression of PTEN impacted the UVC-induced apoptotic cell population in EC (Fig.
3D). Therefore, the significant difference in LD50 of UVC between Vector and PTEN-FL cells could be the ability of repair synthesis associated with DDB2 expression. The protective nature of DDB2 is not through apoptosis but augmented NER activity, particularly against UV-induced DNA damage, was elegantly demonstrated by Alekseev et al. [
43].
We have not seen any significant G1 or G2/M cell cycle arrest in post-UVC-irradiated cells (Fig.
4E) as at the low dose of UVC (5 J/m
2), which was the case here, the functional NER pathway could quickly and efficiently repair the damages without activating the G1 or the G2/M checkpoints [
44]. Interestingly, we noticed that irrespective of the UVC-irradiation, the accumulation of UDS-positive cells was not significantly different in the G1 phase of both Vector and PTEN-FL cells (Fig.
4F), whereas it was markedly different in the G2/M phase of the cell cycle (Fig
S2B). This discrepancy is presumably because the G1 chromatin is more easily accessible to NER machinery than relatively condensed G2/M chromatin [
45]. Additionally, as we previously mentioned, NER responds faster to low UVC dose-induced damage [
44], which might help bring down the percentages of UDS cells closer to baseline after four hours post-irradiation in the G1 cells.
DDB1, CUL4A/B, and RBX, as part of the UV-DDB, form an E3 ubiquitin ligase complex that ubiquitinates histone H2A after DDB2 recognizes the photolesion on the chromatin. This results in the dissociation of DDB2 from the chromatin via auto-polyubiquitination, thereby allowing the recruitment of XPC. In addition, DDB1-CUL4A/B-RBX stabilizes XPC on damaged chromatin through monoubiquitination [
46‐
48]. Therefore, DDB2 must be rapidly degraded by proteasomes in order to transfer the repair process to XPC [
34,
49]. Consequently, our observation of longer retention of DDB2 after UVC-damage might additionally contribute to impeding NER activity along with a lower DDB2 expression in the presence of PTEN. Moreover, our observation that DDB1 is unaffected by UV irradiation is well supported [
46], as its recruitment at the photolesion site is dependent on DDB2 [
50] and not its expression.
Although tumor suppressor p53 is implicated in the transcriptional activation of DDB2 [
51], its involvement was found to be in the late stage of UV-induced damage when most of the NER activity receded [
46]. In this study, we revealed the functional consequences of UVC-irradiation in the absence and presence of PTEN at the early stage of active NER.
PTEN, however, controls chromatin condensation by interacting with histone H1; thus, loss of PTEN induces the state of abnormal chromatin decondensation that leads to gene activation [
52]. More direct evidence of PTEN-associated transactivation came from the recent studies that demonstrated the interaction of PTEN with the transcription machinery [
37,
38] and regulation of genome-wide occupancy of RNA polymerase II in the context of chromatin by PTEN [
36,
38].
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.