Abstract
Background/Aim: Squamous cell carcinoma is one of the major subtypes of esophageal carcinoma, and the 5-year overall survival rate of esophageal squamous cell carcinoma patients who underwent curative treatment remains below 40%. We aimed to detect and validate the prognosticators of esophageal squamous cell carcinoma in patients who underwent radical esophagectomy. Materials and Methods: Comprehensive analysis of transcriptome and clinical data from The Cancer Genome Atlas revealed OPLAH as one of the differentially expressed genes between esophageal squamous cell carcinoma tissues and normal esophageal mucosa. OPLAH expression changes were significantly associated with a patient prognosis. OPLAH protein levels were further evaluated by immunohisto-chemistry in esophageal squamous cell carcinoma tissues (n=177) as well as in serum samples (n=54) using ELISA. Results: OPLAH mRNA was significantly overexpressed in esophageal squamous cell carcinoma tissues compared to normal esophageal mucosa, and patients with high OPLAH mRNA expression have a significantly poorer prognosis, according to The Cancer Genome Atlas data. The high staining intensity of OPLAH protein in esophageal squamous cell carcinoma tissue clearly stratified patient prognosis. According to multivariable analysis, high OPLAH protein expression was an independent prognostic factor for survival after surgery. Pre-neoadjuvant chemotherapy serum OPLAH protein concentrations were significantly associated with clinical tumor depth and node positivity and, consequently, with advanced clinical stage. The serum OPLAH protein concentration was significantly decreased by neoadjuvant chemotherapy. Conclusion: OPLAH protein expression in cancerous tissue and serum may have clinical utility towards stratifying prognosis of patients with esophageal squamous cell carcinoma.
Squamous cell carcinoma (SCC) is one of the major subtypes of esophageal carcinoma, and accounts for up to 90% of esophageal carcinoma cases in eastern Asia and eastern and southern Africa and 50% in Western countries (1, 2). In recent decades, the prognosis of patients with esophageal squamous cell carcinoma (ESCC) has improved due to development of multimodal therapeutic strategies, perioperative adjuvant therapies, and novel antitumor agents (3-7). However, the 5-year overall survival rate of ESCC patients who undergo curative treatment remains below 40% (8). Advanced ESCC has a high recurrence rate even after curative treatment due to the difficulty of treatment for anatomical reasons and the high potential for metastasis. If recurrence occurs, the subsequent prognosis of ESCC patients is dismal.
The elucidation of molecular biomarkers could be of clinical importance both as prognostic factors besides the established clinical staging and as potential targets for treatment (9-11). Malignant phenotype of tumors depends on characteristic metabolic activities, such as the Warburg effect. In recent years, research related to metabolic pathways and metabolites that analyze proteome and metabolome has been reported due to the progress of various mass spectrometers (12, 13). Finding biomarkers in molecules associated with metabolic pathways could be an important first step toward the subsequent development of therapeutic agents. Through comprehensive analysis of transcriptome and clinical data from The Cancer Genome Atlas in this study, we identified 5-oxoprolinase, ATP-hydrolysin (OPLAH), which is involved in glutathione metabolism, as a putative prognosticator for ESCC patients. Glutathione metabolism is essential for cell viability due to its antioxidant and detoxifying effects, and these effects favor cell survival (14). The activity of glutathione metabolism is reportedly associated with malignant phenotypes and treatment resistance in some malignant tumors (15-18). However, the prognostic significance of OPLAH expression in malignant tumors, including ESCC, has not been uncovered.
In the current study, we evaluated the association between OPLAH expression and clinicopathological factors in ESCC patients and examined its utility as a prognostic biomarker.
Materials and Methods
Public ESCC datasets. The RNA sequencing data of 185 esophageal carcinoma (EC) tissues and 11 normal esophageal mucosa tissues and the clinical data of these EC patients were obtained from The Cancer Genome Atlas (TCGA) via Broad Institute’s Firehose (URL: http://gdac.broadinstitute.org). EC patient data were divided according to the histological subtype, SCC or adenocarcinoma, and the ESCC patient data were used for analysis. To validate the predictive ability of candidate genes, Kaplan–Meier Plotter (URL: http://kmplot.com/analysis/) was used.
Identification of candidates as prognosticators of ESCC. Using the mRNA sequencing data and clinical data of ESCC patients from TCGA, we extracted candidates for prognosticator that satisfied the following three criteria: 1) the mRNA expression change between ESCC tissues and normal tissues was significant, and the fold change (FC) was not less than 2 or not more than 0.5; 2) the mRNA expression change in ESCC tissue was significantly associated with poor prognosis; and 3) protein coding gene.
Clinical sample collection. For immunohistochemistry (IHC), 177 primary ESCC tissues were collected from patients who underwent radical esophagectomy at Akita University Hospital between February 2000 and July 2011. None of these patients had received neoadjuvant treatment before surgery. The tissue specimens were fixed in paraformaldehyde and embedded in paraffin, and a tissue microarray (TMA) was constructed at the Pathology Institute (Toyama, Japan) (19). Each sample was punched and arrayed on TMA in triplicate.
For enzyme-linked immunosorbent assay (ELISA), 54 serial serum samples were collected before and after neoadjuvant chemotherapy (NAC) from 27 patients who underwent NAC followed by radical esophagectomy at Nagoya University Hospital between January 2018 and August 2020. Eight serum samples were collected from 8 healthy volunteers for comparison. Serum samples were separated by centrifugation, frozen immediately in liquid nitrogen, and stored at −80°C.
Immunohistochemical staining (IHC). TMA samples were incubated overnight at 4°C with a mouse monoclonal antibody raised against OPLAH protein (sc-393570, Santa Cruz, Dallas, TX, USA) diluted at 1:50. Antigen-antibody complexes were visualized using liquid 3,3′-diaminobenzidine (Nichirei, Tokyo, Japan). The staining intensity of ESCC cells was analyzed by two independent observers who were not informed of the clinical data and were categorized into three groups: high, intermediate, and low.
Enzyme-linked immunosorbent assay (ELISA). Serum OPLAH protein concentration was measured using an ELISA Kit (MBS9332625, MyBioSource, San Diego, CA, USA). The procedures were performed according to the manufacturer’s protocol, and the OPLAH concentration was reported in ng/ml.
Statistical analysis. For continuous variables, statistical analyses were performed using the Student’s t-test. Categorical variables were compared using χ2 tests or Fisher’s exact tests. Survival time was evaluated using the Kaplan–Meier method, and survival curves were compared using log-rank tests. Multivariable analysis for survival time was estimated by the Cox proportional hazard model. All statistical analyses were performed using R version 3.4.1 (Vienna, Austria, URL: http://www.R-project.org/). p<0.05 was considered to indicate a significant difference.
Ethics statement. This study was approved by the Institutional Review Board of Nagoya University (Approval Number 2014-0044). All patients provided written informed consent, and experiments with clinical samples were performed under the approved guidelines.
Results
Identification of OPLAH as a putative prognosticator of ESCC. First, the expression of 20,500 genes was compared between ESCC tissues and normal tissues using TCGA RNA sequencing data. A total of 1,612 genes were significantly overexpressed in ESCC tissue with FC not less than 2, and 515 genes were significantly suppressed in ESCC tissue with FC not more than 0.5 compared to normal tissues (Figure 1A). Next, genes whose expression change in ESCC tissue was significantly associated with poor prognosis were extracted by the log-rank test. The cutoff value of each gene expression level for dividing patients into two groups was the median. A total of 408 genes whose overexpression was significantly associated with poor prognosis and 515 genes whose suppression was significantly associated with poor prognosis were sorted. Then, the intersection of the differentially overexpressed genes and the genes whose high expression change was associated with prognosis identified 20 genes (Figure 1B). The intersection of the differentially suppressed genes and the genes whose low expression change was associated with prognosis identified 7 genes (Figure 1B). Among them, 24 protein-coding genes were identified as putative prognosticators of ESCC. In this study, we focused on OPLAH, which is involved in glutathione metabolism pathway, overexpressing in ESCC tissues, because its prognostic significance has not been previously reported in malignancies, including ESCC.
Extraction of putative prognosticators of ESCC. (A) Volcano plot comparing ESCC tissue and normal esophageal mucosa in the TCGA dataset. Red plots indicate genes significantly overexpressed with FC ≥2, and blue plots indicate genes that are significantly down-regulated with FC ≤0.5 in ESCC tissue compared to normal mucosa. (B) The Venn diagram indicates the intersection of significantly overexpressed genes in ESCC (left circle) and the genes whose high expression is associated with poor prognosis (right circle). (C) The Venn diagram indicates the intersection of the significantly suppressed genes in ESCC (left circle) and the genes whose low expression is associated with poor prognosis (right circle).
OPLAH expression in the public database. OPLAH mRNA was significantly overexpressed in ESCC tissues compared to normal tissues (p=0.003, FC=2.237), and patients with high OPLAH mRNA expression had a significantly poorer prognosis (p=0.045) (Figure 2A and B). Consistently, high OPLAH mRNA expression was significantly associated with poor prognosis in the Kaplan–Meier Plotter (p=0.0042, HR=3.81, 95% CI=1.43-10.2) (Figure 2C).
OPLAH mRNA expression in public datasets. (A) Violin plot showing OPLAH mRNA expression in ESCC tissue and normal mucosa. (B) Survival curves for OPLAH mRNA expression in the TCGA dataset. (C) Survival curves for OPLAH mRNA expression in Kaplan–Meier plotter.
OPLAH expression in ESCC tissues. IHC was performed to verify whether OPLAH protein expression was able to predict ESCC patient prognosis. Staining intensity was categorized into three groups: high, intermediate, and low (Figure 3A). The overall survival (OS) of patients with high OPLAH protein expression was significantly shorter than that of patients with low expression (p<0.0001) (Figure 3B). Relapse-free survival (RFS) was also significantly poorer in patients with high expression levels of OPLAH (p<0.0001) (Figure 3C).
OPLAH protein expression in ESCC tissue and survival analysis. (A) The staining intensity for OPLAH protein was categorized into three groups: high, intermediate, and low. (B) Overall survival according to staining intensity. (C) Relapse-free survival according to staining intensity.
Next, the associations between OPLAH protein expression and clinicopathological factors were investigated. OPLAH expression was not significantly associated with any of the factors (Table I). Multivariable analysis identified high OPLAH protein expression as an independent predictive factor of OS after esophagectomy (p<0.0001, HR=4.83, 95% CI=2.88-8.09), together with pathological lymph node metastasis (pN) positive and pathological stage (pStage) defined by Unio Internationalis Contra Cancrum (UICC) 8th edition (Table II).
Association between OPLAH protein expression and clinicopathological parameters.
Multivariable analysis for overall survival.
OPLAH concentration in serum samples. The characteristics of patients from whom serum samples were collected are shown in Table III. The pre-NAC serum OPLAH protein concentration in stage III patients was significantly higher than that in stage II patients (p=0.0341) (Figure 4A). In most patients, the serum OPLAH protein concentration was significantly decreased by NAC regardless of histological grade (p=0.0070) (Figure 4B). The pre-NAC serum OPLAH protein concentration was significantly associated with clinical tumor depth (cT) (p=0.0263), clinical lymph node metastasis (cN) (p=0.0125), and clinical stage (cStage) (Figure 4C). Patients with high pre-operation (pre-op) serum OPLAH protein concentrations tended to have a poor prognosis, but this difference was not statistically significant due to small sample number (Figure 5A). There was no significant correlation between serum OPLAH protein concentration and the tumor markers carcinoembryonic antigen (CEA) and squamous cell carcinoma antigen (SCC) (Figure 5B and C).
Characteristics of the patients for ELISA analysis.
Association between serum OPLAH protein concentration and clinicopathological factors. (A) Violin plot shows the comparison of pre-NAC serum OPLAH protein concentrations in healthy volunteers and patients with cStage II and cStage III ESCC. (B) Change in serum OPLAH protein concentrations by NAC. (C) Violin plot shows the comparison of serum OPLAH protein concentrations with each clinicopathological factor. Factors that have a significant association with serum OPLAH protein concentrations are colored.
Association between serum OPLAH protein concentrations in serum and survival or conventional tumor marker for ESCC. (A) Overall survival according to pre-op serum OPLAH protein concentration. (B) Correlation analysis between pre-NAC serum OPLAH protein concentrations and pre-NAC tumor markers. (C) Correlation analysis between pre-op serum OPLAH protein concentrations and pre-op tumor markers.
Discussion
In this study, OPLAH was identified as a biomarker that was significantly associated with the prognosis of ESCC patients from comprehensive gene expression analysis. The clinical significance of its expression was demonstrated at the protein level, which is more suitable for clinical practice. The prognosis of patients with high OPLAH protein expression in ESCC tissue detected by IHC was significantly poor, and the serum OPLAH protein concentration before NAC was significantly associated with the clinical stage. This study is the first to report the prognostic significance of OPLAH expression in malignant tumors, including ESCC.
OPLAH is one of the enzymes involved in glutathione metabolism that catalyzes the conversion of 5-oxo-L-proline to L-glutamate concurrently with the hydrolysis of ATP to ADP and inorganic phosphate (20). Glutathione has the antioxidant effects and detoxifying effects, and these effects are achieved by capturing and excreting drugs that damage cells (21-23). Antioxidants, including glutathione, have been reported to play an important role in tumor initiation, proliferation, and metastasis (24-26). Cancer cells have been reported to acquire resistance to anticancer drugs and radiation therapy by retaining high concentrations of glutathione and consequently develop their malignant phenotype (27, 28). Chemotherapy and radiation therapy are essential modalities in the definitive or palliative treatment of ESCC. High OPLAH expression may help to explain the poor prognosis due to treatment resistance by the activation of glutathione metabolism.
The association between OPLAH and malignancies has only been reported as a part of the results of a comprehensive analysis, and no studies have reported the clinical significance of OPLAH alone. In colorectal cancer, OPLAH has been reported as one of the DNA loci that are abnormally hypermethylated compared to normal colorectal tissue, but its clinical significance is unknown (29-32). Singh et al. identified OPLAH as one of the genes that was overexpressed as a brown adipose marker that promotes the proliferation of breast cancer cells using a xenograft mice model, however, its clinical significance has not been mentioned (33). In gastric cancer, the significance of OPLAH alone was not verified, although, OPLAH was one of several genes that was identified as a prognostic marker by TCGA analysis (34, 35). Shi et al. reported that glutathione metabolism contributes to cisplatin (CDDP) resistance in non-small cell lung cancer by comparing a cisplatin-resistant cell and its parent cell (36). The mRNA expression of CD13, GPX4, RRM2B, and OPLAH, which are involved in glutathione metabolism, was increased in CDDP-resistant cells, and CDDP sensitivity was restored by their inhibitors or RNA interference. Their results support the usefulness of OPLAH as a prognosticator in ESCC patients. Our study is the first to report the utility of OPLAH alone as a prognosticator.
We identified genes that may be involved in prognosis from comprehensive gene expression analysis. Additionally, to narrow down genes with putative oncogenic functions, differentially expressed genes between ESCC tissues and normal tissues were selected (Figure 1). This selection criterion has potential to identify not only biomarkers but also future therapeutic targets (37). Twenty-four genes were identified as candidate prognosticators of ESCC according to our criteria. OPLAH was the focus of this study because its individual expression significance has not been previously reported.
The results showed that the patients with high OPLAH mRNA expression had significantly poorer prognosis in two public databases. Furthermore, the consistent significance obtained by protein expression analysis of our cohort suggested the robustness of OPLAH as a prognostic biomarker (Figure 2 and Figure 3). Moreover, patient prognosis was clearly stratified by the IHC intensity (Figure 3). This demonstrates that OPLAH expression is a suitable biomarker that predicts the prognosis of ESCC patients. This finding is based on protein expression analysis, which is currently the most frequently used analysis in clinical practice. OPLAH expression was not significantly associated with any of the clinicopathological factors (Table I). These results suggest that OPLAH protein expression in ESCC tissues may be a novel prognostic factor independent of the existing clinicopathological factors. Multivariable analysis revealed that high OPLAH protein expression was an independent prognosticator for OS (Table II). OPLAH protein expression could be a useful prognosticator without confounding with pathological stage and pathological lymph node metastasis positivity.
Quantification of mRNA that requires quantitative polymerase chain reaction is not widely generalized in clinical practice, whereas IHC and ELISA for evaluating protein expression are widely used clinically. So, targeting proteins as biomarkers is suitable in clinical practice. We examined not only protein expression in tissues, but also serum OPLAH protein concentrations, which can be collected repeatedly with less invasiveness. Serum OPLAH protein concentrations before modification by NAC were significantly higher in cStage III than in cStage II (Figure 4A). No significant difference was obtained between healthy volunteers and individuals with cStage III, and this was considered to be due to the small number of healthy samples. We evaluated serial serum samples before and after NAC that were obtained from the same patient cohort to confirm the usefulness of serum OPLAH protein concentration as a biomarker. The serum OPLAH protein concentration was decreased by NAC in most patients regardless of histological therapeutic effect of the primary lesion (Figure 4B), and the pre-NAC serum OPLAH protein concentration was significantly associated with cT, cN and cStage (Figure 4C). There were no significant associations between staining intensity and pathological tumor depth, pN or pStage. These results could be due to the use of IHC to evaluate the local protein expression. Serum protein concentrations are expected to reflect systemic tumor burden, such as tumor volume and micrometastases. The result that the serum OPLAH protein concentration was decreased by NAC, which was performed to control micrometastases, supports this hypothesis. Although no significant difference was indicated due to the limitation of a small sample number, the prognostic curve based on pre-op OPLAH concentrations tended to be similar to that based on OPLAH mRNA expression (Figure 5A). This tendency was not observed in the pre-NAC OPLAH concentrations. These results are consistent with the results of clinical trials showing that patient prognosis is affected by NAC (38). Furthermore, the lack of correlations between tumor markers used in clinical practice and the serum OPLAH protein concentration indicates that OPLAH may be useful as a prognosticator from another aspect that is not affected by currently used prognostic indicators (Figure 5B and C).
This study had certain limitations. First, while our study revealed the utility of OPLAH as a prognosticator of ESCC, its biological function has not been uncovered. However, we demonstrated the clinical significance of OPLAH at the protein expression level, which is directly linked to clinical practice. Furthermore, its applicability in blood tests shows that OPLAH could be a less invasive and repeatedly collectable biomarker. Second, the case number of the two cohorts used for validation for IHC and ELISA is not sufficient, but two public databases concurrently support the robustness of OPLAH as a prognostic biomarker. Assays for the biological function of OPLAH contributing malignant phenotype and treatment resistance and verification of it as a prognostic biomarker in more sample number are required.
In conclusion, ESCC patients with high OPLAH expression had a significantly poorer prognosis, and serum OPLAH protein concentration reflected disease progression. Our results indicate that OPLAH may serve as a useful biomarker for the prognosis of ESCC.
Acknowledgements
The Authors thank Yoko Nishikawa and Yasuko Iguchi for their technical assistance in this study.
Footnotes
Conflicts of Interest
The Authors declare no competing financial interests.
Authors’ Contributions
D. Shimizu conceived and designed the research, performed analysis, and wrote the paper. T. Kishida, S. Nakamura, M. Sasahara, and S. Ueda performed the experimental procedure. M. Kanda, Y. Sato, and S. Motoyama collected clinical samples and data. Y. Inokawa, N. Hattori, M. Hayashi, and C. Tanaka proofread the paper. M. Kanda and Y. Kodera supervised the study.
- Received April 9, 2023.
- Revision received May 4, 2023.
- Accepted May 19, 2023.
- Copyright © 2023 The Author(s). Published by the International Institute of Anticancer Research.
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).
