Skip to main content
NCBI home page
Medicine logoLink to Medicine
. 2020 Jan 31;99(5):e18666. doi: 10.1097/MD.0000000000018666

Low EIF2B5 expression predicts poor prognosis in ovarian cancer

Lin Hou a, Yan Jiao b, Yanqing Li c, Zhangping Luo c, Xueying Zhang d, Guoqiang Pan e, Yuechen Zhao f, Zhaoying Yang g,∗, Miao He h,∗
Editor: Hua Yang
PMCID: PMC7004721  PMID: 32000373

Abstract

Ovarian cancer has the highest mortality among gynecological cancers. Although ovarian cancer usually responds well to chemotherapy, most patients still have a poor prognosis. EIF2B5 is a crucial molecule in posttranscriptional modifications involved in tumor progression, and here we investigated the prognostic role of EIF2B5 in ovarian cancer. We examined the differential expression of EIF2B5 mRNA in ovarian cancer by exploring The Cancer Genome Atlas (TCGA) database. The chi square test was used to identify a clinical correlation. Survival analysis and Cox regression model were performed to determine the association between EIF2B5 expression and overall survival (OS) in ovarian cancer patients. As a result, Low EIF2B5 expression was found in ovarian cancer tissues and correlated with survival status. Survival analysis showed that ovarian cancer patients with low EIF2B5 expression had a short OS. Moreover, Cox regression analysis indicated that low EIF2B5 expression was an independent risk factor for a poor prognosis in ovarian cancer. Additionally, according to gene set enrichment analysis, mesenchymal transition, angiogenesis, coagulation, and bile acid metabolism were differentially enriched in ovarian cancer with high EIF2B5 expression. In conclusion, Low EIF2B5 expression is an independent risk factor for a poor prognosis in ovarian cancer patients.

Keywords: data mining, EIF2B5, ovarian cancer, prognosis, TCGA

1. Introduction

Ovarian cancer is the most lethal gynecological cancer globally,[1,2] and despite rapid advancements in treatment methods, the prognosis of ovarian cancer patients remains poor, with few effective prognostic biomarkers available at present.[3–5] Therefore, there is an urgent need for new molecular markers that can be used to predict the prognosis of ovarian cancer patients for the purpose of guiding treatment planning.

As a crucial molecule in posttranscriptional modifications, eukaryotic translation initiation factor 2B subunit 5 (EIF2B5) is important in cancer progression.[6] Early research regarding EIF2B5 mainly aimed to study the roles of its expression in multiple sclerosis,[7,8] ovarioleukodystrophy,[9] and vanishing white matter (VWM) syndrome.[10–12] Additionally, recent studies indicated that high EIF2B5 expressed in few cancerous tissue (lung cancer,[13] breast cancer,[14] and liver cancer[15]) and serve as a prognostic biomarker in hepatocellular carcinoma.[15] Nonetheless, the role of EIF2B5 expression in ovarian cancer remains unclear.

To evaluate the clinical significance of EIF2B5 expression in the prognosis of ovarian cancer patients, we analyzed the prognostic value of EIF2B5 mRNA expression in The Cancer Genome Atlas (TCGA) cohort of ovarian cancer patients. First, we analyzed the differential expression of EIF2B5 in ovarian cancer patients and then studied its correlation with overall survival (OS) among the patients.

2. Materials and methods

2.1. Data source

The clinical data for EIF2B5 expression in normal ovarian tissues and ovarian cancer tissues were downloaded from the TCGA (https://cancergenome.nih.gov/) and GTEx (www.gtexportal.org/) databases in May 2018.

2.2. Data mining

All data mining was conducted using R (version 3.5.1).[16] The differences in EIF2B5 expression according to clinical features are shown in boxplots drawn using the ggplot2 package.[17] To determine the high and low EIF2B5 expression groups, and the optimal cutoff value was obtained from ROC curve. Possible clinical correlations between EIF2B5 expression and the clinical characteristics of ovarian cancer patients were evaluated by the chi square test. The survival curves were drawn using Survival Package.[18] The log-rank test was applied to examine the survival difference. Univariate Cox analysis was performed to select relevant variables, and a multivariate Cox model was used to evaluate the independent prognostic role of EIF2B5 expression separate from other clinical characteristics.

2.3. GSEA

Gene set enrichment analysis (GSEA) uses predefined gene sets to rank target genes according to the degree of differential expression between the two types of samples, and then to test whether the pre-defined gene sets are at the top or bottom of the sorting table.[19] In the present study, we used GSEA 3.0 software to analyze the data of ovarian cancer patients. We obtained standardized enrichment fractions (NESs) by using permutation analysis 1000 times.

2.4. Ethical approval

Ethics committee approval was not necessary because all clinical data used in this study were obtained from a public database and are available for research.

3. Results

3.1. Differential expression of EIF2B5 in ovarian cancer

The data for EIF2B5 expression and clinical features including age, subdivision of cancer, cancer stage, longest dimension, lymphatic invasion, histologic grade, occurrence type, sample type, vital status, and EIF2B5 expression are presented in Table 1 and Figure 1A. From the prepared boxplots, EIF2B5 expression was low in ovarian cancer tissues compared with that observed in normal ovarian tissues. Additionally, low EIF2B5 expression was observed in deceased patients, suggesting a potential connection between the survival status and EIF2B5 expression (Fig. 1 and Table 2).

Table 1.

Demographic and clinical characteristics of TCGA ovarian cancer cohort.

3.1.

Open in a new tab

Figure 1.

Figure 1

Open in a new tab

Differential EIF2B5 expression in ovarian cancer. The EIF2B5 expression in all ovarian cancer cases and different groups according to histologic grade, occurrence type, subdivision, lymphatic invasion, patient age, stage, and vital status.

Table 2.

Correlation between EIF2B5 expression and clinicopathologic characteristics in ovarian cancer.

3.1.

Open in a new tab

3.2. Correlation of EIF2B5 expression and survival

To explore possible correlations of EIF2B5 expression with clinical factors, we completed the chi square test and found a specific correlation between vital status and expression of EIF2B5 (Table 2). Moreover, patients with shorter OS time had much lower expression of EIF2B5 (Fig. 2, P = .034), which is consistent with results of subgroup analysis, especially among the elderly patients (Fig. 2, P = .022).

Figure 2.

Figure 2

Open in a new tab

Survival analysis for groups of ovarian cancer cases with differing EIF2B5 expression in ovarian cancer and subgroup analysis according to early stage, advanced stage, G1 and G2, G3 and G4, lymphatic invasion, non-lymphatic invasion, younger, and older.

The univariate Cox model revealed several potential survival-related variables including age, occurrence type, and EIF2B5 expression. The Multivariate Cox model suggested that low EIF2B5 expression was an independent risk factor for a poor prognosis in ovarian cancer patients, based on its association with a shorter OS (hazard ratio [HR] = 1.82, P = .008, Table 3).

Table 3.

Univariate and multivariate Cox regression analyses of overall survival duration.

3.2.

Open in a new tab

As shown in Table 4, GSEA revealed significant differences in the enrichment of the MSigDB Collection (NOM P < .05, false discovery rate [FDR] < 0.25). We chose the most essential signaling pathways based on NES (Table 4; Fig. 3). Figure 3 shows that mesenchymal transition, angiogenesis, coagulation, and bile acid metabolism were enriched in low EIF2B5 expression phenotype, respectively.

Table 4.

Gene set enrichment with low EIF2B5 expression.

3.2.

Open in a new tab

Figure 3.

Figure 3

Open in a new tab

Enrichment plots from GSEA.

4. Discussion

Although many advances in treatment strategies for ovarian cancer have been explored, the OS of these patients has not been improved. Thus, novel biomarkers that can be used to predict the prognosis of these patients remain urgently needed.[20–26] According to the results of the present study, low EIF2B5 expression is an independent risk factor for a poor prognosis among ovarian cancer patients.

Early studies of EIF2B5 mainly focused on its role in VWM diseases, which involve downregulation of EIF2B5.[8,10,11] Recently, several studies began investigating the role of EIF2B5 in various cancers, including lung cancer,[13] breast cancer,[14] and liver cancer.[15] In these studies, EIF2B5 was shown to be overexpressed at both mRNA and protein levels in the cancerous tissues. In contrast, in the present study we observed the opposite phenomenon in which EIF2B5 expression was lower in ovarian cancer tissues than in normal ovarian tissues. This discrepancy may be due to differences in the cancer types, which might suggest exclusive functions and mechanisms of EIF2B5 in ovarian cancer. Moreover, EIF2B5 expression showed a decreasing trend from stage I to stage IV ovarian cancer, suggesting that the function of EIF2B5 may change throughout different stages of ovarian cancer. To better understand the dynamics of EIF2B5 expression in ovarian cancer, a subgroup analysis is necessary. Additionally, considering that the low EIF2B5 expression continued to decline with disease progression, the relationship between EIF2B5 and survival needs to be further studied.

Previous research also linked EIF2B5 expression with cancer patients’ prognosis. A previous study demonstrated that high EIF2B5 expression is associated with a shorter survival time in colorectal cancer cases.[27] Also, expression of minor alleles of EIF2B5 was found to improve the prognosis of ovarian cancer patients via the inhibition of angiogenesis and tumor growth.[28] However, the association between EIF2B5 expression and OS remains unknown in ovarian cancer. In the present study, we found that the overall survival time of ovarian cancer patients with low EIF2B5 expression was short, while subgroup analysis revealed the same phenomenon with differences in the stage and histologic grade of ovarian cancer. Interestingly, we found that the survival difference was especially significant in older patients. However, this study doesn’t contain the variables like race and cancer type, because the races information of TCGA is absent, and only epithelial type exists. Future study needs to explore these variables in other population. Moreover, mesenchymal transition, angiogenesis, coagulation, and bile acid metabolism may be signaling pathways of EIF2B5 in ovarian cancer.

To the best of our knowledge, this is the first study analyzing the prognostic value of the EIF2B5 expression in ovarian cancer. Together with other studies of EIF2B5, our study provides insight into the role of EIF2B5 expression in various cancer types. However, as we did not explore the underlying mechanism of the function of EIF2B in ovarian cancer, future in vitro and in vivo experiments are needed to explore the mechanism in depth.

5. Conclusion

In the present study, we investigate the predictive value of EIF2B5 expression for ovarian cancer patients’ prognosis. We found that low EIF2B5 expression was an independent risk factor for a shorter survival time among ovarian cancer patients. Our future research will include in vitro and in vivo experiments to explore the underlying mechanism of this relationship in depth.

Author contributions

Conceptualization: Lin Hou, Yan Jiao, Zhaoying Yang, Miao He.

Data curation: Yanqing Li.

Formal analysis: Yanqing Li.

Funding acquisition: Zhaoying Yang.

Investigation: Zhangping Luo.

Project administration: Zhangping Luo, Miao He.

Resources: Zhangping Luo, Miao He.

Software: Yanqing Li.

Supervision: Zhangping Luo, Yuechen Zhao.

Validation: Xueying Zhang, Yuechen Zhao, Zhaoying Yang, Miao He.

Visualization: Yanqing Li, Xueying Zhang.

Writing – original draft: Yan Jiao.

Writing – review & editing: Lin Hou, Guoqiang Pan, Zhaoying Yang, Miao He.

Footnotes

Abbreviations: EIF2B5 = eukaryotic translation initiation factor 2B subunit 5, GSEA = gene set enrichment analysis, NES = normalized enrichment score, OS = overall survival, TCGA = the cancer genome atlas.

How to cite this article: Hou L, Jiao Y, Li Y, Luo Z, Zhang X, Pan G, Zhao Y, Yang Z, He M. Low EIF2B5 expression predicts poor prognosis in ovarian cancer. Medicine. 2020;99:5(e18666).

This study was supported by Science and Technology of Jilin Province Health and Family Planning Commission Project 2017Q035 (ZY).

The authors report no conflicts of interest in this work.

References

  • [1].Fidler MM, Bray F, Soerjomataram I. The global cancer burden and human development: A review. Scand J Public Health 2018;46:27–36. [DOI] [PubMed] [Google Scholar]
  • [2].Torre LA, Islami F, Siegel RL, et al. Global cancer in women: burden and trends. Cancer Epidemiol Biomarkers Prev 2017;26:444–57. [DOI] [PubMed] [Google Scholar]
  • [3].Eisenhauer EA. Real-world evidence in the treatment of ovarian cancer. Ann Oncol 2017;28: suppl_8: viii61–5. [DOI] [PubMed] [Google Scholar]
  • [4].Mallen A, Soong TR, Townsend MK, et al. Surgical prevention strategies in ovarian cancer. Gynecol Oncol 2018;151:166–75. [DOI] [PubMed] [Google Scholar]
  • [5].Pignata S, C Cecere S, Du Bois A, et al. Treatment of recurrent ovarian cancer. Ann Oncol 2017;28: Suppl 8: viii51–6. [DOI] [PubMed] [Google Scholar]
  • [6].Brady LK, Wang H, Radens CM, et al. Transcriptome analysis of hypoxic cancer cells uncovers intron retention in EIF2B5 as a mechanism to inhibit translation. PLoS Biol 2017;15:e2002623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Zahoor I, Asimi R, Haq E. No evidence for a role of Ile587Val polymorphism of EIF2B5 gene in multiple sclerosis in Kashmir Valley of India. J Neurol Sci 2015;359:172–6. [DOI] [PubMed] [Google Scholar]
  • [8].Zahoor I, Haq E, Asimi R. Multiple Sclerosis and EIF2B5: a paradox or a missing link. Adv Exp Med Biol 2017;958:57–64. [DOI] [PubMed] [Google Scholar]
  • [9].Ibitoye RT, Renowden SA, Faulkner HJ, et al. Ovarioleukodystrophy due to EIF2B5 mutations. Pract Neurol 2016;16:496–9. [DOI] [PubMed] [Google Scholar]
  • [10].Esmer C, Blanco Hernandez G, Saavedra Alanis V, et al. Association between homozygous c.318A>GT mutation in exon 2 of the EIF2B5 gene and the infantile form of vanishing white matter leukoencephalopathy. Bol Med Hosp Infant Mex 2017;74:364–9. [DOI] [PubMed] [Google Scholar]
  • [11].Raini G, Sharet R, Herrero M, et al. Mutant eIF2B leads to impaired mitochondrial oxidative phosphorylation in vanishing white matter disease. J Neurochem 2017;141:694–707. [DOI] [PubMed] [Google Scholar]
  • [12].Bektas G, Yesil G, Ozkan MU, et al. Vanishing white matter disease with a novel EIF2B5 mutation: A 10-year follow-up. Clin Neurol Neurosurg 2018;171:190–3. [DOI] [PubMed] [Google Scholar]
  • [13].Xue D, Lu M, Gao B, et al. Screening for transcription factors and their regulatory small molecules involved in regulating the functions of CL1-5 cancer cells under the effects of macrophage-conditioned medium. Oncol Rep 2014;31:1323–33. [DOI] [PubMed] [Google Scholar]
  • [14].Yang S, Zhang H, Guo L, et al. Reconstructing the coding and non-coding RNA regulatory networks of miRNAs and mRNAs in breast cancer. Gene 2014;548:6–13. [DOI] [PubMed] [Google Scholar]
  • [15].Jiao Y, Fu Z, Li Y, et al. High EIF2B5 mRNA expression and its prognostic significance in liver cancer: a study based on the TCGA and GEO database. Cancer Manag Res 2018;10:6003–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Team RDCJC. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2009;14:12-21. [Google Scholar]
  • [17].Wickham H. Ggplot2: elegant graphics for data analysis. J R Stat Soc 2011;174:245–6. [Google Scholar]
  • [18].Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. Vol 97. New York: Springer; 2000. [Google Scholar]
  • [19].P. T, VK M, S M, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. A Subramanian A 2005;102:15545–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Jiao Y, Fu Z, Li Y, et al. Aberrant FAM64A mRNA expression is an independent predictor of poor survival in pancreatic cancer. PloS One 2019;14:e0211291. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Jiao Y, Li Y, Lu Z, et al. High trophinin-associated protein expression is an independent predictor of poor survival in liver cancer. Dig Dis Sci 2019;64:137–43. [DOI] [PubMed] [Google Scholar]
  • [22].Jiao Y, Li Y, Liu S, et al. ITGA3 serves as a diagnostic and prognostic biomarker for pancreatic cancer. Onco Targets Ther 2019;12:4141–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Jiao Y, Li Y, Jiang P, et al. PGM5: a novel diagnostic and prognostic biomarker for liver cancer. Peer J 2019;7:e7070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Li Y, Jiao Y, Fu Z, et al. High miR-454-3p expression predicts poor prognosis in hepatocellular carcinoma. Cancer Manag Res 2019;11:2795–802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Sun Z, Sun L, He M, et al. Low BCL7Ax expression predicts poor prognosis in ovarian cancer. J Ovarian Res 2019;12:41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [26].Hou L, Zhang X, Jiao Y, et al. ATP binding cassette subfamily B member 9 (ABCB9) is a prognostic indicator of overall survival in ovarian cancer. Medicine (Baltimore) 2019;98:e15698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [27].Palaniappan A, Ramar K, Ramalingam S. Computational identification of novel stage-specific biomarkers in colorectal cancer progression. PloS One 2016;11:e0156665. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [28].Goode EL, Maurer MJ, Sellers TA, et al. Inherited determinants of ovarian cancer survival. Clin Cancer Res 2010;16:995–1007. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Medicine are provided here courtesy of Wolters Kluwer Health

RESOURCES