Identification And Validation Of A Hypoxia ‑ Related Prognostic Signature In Clear Cell Renal Cell Carcinoma Patients

doi:10.21203/rs.3.rs-48542/v1

Download PDF

Primary research

Identification And Validation Of A Hypoxia ‑ Related Prognostic Signature In Clear Cell Renal Cell Carcinoma Patients

https://doi.org/10.21203/rs.3.rs-48542/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Increasing evidence has shown that hypoxia is closely related to the development, progression and prognosis of clear cell renal cell carcinoma (ccRCC). Nevertheless, reliable prognostic signatures based on hypoxia have not been well-established. This study aimed to construct an optimized prognosis nomogram based on hypoxia-related genetic signatures for patients with ccRCC.

Method: We accessed hallmark gene sets of hypoxia, including 200 genes, and an original RNA seq dataset of ccRCC cases with integrated clinical information obtained by mining the Molecular Signatures Database, the TCGA database and the ICGC database. Univariate Cox regression analysis and multivariate Cox proportional hazards regression were performed to identify prognostic hypoxia-related genetic signatures and further generate RiskScore, a new independent prognosis predictor for optimizing prognosis models. External validation of the optimized prognosis model was performed in independent cohorts from the ICGC database.

Result: ANKZF1, ETS1, PLAUR, SERPINE1, FBP1 and PFKP were selected as hypoxia-related genetic signatures, and the resultant formula based on those genetic signatures and their respective coefficients helped generate RiskScore. The results of receiver operating characteristic (ROC) curve, risk plot, survival analysis and so on suggested that RiskScore based on hypoxia-related genetic signatures was an independent risk factor. A novel prognosis nomogram optimized via RiskScore showed its promising performance in both a TCGA-ccRCC cohort and an ICGC-ccRCC cohort.

Conclusions: Our study reveals that the differential expressions of hypoxia-related genes are associated with the overall survival of patients with ccRCC. RiskScore based on hypoxia-related genetic signatures was an independent risk factor beyond TNM staging and grading. The novel nomogram optimized via RiskScore exhibited a promising prognostic ability. It may be able to serve as a prognostic tool for guiding clinical decisions and selecting effective individualized treatments.

Cancer Biology

clear cell renal cell carcinoma

hypoxia

prognosis

prognostic model

nomogram.

Renal cell carcinoma (RCC) is a common urological tumor, with the sixth highest incidence among male cancer. It is estimated that 73,750 new cases and 14,830 deaths due to RCC will occur in the United States in 2020 (1). There are various histological subtypes of RCC, among which clear cell renal cell carcinoma (ccRCC) accounts for approximately 70% (2). The etiology of RCC is still unclear, although heredity, smoking, obesity and hypertension are recognized risk factors (3–5). Patients with ccRCC lack symptoms in the early stage, so it is usually found incidentally on imaging. When typical symptoms such as hematuria, lumbago and abdominal mass take place, many patients are diagnosed with metastatic carcinoma, whose five-year overall survival rate is dismal at 8–12% (6). In addition, 20–30% localized ccRCC patients will progress to metastasis despite nephrectomy (7). Thus, there is an exigence for new molecular markers to identify patients with a high risk of progress and poor prognosis to alert clinicians.

Hypoxia is a characteristic of solid tumors that profoundly affects the expression of many noncoding RNAs (8, 9). It is reported that hypoxia directly contributes to many hallmarks of cancer, including reprogramming metabolism, proliferation, invasion and metastasis, apoptosis and resistance to therapy (10, 11). Hypoxia inducible factor (HIF) is a protein complex, composed of either HIF-1a or HIF-2a and HIF-1b/ARNT subunits, that centrally regulates cellular oxygen detection and adaptation (12). Hypoxia and overexpression of HIF are associated with prognosis in various cancers. For example, in one study a poor disease outcome group had significantly higher HIF-1α expression in stage IIB–IIIB cervical cancer (13). With the further research of hypoxia and ccRCC progress mechanism, it is found that the HIF family of proteins (either directly or combined with glucose uptake and the glycolysis pathway) regulates many downstream targets to contribute to the development and progression of ccRCC (14). Therefore, hypoxia-related genes may affect the prognosis of typical cancer patients, such as gastric cancer (15). However, whether hypoxia-related genes can be used as a prognostic indicator of ccRCC is still a mystery.

In this study, we obtained and processed RNA transcriptome data and clinical information of ccRCC patients from the Cancer Genome Atlas (TCGA) database and the International Cancer Genome Consortium (ICGC) database. We then further established a hypoxia-related prognosis signature and an optimized nomogram convenient for clinicians to use, which was verified using ICGC data.

2.1 Data collection and processing

The hallmark gene sets of hypoxia, including 200 genes, were downloaded from the Molecular Signatures Database (MSigDB version 6.0; http://software.broadinstitute.org/gsea/msigdb/index.jsp). The original RNA seq data set and clinical information of the ccRCC dataset were downloaded from the TCGA database (https://www.cancer.gov/tcga) and ICGC database (https://icgc.org/). R software (version 4.0.0; https://www.r-project.org/) was used to standardize and process data.

2.2 Screening and visualization of hypoxia‑related differentially expressed genes (DEGs)

We used the “limma” R package to screen the RNA raw data and excluded genes whose average count value was lower than one. All hypoxia genes whose |log2 fold change| was above one and false discovery rate (FDR) was lower than 0.05 were defined as hypoxia‑related DEGs. The “pheatmap” algorithm was used to draw a volcano plot and a heatmap for the hypoxia‑related DEGs.

2.3 Functional enrichment and protein–protein interaction (PPI) network analysis

We evaluated the hypoxia‑related DEGs according to their gene ontology (GO), including analysis of cellular components, molecular function and biological process involvement, and also based on the Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis. Both enrichment analyses were performed using "clusterProfiler," "org.Hs.eg.db," "enrichplot," "ggplot2" and "GOplot" packages in R software. P < 0.05 and FDR < 0.05 were considered statistically significant. Then, we used the string database (https: //string-db.org/) to construct the PPI network of hypoxia-related DEGs (CombinedScore = 0.40) and imported the data into Cytoscape software (version 3.7.2; https://cytoscape.org/) to visualize the interaction of the PPI network.

2.5 Construction of prognostic model

Among a total of 537 ccRCC patients, data related to seven patients were excluded from analysis because they were missing some information. Using the “survival” R package, we performed univariate Cox regression analysis on all hypoxia-related DEGs to narrow the candidate list to 16 hub hypoxia genes related to ccRCC prognosis (P < 0.001). We also evaluated the impact of these hypoxia genes on the survival time and outcome of patients through multivariate Cox proportional hazard regression. Next, we constructed a predictive model using expression data of the six hub hypoxia genes. The risk score of each patient was calculated according to the following equation:

where β represents a coefficient value that was identified based on the impact of each gene, and Exp represents the expression level of each hypoxia gene. The patients were divided into two subgroups (low-risk and high-risk) using the median risk score as a threshold. Overall survival (OS) of the two groups was evaluated by the Kaplan-Meier plot and log-Rank test using “survival” and “survminer” packages in R. The plots of the survival status of patients and heatmap were drawn using the “pheatmap” package in R. Additionally, we carried out receiver operating characteristic (ROC) curve analysis using the “survivalROC” package to evaluate the overall diagnostic performance of the proposed model. Moreover, we performed univariate Cox regression analysis and multivariate Cox proportional hazards regression with the “survival” package to investigate whether the risk score was an independent prognosis predictor of ccRCC. RiskScore, age, gender, tumor subtype, pathological stage and histological grade were used as covariates.

2.6 Correlation between the expression of the six hypoxia-related genes and survival time as well as clinical characteristics of ccRCC patients

We obtained the relationship between the expression levels of these six hypoxia-related genes and the survival time of ccRCC patients based on the Gene Expression Profiling Interactive Analysis database (GEPIA; http://gepia.cancer-pku.cn/). After that, the analysis of the six genes combined with clinical characteristics was performed. In order to further confirm the differential expression of the six genes between tumor and normal tissues, as well as their correlations with pathological characteristics, data of immunohistochemistry in ccRCC and normal kidney tissue from the Human Protein Atlas database (HPA, https://www.proteinatlas.org/) were used to verify the expression of the six genes at the translational level.

2.7 The establishment and validation of nomogram

Age, gender, stage, grade, T, N, M and RiskScore were used to construct the nomogram, using the“rms” and “survival” packages to ease the use of this new prognostic tool and allow quantitative prognostic evaluation of patients. We refer to this optimized nomogram as the new model and the corresponding nomogram without the RiskScore variable as the old model. Then, C-index, calibration curves, decision curve analysis (DCA), integrated discrimination improvement (IDI) and net reclassification index (NRI), which were used as indices to evaluate the advantages and disadvantages of the new and old models, were calculated and drawn to determine which model is better. After that, C-index and calibration curves were calculated and drawn to assess the consistency between actual and predicted survival between TCGA-ccRCC and ICGC-ccRCC cohorts, respectively.

3.1 Differentially expressed hypoxia-related genes

In this study, we used several advanced algorithms to identify hypoxia-related DEGs. Figure 1 shows the workflow of our research. Transcriptome data from 539 tumor samples and 72 normal tissue samples were downloaded from the TCGA database. A total of 537 and 91 patients from the TCGA and ICGC databases, respectively, were analyzed in the present study. The characteristics of ccRCC patients from TCGA and ICGC databases are shown in Table 1. We identified 57 upregulated and 16 downregulated hypoxia genes that were eligible for further screening via the “limma” package in R. The identified hypoxia-related DEGs are shown in heatmap and volcano plots in Figure 2A and B.

3.2 Enrichment analysis and construction of the PPI network

Next, we performed GO and KEGG enrichment analysis. The GO analysis results showed that hypoxia-related DEGs can be enriched in several basic biological processes (BP), including “response to hypoxia,” “response to decreased oxygen levels,” “response to oxygen levels” and “monosaccharide metabolic process” (Figure 3A). The KEGG pathway analysis results showed that the identified hypoxia-related DEGs were involved in the “HIF−1 signaling pathway,” “Glycolysis/Gluconeogenesis,” “Ras signaling pathway” and “AMPK signaling pathway” (Figure 3B). The PPI network, which included hypoxia-related DEGs, consisted of 65 nodes and 254 edges, as shown in Figure 3C.

3.4 Identification of prognostic hypoxia-related genes

Based on the 73 important hypoxia-related DEGs, we conducted univariate Cox regression analysis to investigate the prognostic value of these genes. The analysis revealed 16 candidate genes potentially associated with ccRCC prognosis (Figure 4A and B; P < 0.001). Next, we evaluated the association between these identified prognostic-associated candidate genes and patient survival time and clinical outcomes through multiple stepwise Cox regression. This analysis revealed six hub hypoxia genes (ANKZF1, ETS1, FBP1, PFKP, PLAUR and SERPINE1) that showed potential as independent predictors for ccRCC prognosis (Table 2).

3.5 Construction of hypoxia-based prognostic model

Next, we constructed a hypoxia-based model using the six identified hub genes. The following formula was calculated to generate the RiskScore for each patient:

To further evaluate the predictive potential of the proposed predictive model, we evaluated the survival outcome of patients with ccRCC according to their calculated RiskScore (Figure 5A). The results showed that patients with high RiskScores had poor OS, whereas patients with low RiskScores had longer OS. Furthermore, a time-dependent ROC curve analysis revealed that the AUC value, which predicted 3-year and 5-year OS, was 0.718 and 0.742, respectively (Figure 5B), suggesting that the proposed prognostic model had good ability to predict the long-term outcome of ccRCC patients. Figure 5C shows the expression heatmap, patient survival status and risk scores of the six hypoxia-related signatures in the low- and high-risk patient subgroups. These results indicate that the hypoxia-related model has a good prediction ability.

3.6 Independent analysis of the hypoxia-related model

To explore whether the RiskScore is independent of other clinical characteristics such as staging, we conducted an independence test for this model. The prognostic values of different clinical characteristics of ccRCC patients were also evaluated by univariate Cox regression analysis, which revealed that age, cancer grade, stage, TNM and RiskScore were related to OS (Figure 6A, P < 0.05). Moreover, multiple regression analysis showed that prognostic factors such as age and RiskScore could independently predict OS (Figure 6B, P < 0.001). Overall, the RiskScore was a robust predictor that was independent of the staging and grading of the patients.

3.7 Verification of correlation between the expression of the six hypoxia-related genes and survival time as well as clinical characteristics of ccRCC patients

We investigated the correlation between the six hypoxia-related genes and clinical characteristics, and the results revealed that these genes were almost related to the patient’s staging and grading (Figure 7A–E). We also evaluated the correlation between the expression of the six hypoxia-related genes and survival time via GEPIA online tools, which revealed that the expression levels of ANKZF1, ETS1, FBP1, PFKP and PLAUR had a positive or negative correlation with the ccRCC patient’s OS (Figure 8A–E). The above results clearly indicated that the expression of these six genes had a robust correlation with the prognosis of ccRCC patients. In particular, using data from the HPA database, we could observe that the expressions of ANKZF1, FBP1, PLAUR and PFKP were significantly reduced in ccRCC samples, whereas normal tissues showed the opposite trend (Figure 11A and C–E). No significant differences in ETS1 nor SERPINE1 protein levels were observed between tumor and normal renal tissues (Figure 11B and F, both were negative). The above results indicate that the six hypoxia-related genes we identified were suitable for predicting the prognosis of ccRCC patients.

3.8 Validation of the nomogram

We also designed a nomogram for ccRCC prognosis based on the signatures of the six hypoxia-related genes (Figure 10A). The point of each variable could be found in a horizontal line, which was then summed and normalized into a distribution of 0–100 for each patient. The nomogram-predicted probability of 3- and 5-year OS for ccRCC patients could be obtained after calculating the total nomogram score. The C-index of the new model and old model was 0.773 (95% CI: 0.724–0.822) and 0.758 (95% CI: 0.705–0.811), respectively. IDI results showed that the new model was 3.7% better than the old model (P < 0.01). Calibration curve, DCA and NRI plots also revealed that the new model was better than the old model, as shown in Figure 10B–E and Figure 9A–D. After we concluded that the new model was better than the old model, we conducted external verification of the new model. The 3-year and 5-year calibration curves of the ICGC dataset are shown in Figure 10G and E. The indication of the calibration curve matches well. The C-index of the ICGC dataset was 0.710 (95% CI: 0.592–0.828). The results also showed that the new model has a better ability to predict the survival rate of ccRCC patients.

A novel and robust prognostic signature for patients with ccRCC has been in high demand. Different from previous prognostic systems, either comprehensive or simply based on clinicopathologic characteristics and the current TNM staging method, ideas about the prognostic implications of hypoxia-related gene signatures are the focus of this study.

Since the 1990s, hypoxia-induced pathways have received great attention from researchers. Rapidly-proliferating tumor cells, incremental oxygen consumption and structurally and functionally abnormal vasculature inside tumor tissue co-formed a unique hypoxia micro-environment of solid tumors, distinguishing cancer from normal tissue and also partly contributing to a poor clinical outcome (16). Interestingly, the hypoxia signaling pathway can sometimes also be activated by genetic alterations in cancers; for example, the disfunction of VHL can be observed in ccRCC (17).

Increasing evidence demonstrates that there are direct or indirect interactions between the hypoxia signaling pathway and tumorigenesis or tumor progression. However, adaptive responses triggered by hypoxia containing hundreds of related proteins and genes can be both diverse and specific. Besides, intersecting with other signaling mechanisms also makes the hypoxia signaling pathway intricate and obscure. It is impractical to figure out all the steps involved in the hypoxia signaling pathway relying on present theory. Accordingly, we focused our attention to the hypoxia-related genes of the hypoxia signaling pathway.

Hypoxia-induced up-transcription of genes has shown a demonstrable influence on a series of tumor biological processes, including cell uncontrollable proliferation and immortalization (18), angiogenesis (18), glucose metabolism (19), immune escape (20), tumor invasion and metastasis (21) and radiation resistance (22). This agrees with our assumption that the differential expression of hypoxia activating genes is a considerable risk feature for patient survival.

Here, we tried to investigate the correlation between the survival of patients with ccRCC and hypoxia-related genes and subsequently identify genes significantly contributing to clinical outcome. After a series of biological and statistical analyses, ANKZF1, ETS1, PLAUR, SERPINE1, FBP1 and PFKP were selected, and we hypothesized that they could be used together to generate an integrated hypoxia-related RiskScore for optimizing the previous prediction model.

Subsequently, a series of analyses including ROC curve, risk plot and survival analysis were performed to evaluate RiskScore, resultantly suggesting that RiskScore had ideal overall diagnosis and prognosis performance. In addition, results of univariate Cox regression analysis and multivariate Cox proportional hazards regression analyses revealed that RiskScore based on hypoxia-related genetic signatures was an independent risk factor beyond TNM staging and grading. A multiple risk factors ROC curve also suggested that RiskScore was an excellent independent prognosis predictor for the five-year survival rate of patients compared with other risk factors such as TNM staging, grading and age. A novel prognosis nomogram optimized via RiskScore, a new independent prognosis predictor, is feasible.

Notably, some genetic signatures chosen for RiskScore calculation seemed to be expressed more in high staging tumors, whereas others were expressed more in low staging tumors. Whether negative or positive, certain correlations between the gene signatures and the tumor clinical grades and TNM stages could be observed in the present study, indicating that those prognostic hypoxia-related gene signatures play anti-tumor or pro-tumor roles in the progression of ccRCC.

Differential expression of the abovementioned six gene signatures and their impacts in cancer have been reported previously. Over-expression of ETS1 (E26 transformation specific-1), which belongs to the large family of transcription factors with ETS domain, was identified in a variety of solid tumors, including breast cancer, lung cancer and, specifically, renal carcinoma (23). Further research provided evidence that ETS1 played a part through regulating the expression of HIF-related genes, consistent with the results of GO and KEGG enrichment analyses in our research (24). ETS1 was considered as an oncogene that was involved in many biological processes, such as angiogenesis, drug resistance and tumorigenesis, and thus was linked to unfavorable survival (23). However, our research came to the conflicting conclusion that the elevated expression of ETS1 was a protective feature for prognosis. Similar experimental conclusions have also been reported for colon cancer cells and invasive breast cancer (25, 26). This may be related to the heterogeneity of the tumor. The apoptosis-inducing activity of ETS1 may also account for that. Pro-apoptotic genes, including CDKN1A (encoding p21 protein), CDKN1B (encoding p27 protein) and caspase I (27), and tumor suppressor protein genes, including p16INK4A (28), were validated as targets of ETS1 (29).

Plasminogen activator urokinase receptor (PLAUR) was another gene among those identified as having prognostic features and over-expressing in a number of tumors. However, we inferred that what really caused poor survival was not PLAUR but its activator, urokinase type plasminogen activator (uPA). A series of integrins and proteases were identified as having the capacity to enable tumor tissues to penetrate the basement membrane (BM) and extracellular matrix (ECM), surround them and metastasize to other anatomical sites and form secondary tumor foci [18]. One of those proteases mediating invasion and metastasis-related molecular processes was uPA, making disseminated tumors hard to eradicate successfully and causing serious clinical problems (30, 31). Serpin Family E Member 1 (SERPINE 1), also known as plasminogen activator inhibitor type 1 (PAI-1), is the main inhibitor of the uPA/uPAR complex, regulating the adhesion balance of ECM in both tumor and normal tissue. Similar to PLAUR, SERPINE 1 has been validated to play a critical role in a variety of tumors (32, 33). Fructose-bisphosphatase1 (FBP1), a rate-limiting enzyme involved in gluconeogenesis, and phosphofructokinase-platelet (PFKP), an isoform of phosphofructokinase-1(PFK-1), are both involved in glucose metabolism regulation and were reported to be associated with the Warburg effect, which describes a trend to generate energy mainly via glycolysis in tumors, initially proposed by Otto Warburg (34–36). Last but not least, ANKZF1 is a cofactor that binds to p97 and regulates its cellular biological functions, such as protein quality control, apoptosis, autophagy, DNA damage repair and transcriptional activation (37). Therefore, the abovementioned six gene signatures have the potential to be reliable biomarkers for prognosis.

Necessarily, we compared the optimized prognosis nomogram and the previous one with several evaluations, including C-index, calibration curves, NRI, IDI and, especially, decision curve analysis (DCA), which was more practical than the typical ROC curve analysis since it took the consequences of clinical strategies into consideration (38). The optimized nomogram was clearly the superior one, with better accuracy and discrimination. Unsurprisingly, subsequent external validation in the ICGC database showed the promising performance of the optimized nomogram. We believe that the findings of our research could provide robust prognostic indicators and underlying therapeutic targets for patients with ccRCC and, more importantly, provide insight into the correlation between hypoxia-related gene expression and clinical outcomes. We believe that patients may benefit from our study.

Although we made several accomplishments in the investigation of novel prognostic biomarkers, there were several limitations in this study. For example, those components participating in the hypoxia signaling pathway and influencing the regulation of hypoxia-inducible genes, whether directly or indirectly, such as pVHL and PHD, were not involved in the construction of the prognostic system, meaning their impact and underlying biological process information were ignored. Further studies on the hypoxia-related signaling pathway and ccRCC are ongoing, and well-designed and prospective studies for enriching relevant theories are warranted.

Our study reveals that the differential expression of hypoxia-related genes is associated with the overall survival of patients with ccRCC. And the novel nomogram optimized via the hypoxia-related genes-based RiskScore exhibited promising prognostic ability. It may be able to serve as a prognostic tool for guiding clinical decisions and developing effective individualized treatment.

ccRCC: clear cell renal cell carcinoma; DEGs: Differentially expressed genes; TCGA : The Atlas Cancer Genome; OS: Overall survival; GO: Gene Ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes; PPI: Protein–protein interaction; ICGC: International Cancer Genome Consortium; BP: Biological process; CC: Cell components; MF: Molecular functions; DCA: decision curve analysis; IDI: integrated discrimination improvement; NRI: net reclassification index; HPA: the Human Protein Atlas; GEPIA: the Gene Expression Profiling Interactive Analysis.

AVAILABILITY OF DATA AND MATERIALS

The datasets used and/or analyzed in the current study are available from the corresponding author upon reasonable request.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

Not applicable.

CONSENT FOR PUBLICATION

Not applicable.

CONFLICTS OF INTEREST

The authors declare that they have no competing interests.

FUNDING

This work was financially supported by the Natural Science Foundation of China (Grant No. 81660372) and the Natural Science Foundation of Guangxi (Grant No. AB19110030; Grant No. 2017GXNSFAA198159).

AUTHOR CONTRIBUTIONS

Gang Du designed the experiments. Zhengtian Li, Chan Li, Jun Huang, Zhenpei Wen and Wenkang Yang performed research. Zhengtian Li and Lingling Jiang analyzed the data. Zhengtian Li and Rong Zhao and Wenkang Yang wrote the manuscript. All of the authors reviewed the article.

ACKNOWLEDGMENTS

The results of this study are based on data from TCGA (https://www.cancer.gov/tcga) and ICGC (https://icgc.org/). We thank the authors who provided the data for this study. We thank LetPub (www.letpub.com) for its linguistic assistance during the preparation of this manuscript.

Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin. 2020;70(1):7–30.
Znaor A, Lortet-Tieulent J, Laversanne M, Jemal A, Bray F. International variations and trends in renal cell carcinoma incidence and mortality. European urology. 2015;67(3):519–30.
Weikert S, Boeing H, Pischon T, Weikert C, Olsen A, Tjonneland A, et al. Blood pressure and risk of renal cell carcinoma in the European prospective investigation into cancer and nutrition. Am J Epidemiol. 2008;167(4):438–46.
Renehan AG, Tyson M, Egger M, Heller RF, Zwahlen M. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet. 2008;371(9612):569–78.
Jonasch E, Gao J, Rathmell WK. Renal cell carcinoma. BMJ. 2014;349:g4797.
Choueiri TK, Motzer RJ. Systemic Therapy for Metastatic Renal-Cell Carcinoma. N Engl J Med. 2017;376(4):354–66.
Thorstenson A, Bergman M, Scherman-Plogell AH, Hosseinnia S, Ljungberg B, Adolfsson J, et al. Tumour characteristics and surgical treatment of renal cell carcinoma in Sweden 2005–2010: a population-based study from the national Swedish kidney cancer register. Scandinavian journal of urology. 2014;48(3):231–8.
D'Ignazio L, Batie M, Rocha S. Hypoxia and Inflammation in Cancer, Focus on HIF and NF-κB. Biomedicines. 2017;5(2).
Gilkes DM, Semenza GL, Wirtz D. Hypoxia and the extracellular matrix: drivers of tumour metastasis. Nature reviews Cancer. 2014;14(6):430–9.
Rankin EB, Giaccia AJ. Hypoxic control of metastasis. Science. 2016;352(6282):175–80.
Roy S, Kumaravel S, Sharma A, Duran CL, Bayless KJ, Chakraborty S. Hypoxic tumor microenvironment: Implications for cancer therapy. Experimental biology and medicine (Maywood, NJ). 2020:1535370220934038.
Choudhry H, Harris AL. Advances in Hypoxia-Inducible Factor Biology. Cell Metabol. 2018;27(2):281–98.
Yan B, Ma QF, Tan WF, Cai HN, Li YL, Zhou ZG, et al. Expression of HIF-1α is a predictive marker of the efficacy of neoadjuvant chemotherapy for locally advanced cervical cancer. Oncol Lett. 2020;20(1):841–9.
Bacigalupa ZA, Rathmell WK. Beyond glycolysis: Hypoxia signaling as a master regulator of alternative metabolic pathways and the implications in clear cell renal cell carcinoma. Cancer letters. 2020;489:19–28.
Liu Y, Wu J, Huang W, Weng S, Wang B, Chen Y, et al. Development and validation of a hypoxia-immune-based microenvironment gene signature for risk stratification in gastric cancer. Journal of translational medicine. 2020;18(1):201.
Semenza GL. Hypoxia-inducible factors in physiology and medicine. Cell. 2012;148(3):399–408.
Majmundar AJ, Wong WJ, Simon MC. Hypoxia-inducible factors and the response to hypoxic stress. Molecular cell. 2010;40(2):294–309.
Mak P, Leav I, Pursell B, Bae D, Yang X, Taglienti CA, et al. ERbeta impedes prostate cancer EMT by destabilizing HIF-1alpha and inhibiting VEGF-mediated snail nuclear localization: implications for Gleason grading. Cancer Cell. 2010;17(4):319–32.
Luo W, Hu H, Chang R, Zhong J, Knabel M, O'Meally R, et al. Pyruvate kinase M2 is a PHD3-stimulated coactivator for hypoxia-inducible factor 1. Cell. 2011;145(5):732–44.
Lukashev D, Ohta A, Sitkovsky M. Hypoxia-dependent anti-inflammatory pathways in protection of cancerous tissues. Cancer Metastasis Rev. 2007;26(2):273–9.
Chan DA, Giaccia AJ. Hypoxia, gene expression, and metastasis. Cancer Metastasis Rev. 2007;26(2):333–9.
Moeller BJ, Richardson RA, Dewhirst MW. Hypoxia and radiotherapy: opportunities for improved outcomes in cancer treatment. Cancer Metastasis Rev. 2007;26(2):241–8.
Dittmer J. The role of the transcription factor Ets1 in carcinoma. Sem Cancer Biol. 2015;35:20–38.
Salnikow K, Aprelikova O, Ivanov S, Tackett S, Kaczmarek M, Karaczyn A, et al. Regulation of hypoxia-inducible genes by ETS1 transcription factor. Carcinogenesis. 2008;29(8):1493–9.
Li R, Pei H, Papas T. The p42 variant of ETS1 protein rescues defective Fas-induced apoptosis in colon carcinoma cells. Proc Natl Acad Sci USA. 1999;96(7):3876–81.
Ballschmieter P, Braig M, Lindemann RK, Nordheim A, Dittmer J. Splicing variant DeltaVII-Ets1 is downregulated in invasive Ets1-expressing breast cancer cells. Int J Oncol. 2003;22(4):849–53.
Pei H, Li C, Adereth Y, Hsu T, Watson DK, Li R. Caspase-1 is a direct target gene of ETS1 and plays a role in ETS1-induced apoptosis. Cancer research. 2005;65(16):7205–13.
Ohtani N, Zebedee Z, Huot TJ, Stinson JA, Sugimoto M, Ohashi Y, et al. Opposing effects of Ets and Id proteins on p16INK4a expression during cellular senescence. Nature. 2001;409(6823):1067–70.
Teruyama K, Abe M, Nakano T, Iwasaka-Yagi C, Takahashi S, Yamada S, et al. Role of transcription factor Ets-1 in the apoptosis of human vascular endothelial cells. J Cell Physiol. 2001;188(2):243–52.
Brooks SA, Lomax-Browne HJ, Carter TM, Kinch CE, Hall DM. Molecular interactions in cancer cell metastasis. Acta histochemica. 2010;112(1):3–25.
Santibanez JF. Urokinase Type Plasminogen Activator and the Molecular Mechanisms of its Regulation in Cancer. Protein Pept Lett. 2017;24(10):936–46.
Seker F, Cingoz A, Sur-Erdem İ, Erguder N, Erkent A, Uyulur F, et al. Identification of SERPINE1 as a Regulator of Glioblastoma Cell Dispersal with Transcriptome Profiling. Cancers (Basel). 2019;11(11).
Pavón MA, Arroyo-Solera I, Céspedes MV, Casanova I, León X, Mangues R. uPA/uPAR and SERPINE1 in head and neck cancer: role in tumor resistance, metastasis, prognosis and therapy. Oncotarget. 2016;7(35):57351–66.
Jin X, Pan Y, Wang L, Zhang L, Ravichandran R, Potts PR, et al. MAGE-TRIM28 complex promotes the Warburg effect and hepatocellular carcinoma progression by targeting FBP1 for degradation. Oncogenesis. 2017;6(4):e312.
Lang L, Chemmalakuzhy R, Shay C, Teng Y. PFKP Signaling at a Glance: An Emerging Mediator of Cancer Cell Metabolism. Adv Exp Med Biol. 2019;1134:243–58.
Kim NH, Cha YH, Lee J, Lee SH, Yang JH, Yun JS, et al. Snail reprograms glucose metabolism by repressing phosphofructokinase PFKP allowing cancer cell survival under metabolic stress. Nature communications. 2017;8:14374.
Stapf C, Cartwright E, Bycroft M, Hofmann K, Buchberger A. The general definition of the p97/valosin-containing protein (VCP)-interacting motif (VIM) delineates a new family of p97 cofactors. J Biol Chem. 2011;286(44):38670–8.
Fitzgerald M, Saville BR, Lewis RJ. Decision curve analysis. Jama. 2015;313(4):409–10.

Table 1

Characteristics of patients from TCGA and ICGC databases.
Characteristic	TCGA (%)	ICGC (%)
AJCC stage
I	269(50.1)	48(52.7)
II	57(10.6)	12(13.2)
III	125(23.3)	13(14.3)
IV	83(15.5)	9(9.9)
Not available	3(0.6)	9(9.9)
Grade
I	14(2.6)	13(14.3)
II	230(42.8)	48(52.7)
III	207(38.5)	15(16.5)
IV	78(14.5)	14(15.4)
Not available	8(1.5)	1(1.1)
Sex
Male	346(64.4)	52(57.1)
Female	191(35.6)	39(42.9)
Age (mean ± standard deviation)	60.59 ± 12.14	60.47 ± 10.03
Total number of patients	537	91
AJCC, American Joint Committee on Cancer; ICGC, International Cancer Genome Consortium; TCGA, The Cancer Genome Atlas.

Table 2

Prognostic-related genes via the multivariate Cox regression analysis
Gene symbol	Coefficient	HR	P-value
ANKZF1	0.054262	1.055762	4.24E-05
ETS1	-0.00974	0.990305	0.009003
FBP1	-0.01052	0.989535	0.016046
PFKP	-0.00426	0.995746	0.012756
PLAUR	0.048648	1.04985	1.41E-07
SERPINE1	0.000519	1.000519	0.048343

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification And Validation Of A Hypoxia ‑ Related Prognostic Signature In Clear Cell Renal Cell Carcinoma Patients

Status:

Version 1

Abstract

Figures

Background

Methods

Results

Discussion

Conclusions

Abbreviations

Declarations

References

Tables

Status:

Version 1