Abstract
Background/Aim: PDIA6 is a disulphide isomerase of the PDI family, known to mediate disulphide bond formation in the endoplasmic reticulum. However, PDI-related proteins also function in other parts of the cell and PDIA6 has been shown to be involved in many types of cancers. We previously identified PDIA6 as a putative Maspin interactor. Maspin has itself been implicated in prostate cancer progression. Our aim was to further explore the roles of Maspin in prostate cancer and establish whether PDIA6 is also involved in prostate cancer. Materials and Methods: RNA levels of PDIA6 and Maspin in prostate cell lines were measured using RT-PCR. Bioinformatics analysis of the TCGA database was used to find RNA levels of PDIA6 and Maspin in prostate cancer. siRNAs were used to knock-down PDIA6, and proliferation and migration assays were conducted on those cells. Results: PDIA6 and Maspin RNA were shown to be expressed at varying levels in prostate cell lines. RNAseq data showed that PDIA6 expression was significantly increased in prostate adenocarcinoma samples, while Maspin RNA expression was decreased. When PDIA6 expression was knocked-down using siRNA in prostate cell lines, proliferation was decreased substantially in the two prostate cancer cell lines (DU145 and PC3) and also decreased in the normal prostate cell line (PNT1a), though less strongly. Conclusion: PDIA6 expression is higher in prostate cancer cells compared to normal prostate cells. Decreasing PDIA6 expression decreases proliferation. Thus, PDIA6 is a promising target for prostate cancer therapeutics.
Disulphide isomerases of the protein disulphide isomerase (PDI) family have been widely studied as molecular chaperones, mediating correct protein folding and disulphide bond formation in the endoplasmic reticulum (ER) (1-4). However, PDI-related proteins also function outside the ER. Recently, PDIA6 (also known as P5, PDI-P5, ERP5, TXNDC7) has been identified as an unconventional RNA-binding protein, with its tumorigenic properties in cutaneous melanoma requiring RNA binding (5). PDIA6 has been shown to be highly expressed in many types of cancer including non-small lung cancer (NSLC or NSCLC; 6, 7), pulmonary squamous cell carcinoma (SCC; 8), nonfibrotic hepatocellular carcinoma (nfHCC; 9), primary ductal breast cancer (10), bladder cancer (11), ovarian cancer (12), leukemia, glioblastoma, breast, colon, ovarian, and uterine cancer cells (13), gastric cancer (14), oral squamous cell carcinoma (OSCC; 15), pancreatic cancer (16), renal cell carcinoma (17), and multiple myeloma (18), but has not been previously identified in prostate cancer.
PDIA6 has been associated with gradually increasing levels during cancer progression. In breast cancer, lymph node metastasis showed higher levels of PDIA6 expression (10). Expression of PDIA6 (and all PDI family members tested) increased during ovarian cancer progression (12). In multiple myeloma, PDIA6 progressively increased in expression at both RNA and protein levels with more severity (18). In NSCLC, PDIA6 expression showed a gradual increase as tumor stage progressed (7). Malignant clinicopathologic features in OSCC were found to be associated with higher PDIA6 expression (15). In pancreatic cancer, higher levels of PDIA6 were associated with poorer patient prognosis, when compared with lower levels of PDIA6 (16).
In addition, PDIA6 expression can be used for predicting cancer prognosis and survival (7, 15, 16). Knocking down PDIA6 expression enhanced cell death and chemosensitivity in cisplatin-resistant cancer cells in NSCLC (6, 7) and gastric cancer (14) as well as chemosensitivity to DDP in pancreatic cancer (16). Knockdown of PDIA6 expression also caused a decrease in cancer cell proliferation (7, 11, 13-17), migration (13, 15, 16), and invasion (11, 15, 16). Knocking down PDIA6 expression reduced clonogenicity, anoikis resistance, and 3D-growth in a number of melanoma cell lines and substantially decreased the ability of melanoma cells to metastasize to the lung (5). Accordingly, when PDIA6 was overexpressed, there was an increase in cancer cell proliferation (7, 14-16, 19), migration (15, 16), and invasion (15, 16). PDIA6 was expressed more highly in a subset of hepatocellular carcinoma that includes portal vein tumor thrombosis (20). PDIA6 appears to be signaling through the Wnt/Beta-catenin pathway (11, 14, 16, 19), as well as the JNK/c-Jun pathway through MAP4K1 (7). Knockdown of PDIA6 expression inhibited tumorigenesis (7, 11, 16) while overexpression of PDIA6 promoted tumorigenesis (15, 16). Thus, inhibition of PDIA6 is a strong candidate for cancer therapy as PDIs appear to be druggable targets (21). Indeed, PDIA6 is one of the proteins inhibited by zafirlukast, a pan-thiol isomerase inhibitor (22). Zafirlukast has been used in a pilot clinical trial for women with relapsed ovarian cancer (23).
We previously identified a putative interaction between PDIA6 and Maspin (24). Maspin is a non-inhibitory member of the serpin family that has been shown to have effects on a range of cell types consistent with anti-metastatic activity (25). It was originally identified as a tumor metastasis suppressor in breast cancer (26), but Maspin has since been implicated in the progression of cancers from other origins, including those of the prostate (27-29). Maspin has been shown to influence a range of cell functions that impact on cell migration and survival; it decreased cell migration, invasion, proliferation, and angiogenesis while it increased cell adhesion and sensitivity to apoptotic stimuli (29). In this study, we have aimed to further explore the roles of Maspin in prostate cancer.
Though PDIA6 has not previously been identified in prostate cancer, other PDI family members have been, including PDIA2, PDIA4, and PDIA3. PDIA2 levels were shown to be a good prognostic marker: higher levels were correlated with poorer prostate cancer survival (30). PDIA4 was overexpressed in prostate cancer cells that were docetaxel resistant (PC-3 and C4-2B), while upregulation of PDIA4 increased the docetaxel resistance in those same cell lines (31). PDIA3 was more highly expressed in more severe prostate cancers (32). PDIA3 was also identified as a potential membrane receptor for Vitamin D or metabolites – Vitamin D3 being a potential treatment for prostate cancer (33, 34). PDIA3 RNA was found in LNCaP, PC3, and PNT-2 prostate cancer cell lines and PDIA3 protein in LNCaP cells (33). In prostate tumor tissues, PDIA3 and LEDGF/p75 proteins were found to be elevated and LEDGF/p75 bound to and transactivated the PDIA3 promoter, regulating PDIA3 expression (35). A new transcript isoform of PDIA3 called PDIA3N was detected in five prostate cancer cell lines and the higher levels of PDIA3N compared to PDIA3 corresponded to more severe cancer (36). Zafirlukast, the pan-thiol isomerase inhibitor mentioned above, inhibits PDIA1, PDIA2, PDIA3, PDIA4, and PDIA6 (22), which might make it very effective for prostate cancer treatment. Indeed, zafirlukast was found to inhibit growth of multiple cancer cell lines, including the PC3 prostate cancer cell line (23). Though PDIA2, PDIA3, and PDIA4 have been identified in the context of prostate cancer, PDIA6 has not. Thus, in this study, we have aimed to find out whether PDIA6 is also involved in prostate cancer, as it has been previously found in other cancers.
Materials and Methods
Cell lines. DU145, PC3 and LNCaP cell lines were obtained from ATCC and cultured in RPMI, supplemented with 10% (v/v) fetal calf serum (FCS) and penicillin/streptomycin. PNT1a were a kind gift from Prof. Edwards (University of East Anglia, UK); they were cultured in the same way as the other prostate derived lines, except that the medium was also supplemented with 10mM HEPES. Cell lines were authenticated by short tandem repeat profiling (DDC, DNA Diagnostics Centre, London, UK). All cell culture reagents were from ThermoFisher Scientific (Paisley, UK).
Reverse transcription (RT)-PCR. Specific oligonucleotide primers and fluorogenic probes (Table I) were designed for Maspin, PDIA6, and 18S by Primer Design Ltd, (Southampton, UK) using probes labeled with a FAM fluorophore reporter (emission 520 nm) at the 5′ end and a TAMRA (quencher emission 583 nm) at the 3′ end to allow measurement by quantitative real-time RT-PCR using a BioRad CFX96.
Primers and probes for RT-PCR.
Bioinformatics analyses. Transcriptomics data was obtained by The Cancer Genome Atlas (TCGA) (37), the largest data pool of cancer genetic data that is publicly available, and it was accessed via the Xena UCSC portal (38). RNAseq analysis was performed on 595 prostate tissues comprising 100 normal and 495 prostate adenocarcinoma (primary tumor) samples. The sequencing reads were obtained by the Cancer Genomics Project (37) using the Illumina platform and normalized by RSEM (RNA-Seq by Expectation Maximization) (39) and log transformed (log2). A Welch Two Sample t-test was applied on the two groups (Normal tissue/Primary Tumor) by using R version 3.5.1. The results were presented as box and whiskers plots by using ggplot2 version 3.3.5.
siRNA knock-down. Predesigned siRNAs directed against human PDIA6 and AllStars Negative Control siRNA (referred to as control or ctrl) were obtained from Qiagen (Manchester, UK). siRNAs for PDIA6 NM_005742 were referred to as #1 (Hs_TXNDC7_1), #2 (Hs_TXNDC7_2), #3 (Hs_PDIA6_1) and #4 (Hs_PDIA6_2), with the full sequences of the siRNAs here: #1- Hs_TXNDC7_1 SI00753683 223133 AAGATGAAATTTGCTCTGCTA; 2-Hs_TXNDC7_2 SI00753690 223135 ACGGGATTAGAGGATTTC CTA; 3- Hs_PDIA6_1 SI03097871 223138 CTGGCAGTGA ATGGTCTGTAT; 4- Hs_PDIA6_2 SI03101812 223139 GACGA CAGCTTTGATAAGAAT.
Cells were transfected with 100 nM +100 nM siRNA (of 2 different siRNAs or 200 nM control) using oligofectamine, according to manufacturer’s instructions (Invitrogen). After 4 h, complete medium was added. 72 h after transfection with siRNA, cells were used. Some cells were lysed to perform Western blots. The Western blots were carried out as reported previously (40). Commercially available mouse monoclonal antibodies were used to detect PDIA6 (PDI-P5; PA3-008, Cambridge Bioscience, Cambridge, UK). Secondary antibodies were from Dako (Ely, UK). For proliferation assays, cells were then seeded as below and analyzed on the Xcelligence immediately. For migration assays, cells were then seeded as below and analyzed the next day.
Proliferation assay. The proliferation assay was performed as described previously (41), except that cells were seeded at 10,000 cells/well 3 days after siRNA treatment. To account for differences in batches, control cell index values for the three control triplicates per experiment were averaged and that mean was subtracted from the controls and the experimental wells to present in the graph and for statistical analyses.
Migration assay. The migration assay was performed as described previously (41), except that cells were seeded at 7,500 cells/well (in a 24-well-plate) 3 days after siRNA treatment, and then imaged starting 24 h later. Ten cells were counted per well in 3 wells per condition. To account for differences in batches, control migration values in microns/min for the three control triplicates per experiment were averaged and that mean was subtracted from the control and the experimental wells to present in the graph and for statistical analyses.
Statistical analyses. Data are presented as mean±standard error of the mean (SEM). Significance was judged using Generalized Linear Models in SPSS, unless otherwise stated. A statistically significant difference was defined as p<0.05.
Results
PDIA6 and Maspin were expressed in normal and cancerous prostate cell lines. To begin investigating the possible roles of Maspin and PDIA6 in prostate cancer, their expression was examined in prostate cell lines (Figure 1). LNCaP, PC3, and DU145 were prostate cancer cell lines (42-44), while PNT1a cells were immortalized prostate luminal epithelial cells (45).
Study design. (a) Experimental design for studies involving prostate cell lines. (b) Experimental design for bioinformatics analysis of prostate RNAseq data.
PDIA6 RNA levels were decreased in LNCaP cells compared to the other cell lines (Figure 2a). Maspin, on the other hand, did not appear to be expressed at all in DU145 cells (Figure 2b), consistent with previous research (46). PNT1a prostate cells had much lower Maspin RNA expression compared to LNCaP and PC3 cells. Thus, both PDIA6 and Maspin were found to be expressed at varying RNA levels in normal and cancerous prostate cell lines (Figure 2a-b).
PDIA6 and Maspin are expressed in normal and cancerous prostate cells. (a) RNA expression of PDIA6 in prostate cancer cell lines DU145, PC3, and LNCaP, as well as normal prostate cell line PNT1a as analyzed by RT-PCR. (b) RNA expression of Maspin in prostate cancer cell lines DU145, LNCaP, and PC3, as well as normal prostate cell line PNT1a as analyzed by RT-PCR. (c) RNAseq analysis of PDIA6 RNA levels in normal prostate and primary prostate tumors. n=595 prostate tissues comprising 100 normal and 495 prostate adenocarcinoma (primary tumor) samples. ***p<2.2e-16. (d) RNAseq analysis of Maspin RNA levels in normal prostate and primary prostate tumors. n=595 prostate tissues comprising 100 normal and 495 prostate adenocarcinoma (primary tumor) samples. ***p<1.818e-05.
To further explore the roles of PDIA6 and Maspin in prostate cancer, we analyzed the expression of PDIA6 and Maspin in 100 normal and 495 prostate adenocarcinoma (primary tumor) samples using publicly available transcriptomics data from The Cancer Genome Atlas (TCGA) (37) (Figure 1). PDIA6 expression was significantly and substantially increased in the tumor samples [p<2.2e-16; t(593)=−30.316; 95% CI=−1.675 to −1.470; Figure 2c]. In contrast, the expression of Maspin was slightly, but significantly, reduced in the tumor samples [p<1.818e-05; t(593)=4.436; 95% CI=0.516 to 1.346; Figure 2d]. Thus, Maspin decreased, consistent with its previously identified role as a tumor suppressor in prostate cancer (Figure 2d), in contrast, PDIA6 increased (Figure 2c).
Knock-down of PDIA6 RNA levels decreased prostate cell proliferation. Cancer cells show higher levels of proliferation and PDIA6 showed increased expression in prostate cancer (Figure 2c). Thus, PDIA6 could be involved in causing increased levels of proliferation in cancer. To investigate the role of PDIA6 in proliferation, siRNA treatment was used. Combinations of two siRNAs targeted to the PDIA6 sequence (1+3 or 2+4, see Materials and Methods section; Figure 1) were able to largely decrease protein expression of PDIA6 in DU145, PC3, and PNT1a cells compared to the scrambled siRNA control, though none of the siRNA treatments showed a complete loss of PDIA6 protein (Figure 3a). Proliferation of these PDIA6 knock-down cells was then examined and significant differences were seen based on cell type and specific siRNA treatment [cell type overall effect: χ2(2)=26.97, p<0.001, treatment overall effect: χ2(2)=80.9, p<0.001, Figure 3b]. The different siRNA treatments had different effects on the different cell types [interaction effect: χ2(4)=18.1, p=0.001, Figure 3b]. The strongest decreases in proliferation were seen in DU145 cells, where both sets of siRNA combinations caused substantially and significantly decreased proliferation when compared to the control siRNA-treated cells (control vs. both siRNA treatments: p<0.001; Figure 3b). The other cancerous cell line, PC3, also showed significant decreases in proliferation, though the 2+4 showed a smaller, though still significant, decrease when compared to controls (control vs. 1+3: p<0.001; control vs. 2+4: p=0.044; Figure 3b). Interestingly the PNT1a cells that are derived from normal prostate tissue showed the least decrease in proliferation upon PDIA6 knockdown (Figure 3b). The 2+4 siRNA treatment actually showed no significant decrease in proliferation at all, while the 1+3 siRNA treatment only showed a modest, though significant decrease in proliferation (p=0.001; Figure 3b). Thus, PDIA6 knockdown in prostate cell lines resulted in significant decreases in proliferation, with more substantial decreases in the prostate cancer cell lines than the noncancerous PNT1a prostate cells (Figure 3).
Effect of PDIA6 knockdown on the proliferation of prostate cells. (a) Western blots with anti-PDIA6 showing the extent of the siRNA knockdown of PDIA6 protein in the two cancerous prostate cell lines DU145 and PC3 as well as the ‘normal’ prostate cell line PNT1a. Individual siRNAs were not effective, so combinations of 1+3 and 2+4 were used. (b) Proliferation of cells normalized to control siRNA-treated cells for DU145, PC3, and PNT1a prostate cells. n=9 wells for each cell line and siRNA treatment. *p<0.05, **p<0.01, ***p<0.001.
Knock-down of PDIA6 RNA levels had no consistent effect on prostate cell migration. As cell migration is an important characteristic of metastatic cancer cells and Maspin is thought to be involved in preventing cancer metastasis, particularly affecting epithelial-to-mesenchymal transition (47), the effects of PDIA6 knock-down were also assessed on prostate cell migration (Figure 1). The 1+3 and 2+4 siRNA combinations were used again to knock-down PDIA6 RNA leading to decreased PDIA6 protein levels (examples in Figure 3a). In these experiments, prostate cell migration over time was analyzed. Significant differences were seen between treatments and their effects on the different cell types, but no significant differences were seen based on cell type [cell type overall effect: χ2(2)=2.7, n.s., p=0.255, treatment overall effect: χ2(2)=12.2, p=0.002, interaction effect: χ2(4)=17.98, p=0.001; Figure 4]. For DU145 and PNT1a cells, neither PDIA6-targeted siRNA combination had any significant effect on cell migration when compared with controls (Figure 4). For PC3 cells, the knock-down in PDIA6 protein showed significant, but inconsistent effects on PC3 cell migration (Figure 4). Treatment with the PDIA6-targeted 1+3 siRNA showed a small but significant decrease in PC3 cell migration (p=0.004), while treatment with the PDIA6-targeted 2+4 siRNA showed a small but significant increase in PC3 cell migration (p=0.018; Figure 4). Thus, knock-down of PDIA6 protein did not consistently affect the migration capacity of either cancerous or normal prostate cell lines (Figure 4).
Effect of PDIA6 knockdown on the migration of prostate cells. Migration of 10 cells per well (microns/min) was averaged and normalized to the control siRNA treated cells for each cell line. n=90 cells for DU145 and PC3 and n=60 cells for PNT1a (for each siRNA treatment for each cell line), *p<0.05, **p<0.01.
Discussion
We previously reported an in vitro interaction between Maspin and PDIA6 (24). In the current work, we showed that PDIA6 and Maspin are both expressed at the RNA level in normal prostate and prostate cancer cell lines and tumors (Figure 2). Consistent with previous studies investigating prostate cancer (48), Maspin in our RNAseq analysis showed decreased levels in prostate adenocarcinoma primary tumors (Figure 2d). However, the first-line treatment for prostate cancer is androgen ablation, which is known to increase maspin expression (49), thus patient cancers will likely have an increased level of Maspin. After treatment, in hormone-refractory prostate cancer, higher levels of Maspin allowed the prostate cancer cells PC3 to respond better to the anti-tumor abilities of curcumin (50).
Though PDIA6 has been extensively studied in a number of different cancer types (21, 51), it has never previously been analyzed in prostate cancer. In this current RNAseq analysis, we have found that PDIA6 RNA expression is significantly increased in prostate adenocarcinoma primary tumors (Figure 2c). This is consistent with the PDIA6 expression increase in many different cancer types (6-15). In addition, PDIA2, PDIA3, and PDIA4 all are increased in prostate cancer (30, 31, 33, 35, 36).
We also showed in this current study that siRNA knock-down of PDIA6 causes a decrease in prostate cancer cell proliferation (DU145 and PC3; Figure 3). The same knock-down of PDIA6 in the normal prostate cell line, PNT1a, decreases proliferation, but much less (Figure 3). This is consistent with the literature showing decreases in proliferation when PDIA6 is knocked down in other cancer types (7, 11, 13-17). In NSCLC (7), OSCC (15), renal (17), and pancreatic cancer (16), PDIA6 is shown to increase proliferation by inhibiting apoptosis and in NSCLC, there appears to be an interaction with MAP4K1 causing a decrease in JNK/c-Jun signaling (7). In bladder (11), gastric (14), renal (17), and pancreatic (16) cancers, PDIA6 appears to increase proliferation by promoting the Wnt/β-catenin pathway. In our current study, we did not find any consistent effect of PDIA6 knockdown on migration in prostate cancer or normal prostate cell lines (Figure 4). In other cancer types or cell lines, PDIA6 knockdown has shown a decrease in cell migration (13, 15, 16). Interestingly, when a human malignant glioma cell line (U87MG) was treated with siRNA to PDIA6, migration and invasion actually increased (52). It may be that PDIA6 knockdown has different effects on migration depending on the cellular and cancer context.
Intriguingly, as mentioned above, zafirlukast, an FDA-approved medication for asthma, is a pan-thiol isomerase inhibitor, shown to inhibit PDIA1, 2, 3, 4, and 6 (22). As thiol isomerases are known to be involved in cancer signaling, progression, and metastasis, zafirlukast has been used in a pilot clinical trial for women with relapsed ovarian cancer (23). The outcome marker was only the tumor marker CA-125 and none of the four women showed a decrease in CA-125, but there was a decreased rate of rise of CA-125 and there were few side effects (23). In addition, zafirlukast inhibited the growth of the PC3 prostate cancer cell line (23). Seeing that zafirlukast is a promising treatment for relapsed ovarian cancer, a specific inhibitor of PDIA6 or a more general thiol isomerase inhibitor like zafirlukast are likely good candidates for prostate cancer treatment as well–particularly since we have shown knocking down PDIA6 expression decreases prostate cancer cell line proliferation.
Conclusion
In summary, our findings revealed PDIA6 could be involved in prostate cancer and that Maspin expression was decreased in prostate adenocarcinoma, consistent with its role as a tumor suppressor. We showed that PDIA6 expression was highly increased in prostate adenocarcinoma and drove proliferation in prostate cancer cell lines. PDIA6 has not previously been shown to be involved in prostate cancer, so our research has shown a novel role for PDIA6 in prostate cancer. Our research identifies Maspin and particularly PDIA6 as targets for further investigation in prostate cancer. PDIA6 is more highly expressed in many cancer types and lowering PDIA6 expression decreases cancer proliferation. Since it is a druggable target, PDIA6 is a strong candidate for prostate cancer therapeutics.
Footnotes
↵* Current address: School of Life Sciences, University of Warwick, Coventry, U.K.
Authors’ Contributions
Tora K. Smulders-Srinivasan: Conceptualization, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization. Sarah E. Jenkinson: Investigation, Writing - Review & Editing. Louise J. Brown: Investigation, Writing - Review & Editing. Vasileios P. Lenis: Conceptualization, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization. Rosemary Bass: Conceptualization, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision, Project administration, Funding acquisition.
Conflicts of Interest
There are no conflicts of interest for the Authors of this research article.
Funding
This work was supported by The JGW Patterson Foundation. The funders had no roles in in study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the article for publication.
- Received September 8, 2023.
- Revision received October 11, 2023.
- Accepted October 30, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
