Comprehensive analysis to identify long non-coding RNA HCP5/MICA axis as a prognostic biomarker in glioblastoma

doi:10.21203/rs.3.rs-1989629/v1

Download PDF

Research Article

Comprehensive analysis to identify long non-coding RNA HCP5/MICA axis as a prognostic biomarker in glioblastoma

https://doi.org/10.21203/rs.3.rs-1989629/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose

To explore the lncRNA-mRNA network based on the IDH1 mutation status and construct a prognostic model for GBM.

Methods

The data of expression, somatic mutation, methylation, immune microenvironment were obtained from TCGA database, CGGA database and GTEx database. The common differential expressed genes (DEGs) between IDH1 mutant GBM and IDH1 wildtype GBM were used to perform a series of bioinformatics analysis to identify the hub gene. LASSO cox regression was used to screen a risk signature and nomogram was used to construct a prognostic model for GBM.

Results

LncRNA HCP5/MICA axis was downregulated in IDH1 mutant GBM and associated with the overall survival (OS) of GBM patients. Further analysis revealed that MICA expression was positively correlated with immune infiltration as well as the expression of several immune checkpoint gene in GBM. We screened a risk signature using MICA-associated genes by LASSO cox regression and the risk score performed well in predicting the OS. Furthermore, the risk score was used to construct a nomogram prognostic model combined with age, gender, IDH mutation, TP53 mutation, radiation and chemotherapy to predict the OS at 1-year, 2-year and 5-year.

Conclusions

This study constructed a prognostic model for GBM with MICA and clinical data for the first time and provided an insight into the molecular therapy of GBM. Further studies are required to uncover the mechanism of regulation between HCP5 and MICA in GBM.

Survival analysis

Glioblastoma

MICA

Prognostic biomarker

Bioinformatics analysis

Glioblastoma (GBM) is a highly aggressive brain tumor with a poor prognosis and causes a tremendous burden to the patients and medical system. The current standard treatment for newly diagnosed GBM relies on the gross total surgical resection and followed by the radio-chemotherapy and adjuvant therapy with temozolomide [1]. Despite the rapid progress of antitumor drugs and novel treatment methods, the median survival of patients with GBM still remains around 15 months [2]. The molecular heterogeneity and pathological heterogeneity of GBM result in a large difference in the survival of patients with GBM. Hence, it’s necessary to explore new prognostic biomarkers for GBM.

Isocitrate dehydrogenase 1 (IDH1) is one of the three isoforms of human IDH, which catalyzes the interconversion of isocitrate and 2-oxoglutarate (2-OG) and reduces the nicotinamide adenine dinucleotide phosphate (NADP⁺) to NADPH [3, 4]. The roles of IDH1 in cancers have been well-elucidated, especially in glioma [5]. IDH1 was mutated in approximately 10% of GBM and most of them are secondary GBM. However, IDH1 was found to be mutated in most of gliomas of WHO grade II or III and probably drive a low-grade glioma to a GBM [6, 7]. Large data suggested that IDH1 mutation was associated with the overall survival of patients with GBM and it has been taken as a prognostic marker for patients with glioma. Considering the crucial roles of IDH1 in gliomas, the 2021 WHO classification of tumors of the central nervous system made a lot of changes compared to the fourth edition [8]. Adult-type diffuse gliomas include only three types of tumor (Astrocytoma, IDH-mutant; Oligodendroglioma, IDH-mutant and 1p/19q-codeleted; and Glioblastoma, IDH-wildtype), and all IDH mutant diffuse astrocytic tumors are considered astrocytoma, IDH-mutant [8].

Although the significance of IDH1 mutation in GBM has been widely accepted, the IDH1 mutation-related genomic alteration remains obscure. In this study, we screened a long non-coding RNA (lncRNA) MICA as a potential biomarker for GBM by analyzing the genomic differences between different IDH1 mutation status and then constructed a prognostic model based on the risk score of MICA-associated genes and the clinicopathological features of patients with GBM.

Data source and processing

The gene somatic mutation data, gene-level RNA-sequencing (RNA-seq) data and corresponding clinical information of GBM patients were obtained from The Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA). The DNA methylation data of GBM patients and the normalized pan-cancer data were downloaded from UCSC Xena (https://xenabrowser.net/), including 60499 tumor samples and 19131 normal tissue samples. The immune scores of immune cells in GBM patients were downloaded from the TIMER database (http://cistrome.dfci.harvard.edu/TIMER/). Transcripts per million (TPM) values or log2 transformations were performed with the RNA-seq data. The somatic mutational phenotypes in GBM were visualized with a waterfall plot.

Identification of Differentially Expressed Genes

All GBM samples were classified into IDH1 mutant GBM (IDH^MUT) and IDH1 wildtype GBM (IDH^WT). The R package ‘limma’ was used to identify the differentially expressed genes (DEGs) between groups, and the thresholds were defined as |log2 (fold change)| > 1 and p value < 0.05. The DEGs were displayed in volcano plots and the common DEGs in TCGA and CGGA were shown in the Venn diagram. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed in the common downregulated DEGs and upregulated DEGs.

Determination of the hub genes

To identify the long non-coding RNA (lncRNA)-mRNA axis in GBM, the lncRNA genes in DEGs were used to matching the potential targeted genes with starBase v2.0 [9]. The common genes in DEGs and the potential targeted genes of lncRNA were defined as the candidate genes. Survival analysis were performed between high expression group and low expression group of sixteen candidate genes in the GBM cohort in TCGA and CGGA. The genes whose p values of survival analysis less than 0.05 in both cohorts were defined as the hub genes. The hub genes and the corresponding lncRNAs were collected in the following analysis.

Expression analyses of HCP5 and MICA in GBM

The expression distribution of HLA complex P5 (HCP5) and MHC class I polypeptide-related sequence A (MICA) in IDH^MUT and IDH^WT, the expression correlation between MICA and HCP5 were drawn by the R package 'ggplot'. The subcellular localization of HCP5 and MICA were predicted by the lncLocator[10]. The pan-cancer expression of HCP5 and MICA in 33 types of cancer and normal tissue samples were analyzed by the Kruskal Wallis test and visualized with violin plots. The interactions between HCP5 and MICA were predicted by starBase v2.0[9].

Correlation analysis between MICA expression and immune infiltration

Tumor infiltrating lymphocytes and immune microenvironment are independent predictors of overall survival in cancer[11]. The correlations between MICA expression and TIMER database score of six types of immune cells (B cell, CD4 + T cell, CD8 + T cell, neutrophil, macrophage and dendritic) in GBM were analyzed using Spearman' s correlation test. The immune score, estimate immune score and stromal score in GBM samples were obtained by 'estimate' package in R and their correlations with MICA expression were also analyzed using Spearman's correlation test.

Correlation analysis between MICA expression and immune checkpoint.

The expression data of more than 40 immune checkpoint genes were extracted from 33 tumor types, and their correlation with MICA expression were analyzed separately. The Pearson’s rho and p values were visualized in a heatmap.

Methylation analysis of MICA in GBM

DNA methylation is one of the epigenetic modifications, which plays important roles in the regulation of gene expression and oncogenesis. In UCSC Xena, the GDC TCGA GBM program was used to investigate the methylation of MICA. 151 patients were selected by three variables (Illumina Human Methylation 450, OS, and OS.time). Survival analysis was performed between groups of high methylation and low methylation of each methylation site, where significance was defined as p < 0.05. Three methylation sites were obtained from the survival analysis and the association between their beta values and MICA expression were investigated by Spearman's correlation test.

Construction of gene signature with MICA-associated genes

String database (https://string-db.org) is a tool to search known and predicted protein-protein interactions. The MICA-associated genes and their interactions were found in String database. The expression data of MICA-associated genes in TCGA were used to perform LASSO cox regression by the 'glmnet' package in R. Then, six genes and their coefficients were determined by the best penalty parameter λ. The risk score of each patient was calculated using the formula:

$$\text{R}\text{i}\text{s}\text{k} \text{s}\text{c}\text{o}\text{r}\text{e}={\sum }_{i=1}^{n=6}c{oef}_{i}\times e{xp}_{i},$$

where coef_i is the coefficient and exp_i is the expression. Each patient in both TCGA (training dataset) and CGGA (validation dataset) obtained a risk score from the formula according to the expression data of six genes. In order to test the predictive ability of the gene signature, survival analyses were performed in TCGA and CGGA, where patients were classified into high risk score group and low risk score group.

Construction of the predictive nomogram

In order to establish a prognostic model for individual patient, we combined risk score and several clinicopathological features (age, gender, IDH1 mutation, TP53 mutation, radiation, chemotherapy) to construct a predictive nomogram. Age and risk score are continuous variables while the others are categorical variables. The nomogram was used to predict the survival probability at 1-year, 2-year and 5-year. The C-index was used to evaluate the prognostic accuracy of the nomogram.

Somatic mutational phenotypes in glioblastoma and IDH1 mutation frequencies in pan-cancer.

Somatic mutational data were used to descript the genomic mutational landscape of 20 genes with the highest mutation frequency in GBM. Missense mutation is the most common alteration. There are 6.6% of GBM patients with IDH1 mutation. The IDH1 mutation frequency in pan-cancer revealed a highest mutation frequency in diffuse glioma and a fourth mutation frequency in GBM. It has been demonstrated that IDH1 mutation is crucial for the diagnosis and prognosis of GBM. Here, we performed the survival analysis to validate the predictive function of IDH1 mutation in TCGA and CGGA. The Kaplan-Meier curves showed that IDH1^MUT had a higher survival probability than IDH^WT in both TCGA and CGGA, log-rank p value < 0.05 was considered as significant (Fig. 1).

Common Differentially Expressed Genes in TCGA and CGGA

Due to the significant roles of IDH1 mutation in GBM, this study aimed to investigate the molecular mechanism differences between IDH1^MUT and IDH1^WT GBM and managed to construct a robust prognostic model for patients with GBM. First, the ‘limma’ package in R was used to identify the DEGs in IDH1^MUT compared to IDH^WT. There were 3106 DEGs in TCGA and 2298 DEGs in CGGA. The blue circles and red circles in volcano plots represent the down-regulated DEGs and up-regulated DEGs (Fig. 2A-B). The Venn diagram showed 1049 common DEGs between two datasets, among which 838 DEGs were down-regulated, 208 DEGs were up-regulated and 3 DEGs were reversely regulated in two datasets (Fig. 2C).

GO and KEGG pathway analysis of common DEGs

838 common down-regulated DEGs and 208 common up-regulated DEGs were used to perform the GO and KEGG pathway analysis. The top5 enrichment GO terms in common down-regulated DEGs were cytokine-cytokine receptor interaction, PI3K-Akt signaling pathway, focal adhesion, pathways in cancers and proteoglycans in cancer. The top5 enrichment KEGG pathways in common down-regulated DEGs were signal transduction, cell adhesion, inflammatory response, immune response and positive regulation of cell proliferation. The top5 enrichment GO terms in common up-regulated DEGs were neuroactive ligand-receptor interaction, cAMP signaling pathway, glutamatergic pathway, basal cell carcinoma and Steroid hormone biosynthesis. The top5 enrichment KEGG pathways in common up-regulated DEGs were negative regulation of transcription from RNA polymerase II promoter, cell adhesion, nervous system development, positive regulation of transcription (DNA templated) and chemical synaptic transmission (Fig. 2D-E). All these pathways are associated with the biological process of cancer.

Determination of the hub genes

In order to further investigate the molecular mechanism differences between IDH1^MUT and IDH1^WT GBM, the starBase v2.0 was used to identify the target genes of the lncRNAs in common DEGs. As a result, twenty-nine lncRNAs in DEGs targeted 697 unique genes. There were 16 common genes in 697 target genes and 1049 common DEGs. The 16 common genes were marked as candidate genes in the following analysis. Survival analysis were performed between high expression group and low expression group of 16 candidate genes in the GBM cohort in TCGA and CGGA. MICA was the only one whose p value less than 0.05 in both datasets (Fig. 3). In this study, MICA was the potential target of lncRNA HCP5. Therefore, we speculated that HCP5/MICA axis may play a significant role in GBM.

Expression of HCP5 and MICA

The expression levels of HCP5 and MICA were analyzed in 33 types of tumor and the corresponding normal tissue samples. The expression data of tumor samples were obtained from TCGA, the majority of normal tissue were obtained from GTEx due to the few normal samples in TCGA. The violin plots showed that the expression of HCP5 is higher in most tumor types than that in normal tissue and that MICA is differentially expressed in different types of tumor and normal tissue, but they are both higher in GBM than that in normal brain tissues (Fig. 4A-B). Then we analyzed the expression of HCP5 and MICA in IDH ^MUT and IDH1^WT GBM, showing a significant higher expression of both genes in IDH ^WT GBM than that in IDH ^MUT GBM (Fig. 4C-D). Considering that the subcellular localization of lncRNA implies the underlying mechanism, we utilized the lncLocator[10] to predict the subcellular localization of HCP5 and MICA. As shown in Fig. 4G, HCP5 was mainly located in cytoplasm and MICA was mainly located in cytoplasm and cytosol, which suggested that HCP5 may improve the expression of MICA (Fig. 4E). Analysis of expression correlation of HCP5 and MICA in GBM revealed that the expression of MICA was strongly correlated with the expression of HCP5 (Fig. 4F-G). Then we analyzed the interactions between HCP5 and MICA using starBase v2.0. There are 9 interaction regions in HCP5 and MICA and three interactions with lowest free energy are shown in Fig. 4H. Their interactions were validated in 3 three supported experiments[12, 13]. These data indicated that HCP5 can improve the expression of MICA in GBM.

Correlation analysis between MICA expression and immune infiltration and immune checkpoints.

Immune infiltration analysis revealed the correlation between MICA expression and scores of six types of immune cell in GBM. MICA expression was significantly correlated with the immune infiltration of macrophage (R = 0.174, p = 0.0327) and dendritic (R = 0.25, p = 0.00195) in GBM (Fig. 5A). In order to explore the roles of MICA in GBM immune microenvironment, the correlation between MICA expression and immune score, estimate immune score and stromal score were analyzed. As shown in Fig. 5B-D, MICA expression was positively correlated with immune score (R = 0.278, p < 0.001), estimate immune score (R = 0.291, p < 0.001) and stromal score (R = 0.303, p < 0.001) in GBM. Furthermore, we analyzed the correlation between MICA expression and the expression levels of 47 immune checkpoint genes in 33 types of tumor. MICA expression was significantly correlated with expression levels of immune check point genes in pan-cancer, especially in LGG (low grade glioma), OV and LIHC. In GBM, there were 21 immune check point genes correlated with the MICA expression (Fig. 5E).

Methylation analysis of MICA in GBM

To further unravel the roles of MICA in GBM, we also explored the correlation of methylation sites of MICA and the survival of GBM patients. The heatmap derived from UCSC Xena displayed the survival days, survival status and 74 methylation sites of MICA in 151 patients with GBM (Fig. 6A). For each methylation site, 151 patients with GBM were divided into high methylation and low methylation groups according to their beta values. Survival analysis between high methylation and low methylation groups determined three methylation sites (cg16203549, cg18188079, cg09210225) that were associated with the OS of GBM. High methylation of cg16203549 and low methylation of cg18188079 and cg09210225 are the favorable prognostic factors in GBM (Fig. 6B-D). Then we analyzed the correlation between MICA expression and the beta value of methylation. As shown in Fig. 6E-G, MICA expression was negatively correlated with the methylation of cg16203549 (R=-0.28, p = 0.025), and is positively correlated with the methylation of cg18188079 (R = 0.43, p < 0.01) and cg09210225 (R = 0.41, p < 0.01). These data suggested that the methylation of cg16203549, cg18188079 and cg09210225 regulated the MICA expression and eventually influence the survival time of GBM patients.

The MICA-associated gene signature in GBM

Ten MICA-associated genes were determined by String database and five protein-protein associations existed in MICA and the proteins encoded by its associated genes (Fig. 7A). Then the expression data of these 11 genes in GBM in TCGA were used to perform the LASSO cox regression. Six genes with non-zero coefficients were determined by the best penalty parameter λ. Four genes (MICA, PDIA6, ULBP2, KLRD1) had positive coefficients, whereas two genes (KLRK1, MICB) had negative coefficients (Fig. 7B-D). We calculated the risk score for each patient according to the formula mentioned in the method. Survival analysis revealed that low risk score group had a superior OS than the high risk score group in GBM, and it was validated in the CGGA dataset (Fig. 7E-F). Therefore, the MICA-associated gene signature consisting of six genes may be a potential prognosis biomarker for GBM.

Prognostic model for GBM

The development and progress of GBM involves intricate mechanism and the prognosis of patients with GBM can be affected by various factors. Here, we subjected the risk score, age, gender, IDH1 mutation, TP53 mutation, radiation and chemotherapy into the nomogram to construct a prognostic model for individual GBM patient. The point for each variable ranges from 0 to 10. Risk score is the variable taking the largest share in the total points. The favorable prognostic factors for variables are younger age, female, IDH1 mutant, TP53 mutant, receiving temozolomide chemoradiation and temozolomide chemotherapy, low risk score. Given a specific total point of a patient with GBM, the survival probabilities at 1-year, 2-year and 5-year can be predicted from the nomogram (Fig. 8). The C-index of the nomogram is 0.76.

Glioblastoma is the most common primary malignant tumor in central nervous system with an annual incidence of 3–5 per 100, 000 people[14]. Most of patients with GBM present without any certain risk factors. The only well-validated non-genetic risk factors for GBM are ionizing radiation (increasing risk) and history of allergies (decreasing risk) [15]. The genetic risk factors for GBM includes a series of gene mutations, one of which is the mutation of IDH1 [15, 3, 16, 17]. In the 2016 WHO classification of CNS tumors and 2021 WHO classification of CNS tumors, IDH-mutation has become the key biomarker for the classification of gliomas [8, 18]. The importance of IDH mutation in GBM has been widely accepted.

Studies showed that lncRNAs played significant roles in the onset and progression of glioma, such as stemness, proliferation, migration, angiogenesis and drug resistance [19, 20]. In this study, we screened the DEGs by RNAseq data and lncRNA-targeted hub genes by starbase. Finally, we identified the lncRNA HCP5 and protein coding gene MICA axis as a potential prognostic biomarker in GBM. The expression of HCP5 and MICA were both upregulated in GBM than normal tissue. However, the expression of HCP5 and MICA were both downregulated in IDH1^MUT GBM than IDH1^WT GBM. Patients with IDH1^MUT GBM generally showed substantially longer overall survival than those with IDH1^WT GBM [21]. Interestingly, HCP5 and MICA shared many interaction regions and their expression presented a strong correlation, which suggested a regulatory relationship between HCP5 and MICA. Therefore, we inferred that HCP5 and MICA were associated with the malignant behaviors of GBM.

A six immunophenotype-related lncRNA signature including HCP5 showed good performance in distinguishing tumor immunophenotypes of GBM and HCP5 was highly expressed in immuno-high subtype of GBM, which possessed stronger tumor immune cell infiltration [22]. In this study, we found that HCP5 was up-regulated in GBM and LGG. Besides, HCP5 was also up-regulated in U87 and U251 glioma cell lines [23]. Consistent with the abnormal expression of HCP5, multiple mechanisms indicated the cancer-promoting roles of HCP5. HCP5 regulated malignant behavior of glioma cells through HCP5-miR-139-RUNX1 feedback loop [23]. In contrast, knockdown of HCP5 increased radiosensitivity through cellular senescence and inhibited cell proliferation by regulating microRNA-128 in gliomas [24].

MICA is a ligand of natural killer group 2D (NKG2D) activating receptor and broadly expressed by various types of cancer cells and serves as a target for adoptive immunotherapy [25]. Natural killer cells recognize GBM cells through NKG2D binding to MICA [26], but the different variants of MICA will give different prognosis for GBM [27]. MICA seemed to have immune-mediated antitumor function via the above pathway in GBM, while it displayed the opposite function in colorectal cancer and prostate cancer [28, 29]. Besides, soluble MICA can also inhibit the natural killer cell killing activity in breast carcinoma cells and increase hepatitis C virus-related hepatocellular carcinoma risk [25, 30]. The dual functions of MICA in different cancers indicates that it involves in multiple pathways. By searching in string database, ten proteins were found to be associated with MICA, most of which were the immune-related proteins. The LASSO cox regression model showed MICA was positively correlated with the risk score, which was proved to be associated with the low survival probability. The nomogram further emphasized the prediction roles of MICA in GBM.

Mainstream of researches on MICA focused on its roles as an NKG2D ligand. Here, we found that HCP5 could bind to MICA and then regulate its expression in GBM. This study constructed a prognostic model for GBM with MICA and clinical data for the first time and provided an insight into the molecular therapy of GBM. Further studies are required to uncover the mechanism of regulation between HCP5 and MICA in GBM.

Ethics approval and consent to participate

All of the raw data used in this study were downloaded from public database. The ethics approve and consent are not applicable.

Consent for publication

Not applicable.

Availability of data and materials

The raw data used in this study are available in TCGA database (https://portal.gdc.cancer.gov/), CGGA database (http://www.cgga.org.cn/), GTEx database (https://www.gtexportal.org/) and TIMER database (https://cistrome.shinyapps.io/timer/).

Competing interests

No competing interests.

Funding

Not applicable.

Authors' contributions

QQ is the first author of this study. XW is the corresponding author supervising this work. QQ and WM designed the study and wrote the first manuscript; WM and XZ collected data and performed the statistical analysis. XW performed the project administration. All authors read and approved the final manuscript.

Acknowledgements

We acknowledge the TCGA database, CGGA database and TIMER database for providing their platforms and the contributors for uploading their meaningful datasets. The authors hereby confirm that neither the manuscript nor any part of it has been published or is being considered for publication elsewhere. We acknowledge that all authors participated sufficiently in the work and take public responsibility for its content.

Alexander BM, Cloughesy TF. Adult Glioblastoma. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2017;35(21):2402-9. doi:10.1200/jco.2017.73.0119.
Weller M, van den Bent M, Tonn JC et al. European Association for Neuro-Oncology (EANO) guideline on the diagnosis and treatment of adult astrocytic and oligodendroglial gliomas. The Lancet Oncology. 2017;18(6):e315-e29. doi:10.1016/s1470-2045(17)30194-8.
Cairns RA, Mak TW. Oncogenic isocitrate dehydrogenase mutations: mechanisms, models, and clinical opportunities. Cancer discovery. 2013;3(7):730-41. doi:10.1158/2159-8290.cd-13-0083.
Hvinden IC, Cadoux-Hudson T, Schofield CJ et al. Metabolic adaptations in cancers expressing isocitrate dehydrogenase mutations. Cell reports Medicine. 2021;2(12):100469. doi:10.1016/j.xcrm.2021.100469.
Karpel-Massler G, Nguyen TTT, Shang E et al. Novel IDH1-Targeted Glioma Therapies. CNS drugs. 2019;33(12):1155-66. doi:10.1007/s40263-019-00684-6.
Yan H, Parsons DW, Jin G et al. IDH1 and IDH2 mutations in gliomas. The New England journal of medicine. 2009;360(8):765-73. doi:10.1056/NEJMoa0808710.
Hartmann C, Meyer J, Balss J et al. Type and frequency of IDH1 and IDH2 mutations are related to astrocytic and oligodendroglial differentiation and age: a study of 1,010 diffuse gliomas. Acta neuropathologica. 2009;118(4):469-74. doi:10.1007/s00401-009-0561-9.
Louis DN, Perry A, Wesseling P et al. The 2021 WHO Classification of Tumors of the Central Nervous System: a summary. Neuro-oncology. 2021;23(8):1231-51. doi:10.1093/neuonc/noab106.
Li JH, Liu S, Zhou H et al. starBase v2.0: decoding miRNA-ceRNA, miRNA-ncRNA and protein-RNA interaction networks from large-scale CLIP-Seq data. Nucleic acids research. 2014;42(Database issue):D92-7. doi:10.1093/nar/gkt1248.
Cao Z, Pan X, Yang Y et al. The lncLocator: a subcellular localization predictor for long non-coding RNAs based on a stacked ensemble classifier. Bioinformatics (Oxford, England). 2018;34(13):2185-94. doi:10.1093/bioinformatics/bty085.
Xu C, Zang Y, Zhao Y et al. Comprehensive Pan-Cancer Analysis Confirmed That ATG5 Promoted the Maintenance of Tumor Metabolism and the Occurrence of Tumor Immune Escape. Frontiers in Oncology. 2021;11(772). doi:10.3389/fonc.2021.652211.
Aw JG, Shen Y, Wilm A et al. In Vivo Mapping of Eukaryotic RNA Interactomes Reveals Principles of Higher-Order Organization and Regulation. Molecular cell. 2016;62(4):603-17. doi:10.1016/j.molcel.2016.04.028.
Lu Z, Zhang QC, Lee B et al. RNA Duplex Map in Living Cells Reveals Higher-Order Transcriptome Structure. Cell. 2016;165(5):1267-79. doi:10.1016/j.cell.2016.04.028.
McKinnon C, Nandhabalan M, Murray SA et al. Glioblastoma: clinical presentation, diagnosis, and management. BMJ (Clinical research ed). 2021;374:n1560. doi:10.1136/bmj.n1560.
Ostrom QT, Adel Fahmideh M, Cote DJ et al. Risk factors for childhood and adult primary brain tumors. Neuro-oncology. 2019;21(11):1357-75. doi:10.1093/neuonc/noz123.
Sturm D, Witt H, Hovestadt V et al. Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma. Cancer cell. 2012;22(4):425-37. doi:10.1016/j.ccr.2012.08.024.
Ohgaki H, Kleihues P. The definition of primary and secondary glioblastoma. Clinical cancer research : an official journal of the American Association for Cancer Research. 2013;19(4):764-72. doi:10.1158/1078-0432.ccr-12-3002.
Louis DN, Perry A, Reifenberger G et al. The 2016 World Health Organization Classification of Tumors of the Central Nervous System: a summary. Acta neuropathologica. 2016;131(6):803-20. doi:10.1007/s00401-016-1545-1.
Peng Z, Liu C, Wu M. New insights into long noncoding RNAs and their roles in glioma. Molecular cancer. 2018;17(1):61. doi:10.1186/s12943-018-0812-2.
Yan Y, Xu Z, Li Z et al. An Insight into the Increasing Role of LncRNAs in the Pathogenesis of Gliomas. Frontiers in molecular neuroscience. 2017;10:53. doi:10.3389/fnmol.2017.00053.
Kristensen BW, Priesterbach-Ackley LP, Petersen JK et al. Molecular pathology of tumors of the central nervous system. Annals of oncology : official journal of the European Society for Medical Oncology. 2019;30(8):1265-78. doi:10.1093/annonc/mdz164.
Gao M, Wang X, Han D et al. A Six-lncRNA Signature for Immunophenotype Prediction of Glioblastoma Multiforme. Frontiers in genetics. 2020;11:604655. doi:10.3389/fgene.2020.604655.
Teng H, Wang P, Xue Y et al. Role of HCP5-miR-139-RUNX1 Feedback Loop in Regulating Malignant Behavior of Glioma Cells. Molecular therapy : the journal of the American Society of Gene Therapy. 2016;24(10):1806-22. doi:10.1038/mt.2016.103.
Wang C, Yu G, Xu Y et al. Knockdown of Long Non-Coding RNA HCP5 Increases Radiosensitivity Through Cellular Senescence by Regulating microRNA-128 in Gliomas. Cancer management and research. 2021;13:3723-37. doi:10.2147/cmar.s301333.
Zhang C, Röder J, Scherer A et al. Bispecific antibody-mediated redirection of NKG2D-CAR natural killer cells facilitates dual targeting and enhances antitumor activity. Journal for immunotherapy of cancer. 2021;9(10). doi:10.1136/jitc-2021-002980.
Smith TM, Jr., Tharakan A, Martin RK. Targeting ADAM10 in Cancer and Autoimmunity. Frontiers in immunology. 2020;11:499. doi:10.3389/fimmu.2020.00499.
Tang J, Shao W, Dorak MT et al. Positive and negative associations of human leukocyte antigen variants with the onset and prognosis of adult glioblastoma multiforme. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology. 2005;14(8):2040-4. doi:10.1158/1055-9965.epi-05-0136.
Espinoza I, Agarwal S, Sakiyama M et al. Expression of MHC class I polypeptide-related sequence A (MICA) in colorectal cancer. Frontiers in bioscience (Landmark edition). 2021;26(10):765-76. doi:10.52586/4986.
Sakiyama MJ, Espinoza I, Reddy A et al. Race-associated expression of MHC class I polypeptide-related sequence A (MICA) in prostate cancer. Experimental and molecular pathology. 2019;108:173-82. doi:10.1016/j.yexmp.2019.04.010.
Lo PH, Urabe Y, Kumar V et al. Identification of a functional variant in the MICA promoter which regulates MICA expression and increases HCV-related hepatocellular carcinoma risk. PloS one. 2013;8(4):e61279. doi:10.1371/journal.pone.0061279.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Comprehensive analysis to identify long non-coding RNA HCP5/MICA axis as a prognostic biomarker in glioblastoma

Status:

Version 1

Abstract

Purpose

Methods

Results

Conclusions

Figures

Introduction

Materials And Methods

Data source and processing

Identification of Differentially Expressed Genes

Determination of the hub genes

Expression analyses of HCP5 and MICA in GBM

Methylation analysis of MICA in GBM

Construction of gene signature with MICA-associated genes

Construction of the predictive nomogram

Results

Somatic mutational phenotypes in glioblastoma and IDH1 mutation frequencies in pan-cancer.

Common Differentially Expressed Genes in TCGA and CGGA

GO and KEGG pathway analysis of common DEGs

Determination of the hub genes

Expression of HCP5 and MICA

Correlation analysis between MICA expression and immune infiltration and immune checkpoints.

Methylation analysis of MICA in GBM

The MICA-associated gene signature in GBM

Prognostic model for GBM

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1