Background
Ovarian cancer is one of the most common malignant tumors of the female reproductive system. The currently established treatment methods include surgery, platinum and paclitaxel-based chemotherapy. Although the widespread use of molecular targeted therapies have helped improve the survival rate of some ovarian cancer patients, the five-year survival rates have remained within the range of 25%–30% [
1‐
3]. Therefore, identification of efficient biomarkers is a key imperative to provide a basis for early diagnosis, targeted therapy, and precise prognostic assessment.
The ubiquitin–proteasome system is mainly composed of ubiquitin, ubiquitin-activating enzymes(E1s), ubiquitin-conjugating enzymes(E2s), ubiquitin ligases(E3s), 26S proteasome and deubiquitinating enzymes, which promote degradation of damaged proteins, and regulate growth and stress responses. The ubiquitin–proteasome system is an indispensable mechanism for regulating intracellular protein degradation, thereby affecting human antigen processing, signal transduction, and cell cycle regulation [
4]. The 26S proteasome consists of a proteolytically active cylindrical particle (20S proteasome) and one or two ATPase-containing complexes (called 19S cap complexes), which are divided into ATPase subunits (PSMC 1 ~ 6) and non-ATPase subunits (PSMD1-14) [
5‐
8]. Recent studies have shown that dysfunction of the ubiquitin–proteasome system can lead to up-regulation and/or down-regulation of
PSMDs genes, and abnormal gene expression is often associated with oncogenes and tumor suppressor genes that regulate tumors [
9‐
11]. As ATP independent molecules, PSMDs complete biological functions without external energy input and obvious conformational changes, and play an important role in a variety of cancers, which attracted our attention. The expression of the
PSMDs has been first studied in breast cancer and bladder cancer, but its carcinogenic or anticancer effect in ovarian cancer has not been reported. Therefore, in this study, we evaluated the role of
PSMD family in ovarian cancer. Our findings may provide a theoretical basis for the early diagnosis, prognostic assessment, and targeted therapy of ovarian cancer.
Materials and methods
GEPIA dataset analysis
The GEPIA database (
http://gepia.cancer-pku.cn/) [
12] contains data pertaining to 9736 tumor samples and 8587 normal samples, based on the UCSC Xena data. We used the GEPIA dataset to validate
PSMDs in ovarian cancer and assessed the differences in mRNA expressions between ovarian cancer tissues and adjacent normal tissues.
P < 0.05 was considered indicative of statistical significance.
TISIDB analysis
The TISIDB database (
http://cis.hku.hk/TISIDB) [
13] is a resource for immunology data including high-throughput screening data, tumor immune-related genes, molecular profiles, and paracancerous multi-omics data. The database can be used to assess the correlation of genes with lymphocyte subsets, immunomodulators, chemokines, etc. We used this database to analyze the relationship between
PSMDs expression and clinical stage in ovarian cancer, and to explore the correlation of
PSMD8 with lymphocytes and immunomodulators.
TCGA and cBioPortal analysis
cBioPortal is an open database based on the TCGA database (
www.cbioportal.org) [
14] for interactive exploration of multiple cancer genomics datasets. The database contains data pertaining to DNA copies, DNA methylation, mRNA and microRNA expression, non- synonymous mutation, and other data. In this study, we used this database to analyze the gene variation of
PSMDs, and evaluate co-expression correlation between members of
PSMDs.
Kaplan–Meier Plotter analysis
As a biomarker evaluation tool, the Kaplan–Meier plotter (
http://kmplot.com) [
15] allows the evaluation of the prognostic significance of molecular biomarkers in cancer samples in terms of survival outcomes in patients with breast, ovarian, lung, gastric, and other cancer types. We analyzed the prognostic value of
PSMDs mRNA in ovarian cancer using hazard ratio (HR) and 95% confidence interval (CI) for survival outcomes. The prognostic value of high and low gene expression groups was evaluated using log-rank test.
P < 0.05 was considered indicative of statistical significance.
GeneMANIA analysis
GeneMANIA (
http://www.genemania.org) [
16] is used to generate a database of gene functions. On the basis of querying genes, GeneMANIA expands the list of genes with similar functions, indicating the relationship between genes and datasets. To draw the interactive functional association network, we used the database to construct
PSMDs gene interaction network in the aspects of physical interaction, co-expression, prediction, co-localization, and genetic interaction, and evaluated the related functions.
Function and pathway enrichment analysis
We obtained
PSMD8 co-expressed genes through the cBioPortal database. GO function and KEGG pathway enrichment analysis were applied using DAVID (
https://david.ncifcrf.gov/) [
17‐
20], which integrated biological data and analysis tools to provide a systematic synthesis Annotation information of biological functions.
P < 0.05 was set as the cut-off criterion.
Sample sources and clinical data
A total of 125 ovarian tissue paraffin specimens from inpatients who underwent surgical resection from 2008 to 2016 in the Department of Obstetrics and Gynecology, Shengjing Hospital Affiliated to China Medical University were selected. There were 80 cases of ovarian epithelial malignant tumor, 18 cases of ovarian epithelial borderline tumor, 16 cases of ovarian epithelial benign tumor, and 11 cases of normal ovarian tissue. There were no significant between-group differences with respect to age (P > 0.05). The pathological types of the malignant group included serous carcinomas (n = 58), mucinous carcinomas (n = 3), endometrioid carcinomas (n = 10), and clear cell carcinomas (n = 9). The malignant group was classified according to histology: 30 cases with high and medium differentiation and 50 cases with low differentiation. The surgical and pathological staging was performed according to the International Federation of Obstetrics and Gynecology (FIGO) criteria. In the malignant group, 66 patients underwent lymph node dissection, 41 patients had lymph node metastasis, 25 patients had no lymph node metastasis; the remaining patients did not undergo lymph node dissection. All the cases were primary, and no radiotherapy and chemotherapy were performed. Complete clinical data was available for all cases.
Immunohistochemistry
The histopathological specimens used in the experiments were fixed in 10% formalin solution, paraffin-embedded, and serially sectioned at 5 µm. The paraffin sections were deparaffinised with xylene and re-hydrated with gradient alcohol solutions, and the antigens were recovered by heating. Subsequently, H2O2, goat serum blocking solution, and anti-PSMD8 antibody (1:100, ab246883, Abcam, Cambridge, UK) was added; the solutions were left to incubate overnight at 4 °C. The following day, the slices were incubated with horseradish peroxidase labelled goat anti-rabbit/mouse secondary antibodies and stained using 3,3-diaminobenzidine (UltraSensitive™ SP Mouse/Rabbit IHC Kit, Fuzhou Maixin Biotech Co. Ltd., Fuzhou, China). Nuclei were stained blue using hematoxylin. The sections were then dehydrated, cleared by xylene, and mounted.
Scoring method: Brownish-yellow, or brown colour of cytoplasm and cell membrane was considered as positive results. Brownish-yellow, brown, light yellow and no staining were scored as 3 points, 2 points, 1 point, and 0 points, respectively according to the coloring intensity. After observing the percentage of stained area in the whole section, the percentage of positive cells, > 75%, 51–75%, 26–50%, 5–25%, and < 5% were recorded as 4 points, 3 points, 2 points, 1 point, and 0 point, respectively. The final score was obtained by multiplying the two items: 0–2 points were recorded as negative expression(-), 3–4 points were recorded as weak positive expression( +), 5–8 points as moderately positive expression (+ +), and 9–12 points as strongly positive expression (+ + +). The sections were evaluated and scored independently by two pathologists who were blinded to the patient's information. The study was approved by the Ethics Committee of Shengjing Hospital Affiliated to China Medical University (2022PS411K). All methods were carried out in accordance with relevant guidelines and regulations.
Cell culture
Ovarian cancer cell lines OVCAR3 and A2780 cell lines (Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, Shanghai, China) were cultured in cell culture medium (Biological Industries, Beit-Haemek, Israel) containing 10% fetal bovine serum (Biological Industries, Beit-Haemek, Israel). When the cells had grown to 80–90% confluence, the culture medium was discarded, cells washed with PBS, and digested with trypsin, and stopped immediately when the cells were about to no longer adhere to the wall. The cells were centrifuged and resuspended, placed into a new culture vessel, and continued to culture.
Establishment of stably overexpressing PSMD8 cell line
Viral transfection was performed in OVCAR3 cell line and A2780 cell line using lentivirus-mediated PSMD8 gene overexpression vector. First 500 µL of complete medium was added to the 24-well plate, followed by addition of PSMD8 overexpressing lentivirus, and the corresponding volume of polybrene to facilitate transfection. The PSMD8 stable high expression cell line OVCAR3-PSMD8-H and its control cell line OVCAR3-PSMD8-H-Mock, A2780-PSMD8-H and its control cell line A2780-PSMD8-H-Mock were constructed.
Establishment of transient low-expressing PSMD8 cell line
PSMD8 siRNA was transfected in OVCAR3 and A2780 cell lines. The PSMD8 siRNA working solution was prepared according to the instructions. On the day of transfection, the cells to be transfected were starved in a 6-well plate with serum-free medium. Serum-free medium, Lipo 3000, 7 µL PSMD8 siRNA-1(5’ GCAUGUACGAGCAACUCAATT 3’ and 3’ UUGAGUUGCUCGUACAUGCTT 5’) or PSMD8 siRNA-2 (5’ GACACUAUCAGGGAUGAGATT 3’ and 3’ UCUCAUCCCUGAUAGUGUCTT 5’) (GenePharma, Suzhou, China) was added to the cells to be transfected. After culturing the stained cells for 6–8 h, serum was added and the cells continued to culture for 48–72 h for cell function experiments. The PSMD8 low-expressing cell lines OVCAR3-PSMD8-L1, OVCAR3-PSMD8-L2 and their control cell lines OVCAR3-PSMD8-Mock, A2780-PSMD8-L1, A2780-PSMD8-L2 and their control cell lines A2780-PSMD8-Mock were constructed.
Scratch test
After observing the cells under the microscope to ascertain the ideal density, the pipette tip was used to make a cross scratch and pictures obtained. After culturing in serum-free medium for 24 h, the six-well plate was taken out, washed with PBS, and photographed to observe the healing of the scratches. The above experiment was repeated 3 times.
MTT experiment
Cells were seeded at a density of 2000 cells per well, and cultured at 37 °C. When the cells had adhered, the initial plate was counted as 0 h. 20 µL of sterile MTT working solution was added to each well, mixed well, and incubated at 37 °C for 4 h. Then, the medium was aspirated, 150 µL of DMSO blue-purple crystals were added, and the absorbance of each well was measured. Attention was paid to avoid light throughout the operation. The above experiment was repeated 3 times.
Invasion assay
The transwell chamber was placed in the culture plate. 1:7.5 diluted Matrigel (356,234, BD Biosciences, Franklin Lakes, NJ) and serum-free cell suspension (containing 4 × 104 cells) were added to the upper chamber of the chamber, and serum-containing medium was added to the lower chamber. After culturing for 48–72 h, the cells in the lower chamber were taken out, fixed and stained. Residual cells left on plate were observed under a microscope. The above experiment was repeated 3 times.
Statistical analysis
SPSS 22.0 (IBM Corporation, Armonk, NY, USA) software was used for data analysis. Chi-squared test and Fisher exact test were used for enumeration data, and t test was used for measurement data. Kaplan–Meier method was used to generate survival curves and between-group differences were assessed using log-rank test. Cox regression models were used to analyze the relationship between PSMD8 expression and clinicopathological parameters. P values < 0.05 were considered indicative of statistical significance.
Discussion
Ubiquitination is an important post-translational modification that controls substrate degradation and can be reversed by deubiquitinating enzymes (DUBs) [
21]. The ubiquitin–proteasome system (UPS) is responsible for the degradation of 80% of intracellular proteins in eukaryotic cells [
22]. The UPS is involved in a wide range of biological functions, such as cell growth, cell cycle progression, DNA transcription, damage, repair, and signal transduction [
23,
24]. Therefore, dysfunction of the UPS or its components can lead to severe disease [
25‐
29].
The proteasome is a multi-subunit complex consisting of a 19S regulatory granule and a 20S core granule, which mainly functions to degrade ubiquitin-tagged proteins [
30]. Among them, the 19S regulatory particles are divided into two parts: the base and the lid, which are connected to the outer surface of the 20S core particles and play the role of recognizing ubiquitinated protein substrates, removing ubiquitin linkages, unfolding proteins, and transporting proteins into the 20S core particles. The 20S core particle is responsible for protein degradation [
31].
PSMDs encodes a family of subunits of the 26S proteasome, which is a non-ATPase subunit in the proteasome structure. It has 14 members in total that participate in the formation of the 19S regulatory complex and perform the functions of catalyzing the unfolding and transport of substrate proteins.
PSMDs play an important role in a variety of cancers, and abnormal gene expression is often associated with tumor-regulating oncogenes and tumor suppressor genes [
9‐
11,
32‐
34]. Thus, these are potential diagnostic and prognostic biomarkers as well as therapeutic targets.
In a study, loss of PSMD1 was found to inhibit the proliferation of breast cancer cells and induce cell cycle arrest by inhibiting the degradation of p53. The upregulation of PSMD1 gene was mainly accompanied by increase of tamoxifen resistance in BRCA cells [
35]. High PSMD1 expression was shown to be significantly associated with disease-free survival (DFS) and overall survival (OS) of gastric cancer patients [
36]. PSMD2 and PSMD7 were shown to regulate breast cancer cell proliferation and cell cycle progression by regulating the proteasomal degradation of p21 and p27 [
37]. PSMD3 regulates breast cancer by stabilizing HER2 degradation [
38]. PSMD4 was shown to affect esophageal cancer by inhibiting endoplasmic reticulum stress and degree of cellular malignancy [
39]. Antioxidant response element-bound nuclear Nrf2 (nNrf2) promotes chemoresistance in colorectal cancer through the EMT pathway via the NF-κB/AKT/β-catenin/ZEB1 cascade by inducing PSMD4 expression [
40]. PSMD4 copy number amplification was associated with sensitivity to Poly (ADP-ribose) polymerase (PARP) inhibitors (PARPi), and it might be a better predictor of PARPi sensitivity than BRCA1/2 mutations [
41]. Inactivation of PSMD5 was shown to promote colorectal tumor progression [
42], TNF-α increases PSMD5 expression through NFκB. Excess PSMD5 directly inhibited the assembly and activity of the 26S proteasome, and TNF-α enhanced the interaction of PSMD5 with PSMC2. In another study, the expression of PSMD6, PSMD9, PSMD11, and PSMD14 was significantly associated with a decreased chance of survival in patients with pancreatic ductal adenocarcinoma [
43]. PSMD7 was considered an oncogene in prostate cancer, esophageal squamous cell carcinoma (ESCC), and breast cancer [
44,
45]. PSMD7 knockout was shown to induce cell cycle arrest in G0/G1 phase, leading to cell senescence and apoptosis or inhibit lung cancer progression by modulating the p53 pathway [
45].
PSMD8, a deubiquitinating enzyme, is a member of the JAMM (JAB1/MPN/Mov34) domain family, and studies have shown that PSMD8 interacts with the sperm adhesin AQN1 to limit polyfertilization [
46]. It was highly expressed in invasive bladder cancer and breast cancer [
47,
48]. High expression of PSMD9 was associated with post-radiotherapy recurrence in cervical and breast cancer [
49], and endogenous PSMD10 interacted with GRP78 to regulate endoplasmic reticulum stress, which might provide a therapeutic target for homocysteine-induced liver injury [
50]. miR-3619-5p inhibited tumor growth in vivo by inducing the phosphorylation of activator of transcription 3 (STAT3) and retinoblastoma protein (Rb1), thereby targeting PSMD10 to inhibit cell proliferation and induce G1 arrest [
51]. PSMD11 and PSMD12 have been extensively studied in the nervous system. The expression of PSMD11 was down-regulated in the hippocampus of epileptic mice, and the lncRNA Peg13 was shown to up-regulate PSMD11 in a miR-490-3p-dependent manner, thereby inactivating the Wnt/β-catenin pathway and relieving epilepsy progression in mice [
52]. Bi et al. [
53] reported that silencing the
PSMD13 gene has the potential to treat neuroinflammatory diseases by regulating the activation of microglia and the production of inflammatory mediators. PSMD12 enhanced the proliferation and invasion of glioma cells through Akt signaling-mediated Nrf2 expression [
54], and PSMD12 was considered to be a key regulator of glioma development and progression. PSMD14 had been shown to play an oncogenic role in the context of ovarian, prostate, hepatocellular, lung adenocarcinoma, and colorectal cancers [
55‐
59]. PSMD14 overcame drug resistance in head and neck squamous cell carcinoma by inhibiting E2F1 ubiquitination and degradation, improving Akt pathway activation and SOX2 transcription [
60].
In this study, analysis of mRNA expression data of PSMD8 in various normal tissues and various cancer tissues showed that PSMD8 was more expressed in normal skeletal muscle, cardiac muscle, and tongue muscle; among malignant tumors, PSMD8 was expressed in ovarian cancer, testicular cancer, glioma, and melanoma. On analyzing the relationship of PSMD family and ovarian cancer, the expression levels of PSMD8 and PSMD14 mRNA in ovarian cancer were found to be significantly higher than those in normal ovarian tissue. Prognostic analysis found that patients with high mRNA expression of PSMD2, PSMD3, PSMD4, PSMD5, PSMD8, PSMD11, PSMD12, and PSMD14 have poor prognosis; among these, PSMD8 showed the best prognostic value in patients with serous ovarian cancer. On analyzing the relationship of PSMD8 expression with different stages, differentiation, and TP53 mutation status, PSMD8 mRNA expression was found to be up-regulated in FIGO stage III-IV and TP53 mutant patients, and the PFS was worse. Furthermore, immunohistochemical experiments demonstrated that PSMD8 was mainly expressed in the cytoplasm, and was highly expressed in ovarian epithelial malignant tumor tissues, and the expression level showed a correlation with FIGO stage. Patients with high PSMD8 expression and advanced FIGO stage showed a poor prognosis. Therefore, PSMD8 had the strongest correlation with the prognosis of ovarian cancer patients, and was closely related to the occurrence and development of ovarian cancer. Through the analysis of spearman correlation of PSMD8 with lymphocytes and immunomodulators of TISIDB database, the patients with CD56dim, Act_CD8, Act_DC and CD56bright showed high expression of PSMD8, indicating that patients with high expression of PSMD8 may have poor effect on immunotherapy. Due to the heterogeneity of tumors, studies have shown that TILs therapy had a good therapeutic effect, while not all tumors responded well which need to be validated in clinical trials. Furthermore, the immunoinhibitors including PVDL2, IDO1, IL10RB and VTCN1, immunostimulators including PVR, TNFRSF4, MICB and CD48; MHC molecules including HLA-A, HLA-C, B2M, and TAP1 indicated the relationship between immune-related molecules and PSMD8. MTT, cell scratch assay, and transwell assay confirmed that PSMD8 overexpression could enhance malignant biological behaviors such as proliferation, migration, and invasion of ovarian cancer cells, indicating that PSMD8 could be used as a potential marker for early diagnosis, disease progression, and prognostic assessment in patients with ovarian cancer. Moreover, it was also a potential therapeutic target. The function and enrichment analysis of PSMD8 genes through the database showed that these genes are mainly involved in energy metabolism, such as glucose metabolism, mitochondrial energy metabolism, DNA replication, protein synthesis and other biological processes. Cancer-related signaling pathways affected the occurrence and development of ovarian cancer.
Based on the above studies, we identified the important role of the PSMD family, especially PSMD8, in the occurrence and development of ovarian cancer. In vitro experiments as well as analysis of the relationship of PSMD8 expression in ovarian cancer with the clinicopathological parameters and survival outcomes showed that PSMD8 overexpression can enhance malignant biological behavior of ovarian cancer. Our findings suggest that PSMD8 is a potential biomarker for early diagnosis, disease progression, and prognostic assessment of patients with ovarian cancer patients. In addition, it is also a potential therapeutic target. However, due to differences in database backgrounds, limited sample size, and lack of relevant experimental foundations, further experiments are required for more in-depth characterization of the specific roles and related mechanisms of PSMDs family and PSMD8 in ovarian cancer.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.