OSC is one of the leading lethal malignancies worldwide. The slow progress of molecular targeted therapy and the absence of effective molecular markers for OSC prognostic monitoring make it necessary to better understand the molecular mechanisms leading to this condition. The exploration of metabolic mechanism opens up and important perspective for OSC[1].
Although the significance of MRGs in cancer development and progression has been well-established, no comprehensive, genome-wide analysis has been conducted to explore its clinical significance and molecular mechanism. Most importantly, a personalized metabolic signature based on the selection, differential expression of MRGs is presented to measure cancer cell development and evaluate potential clinical outcomes. Since the beginning of the "War on Cancer", our understanding of carcinogenesis and clinical management techniques has made remarkable progress, but many aspects of OSC metabolic-related molecular mechanisms still unclear. This comprehensive and complete analysis of MRGs in OSC improve our understanding of its clinical implications and clarifies its underlying molecular features. The large number of OSC samples based on bioinformatics we were exposed to in this study contributed to robust results.
In recent years, with the development of high-throughput sequencing technology, large databases like TCGA, SEER and GEO have emerged, which provide an effective means for the selection of genetic markers. In the current study, we dug into the expression profile of MRGs in TCGA in order to search for molecular markers to detect the prognosis of patients with OSC. We first screened 1466 differentially expressed MRGs in OSC and normal ovarian tissues. Considering that these genes may have a closely association with the development and progression of OSC, we performed GO and KEGG analyses on these genes. Interestingly, functional analysis showed that the KEGG pathway (metabolism-related pathway), the most important of these enriched genes, was reduced. Based on the above results, we speculated that tumor metabolism may play an important role in the process of tumorigenesis. tumor metabolism is of great concern; of particular interest is its multifaceted feature in tumorigenesis. Glutamine, amino acids, glucose and free fatty acids are the basic and significant substances that support the growth and survival of cancer cells. These metabolites are either synthesized in cancer cells or assimilated from the blood circulation[12]. In summary, a better understanding of the relationship between cell origin and its metabolic status, as well as the function of MRGs, will help better map the metabolic profile of OSC.
When performed RWR to the network, ACOT7, CERK, EHMT2, MTAP and PDE8A were also identified in association with the pathogenesis of OSC. In the network of these initial genes, it could interact with 200 metabolic genes considered as the most closely related to the pathogenesis of ovarian cancer. On the basis of univariate cox analysis, a total of 8 genes were closely related to the prognosis of ovarian cancer patients, respectively ENPP1, FH, CYP2E1, HPGDS, ADCY9, NDUFA5, ADH1B and PYGB. Further analysis helped us distinguish high-risk and low-risk group to develop the metabolic-based prognostic index, which could be an independent prognostic indicator for OSC patients. Furthermore, we explored its expression profile, prognostic value and mutation status, and found valuable data for future clinical studies. To explore potential molecular mechanisms corresponding to potential clinical value, we constructed a MRGs network to reveal important and hub MRGs that regulate the tumor metabolism and progression.
ENPP1, FH, CYP2E1, HPGDS, ADCY9, NDUFA5, ADH1B and PYGB it characterized in this network. Given the potential molecular mechanisms of the eight MRG, reports on the functions and mechanisms of HPGDS, ADCY9 and NDUFA5 have not been published on OSC[13–17]. However, five of these eight hub MRGs have been studied, namely, ENPP1, FH, CYP2E1, ADH1B and PYGB. ENPP1 is increased in ovarian cancer and may promote the migration ability[13]. High level of FH could promote the aggressive and metastatic behaviors.[18] Increased activity of CYP2E1 was correlated with raised serum levels of IL-6, IL-8, and TNF-α, which mediated drug metabolism, and may have profound effects for drug development and prescribing in oncological settings[19]. High expression of ADH1B was correlated with markedly higher risk of residual disease in OSC[20], which played a significant role in accelerating ovarian cancer cell infiltration and may enhance the possibility of postoperative residual lesions[21]. PYGB obviously promoted ovarian cancer cell proliferation, invasion and migration via wnt pathway[22]. Therefore, previous studies just provided limited information on the mechanism of 8 MRGs in OSC patient survival. Oxidative phosphorylation pathway is the most important pathway in functional enrichment analysis, and it is speculated that oxidative phosphorylation pathway may play an important role in OSC process. Taken together, this study proposed a signature with metabolic-based prognostic index as the endpoint, which was most suitable for the survival monitoring of OSC patients. In addition, metabolic-based prognostic index is not only a prognostic indicator, but also an indicator of metabolic status.
Nowadays, some of the prognostic characteristics of cancer based on expression profiles have been proposed with the help of large public databases. For example, Zhong et al. also indicated a prognostic marker with 6 genes as a potential survival prediction marker for ER-positive breast cancer patients[23]. Bao et al. analyzed RNA-Seq data of 234 BC patients from TCGA and successfully obtained 4-lncRNA signature, which has prognostic value[24]. Nevertheless, these researches only focused on classic tumor biological behavior and ignored tumor metabolic. We attach enormous interests to the classic biological behavior as well as tumor metabolic. Therefore, prognostic characteristics are expected to be translated into clinical applications. However, the limitation of this study lies in its retrospective nature. Due to the lack of sufficient cases, we were unable to detect the expression of ENPP1, FH, CYP2E1, HPGDS, ADCY9, NDUFA5, ADH1B and PYGB in OSC and normal ovarian tissues. As we look to the future, there are still many problems. For instance, the relationship among metabolomics, immune genomics, proteomics and epigenomics should be investigated to further describe global metabolic alterations in OSC. It is important to further explore the potential relationship between metabolomics disorders and precancerous lesions. We anticipate that this prognostic feature may have important clinical significance. We systematically analyzed the role of MRGs in monitoring the occurrence and prognosis of OSC. Our findings provide new ideas for individual treatment of OSC.