BIRC5 and PRLR Might Be Novel Immune-Related Prognostic Biomarkers in Adrenocortical Carcinoma (ACC)

doi:10.21203/rs.3.rs-621629/v1

Download PDF

Research

BIRC5 and PRLR Might Be Novel Immune-Related Prognostic Biomarkers in Adrenocortical Carcinoma (ACC)

https://doi.org/10.21203/rs.3.rs-621629/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Adrenocortical carcinoma (ACC) is a highly aggressive tumor，which occurs in adrenal cortex. This study aims to identify new immune-related molecular biological markers and therapeutic targets in ACC patients.

Results: First, we identified 134 genes by constructing weighted gene co-expression network. Among them, two genes (BIRC5 and PRLR) were eventually regarded as immune-related prognosis biomarkers on the basis of The Immunology Database and Analysis Portal (ImmPort) database. The validation of mRNA level was meaningful (based on Oncomine database and TCGA-ACC). Four methods (overall survival (OS) analysis, disease-free survival (DFS) analysis, stage plot and expression value comparison) were used in GEPIA to verify that PRLR and BIRC5 have significant correlation with the prognosis of ACC. Several other methods including survival analysis, analysis of variance, the spearman correlation analysis, distance correlation analysis and receiver operating characteristic curve were also used for the external validation of them in GEO datasets (GSE10279, GSE19750, GSE76019 and GSE76021). Furthermore, we evaluated the relationship between hub gene expression of and the degree of immune cell infiltration in ACC by ESTIMATE algorithm and ssGSEA. BIRC5 expression was significantly positively related to activated CD4 T cell and type 2 T helper cell. And PRLR expression was positively related to CD56dim NK cell and type 17 T helper cell. Moreover, Gene set enrichment analysis (GSEA) suggested that PRLR has strong correlation with neuroactive ligand receptor interaction and BIRC5 was significantly enriched in cell cycle. In addition, according to Drug Signatures Database (DSigDB), BIRC5 were significantly enriched in “resveratrol_mcf7_down”, “lucanthone_ctd_00006227”, “8-azaguanine_pc3_down”, “thioguanosine _mcf7_down”, “dasatinib_ctd_00004330”, “monobenzone_pc3_down”.

Conclusions: In summary, BIRC5 and PRLR were shown to be related to the prognosis of patients with ACC, indicating them as novel immune-related prognostic biomarkers and immunotherapy targets in ACC.

Molecular Genetics

Internal Medicine

Adrenocortical carcinoma

immune-related prognostic biomarkers

WGCNA

Adrenocortical carcinoma (ACC) is a highly aggressive tumor,which occurs in adrenal cortex. The overall 5-year survival rate is 35% due to highly aggressive biological behavior(1). It mostly occurs between the ages 40 and 60, with higher prevalence in female (up to 60%)(2). Since ACC has few symptoms in the early stage, many patients have already advanced in the late stage when they are diagnosed(3). At these stages, ACC has extensive invasiveness and metastasis with limited treatment options. Actually, the current treatments for ACC, including surgery, chemotherapy and radiotherapy, shows poor efficacy and prognosis(4). Under current circumstance, it is urgent to study the mechanism that promotes the progress of ACC and find new molecular biological markers to predict the prognosis of ACC.

Many scholars are dedicated to studying the impact of the immune system on tumor progression. In recent years, accumulating studies indicated that the tumor immune microenvironment (TIM) plays a vital role in tumor progression, invasion and metastasis(5). The infiltration of various immune cells might affect the proliferation and metastasis of tumor cells(6). Papewalis C et al demonstrated that autologous DCs may serve as adjuvants for immunotherapy in ACC(7). Kanczkowski W et al revealed the significant reduction in the expression and signal transduction of TLR4 and CD14 in malignant but not benign adrenal tumors, which is a novel feature of adrenocortical tumors and may be used as diagnostic markers(8). Parise IZS et al showed that CD8 + T Lymphocytes could be the prognostic markers for ACC(9). To date, the development of immunotherapy for ACC was rarely documented because of the uncertainty in the relationship between tumor cells and the tumor microenvironment (TME). Therefore, in our study, we aimed to screen out novel diagnostic and prognostic biomarkers related to immune for ACC, which can also be used as therapeutic targets for immunotherapy meanwhile.

Based on the facts above, we identified modules associated with prognosis by weighted gene co-expression network. Then the hub genes in the key modules were selected. Following, we screened out two immune-related genes (baculoviral IAP repeat containing 5 (BIRC5) and prolactin receptor (PRLR)) from them. Then, we conducted internal and external validation, respectively, to further verify our results. It is worth mentioning that we regard survival time and survival status as important prognostic yardstick, so as to screen out prognostic-related modules. Previous studies chosen prognosis-related modules only by survival time or survival status. As far as we know, this study initially explore the potential relationship between immune cell infiltration and prognosis biomarkers by single-sample gene-set enrichment analysis (ssGSEA) in ACC. These two genes could be novel prognostic biomarkers and therapeutic targets in ACC. Moreover, based on Drug Signatures Database (DSigDB)(10), which is a new gene set resource that uses drugs/compounds and their target genes for gene set enrichment analysis (GSEA), we also studied the correlation between hub gene and six kinds of drugs in order to find new targets of anticancer drugs.

Data collection

Gene expression data and the related clinical information for ACC (TCGA-ACC) were collected from the The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). Seventy-nine ACC samples in TCGA database were included to construct co-expression network. Another 4 datasets, which were performed as test sets including GSE10927(11), GSE19750(12), GSE76019(13), and GSE76021(14) were obtained from Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/). In total, 33 ACCs from GSE10927, 44ACCs from GSE19750, 34 ACCs from GSE76019 and 29 ACCs from GSE76021 were included in this research. Furthermore, immune-related genes (IRGs) were acquired from the ImmPort database (https://www.immport.org). In total, there were 2499 IRGs in database, and 1811 IRGs overlapping with TCGA-ACC expression matrix gene list were screened out for subsequent analysis.

Data Preprocessing

The expression matrix was displayed as count number. For subsequent analysis, the raw expression data of TCGA-ACC was preprocessed by R package “DEseq.2”(15), by which we took normalization and log2 transformation too. Moreover, normalization and log2 transformation of GEO datasets (GSE10927, GSE19750, GSE76019 and GSE76021) was conducted using the R package “affy”(16).

Co-expression network construction

Firstly, we use the gsg (goodSamplesGenes) method to check whether the top 5000 genes in the expression profile are suitable for WGCNA. Moreover, we distinguished the outlying samples which were distant from other samples. All outliers were excluded from the ananlysis. Then, co-expression network was constructed by using the “WGCNA” R package(17). An appropriate β (soft threshold power beta) was selected according to the scale-free topology criterion. Next, we calculated topological overlap matrix (TOM) using the adjacency matrix to measure the network connectivity of genes. Then, we conducted three different branch cutting methods to divide genes into gene modules. The cut line of merging modules is based on the dissimilarity of module eigengenes (MEs).

Identification of Clinically Significant Modules

Firstly, the gene significance (GS) and module significance (MS, the average GS for genes in a module) were obtained to quantify the relationship between gene expression and clinical features. In general,, the module which has the largest absolute MS was regarded as the module related to clinical feature. In addition, we calculated the module membership (MM) to assess the correlation between ME and gene expression.

Functional and Pathway Enrichment Analysis

We performed Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis pathway by using R package “clusterProfiler”(18) to demonstrate the mechanism of genes in key modules. P < 0.05 were considered to be cutoff.

Identification of hub genes

After the key modules were chosen, hub genes were defined under the threshold of |MM|≥0.80 and |GS|≥0.20. Then, in order to identify immune-related prognosis biomarkers, we regarded the intersection between the hub genes in WGCNA and IRGs in Immport database as immune-related prognostic biomarkers.

Validation of hub genes

Firstly, based on the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn/), the relationship between hub genes and prognosis was verified by four methods including survival analysis (overall survival (OS), disease-free survival (DFS)), stage plot and expression values comparison.

Then, four additional independent validation GEO datasets (GSE10927, GSE19750, GSE76019 and GSE76021) were obtained to further test the prognostic role of hub genes by six methods. The analysis of variance (ANOVA) test and the Spearman correlation analysis were carried out by using SPSS (version 21.0), which was visualized by R package “ggstatsplot”(19). Meanwhile, the distance correlation analysis was completed by R package “energy”(20). Survival analysis was carried out by R package “survival”(21). We divide ACCs into high gene expression group and low gene expression group and generate Kaplan-Meier survival curve. In addition, by virtue of R package “plotROC”(22), we performed receiver operating characteristic curve (ROC) analysis and calculated the area under the curve (AUC) to measure the resolving power for distinguishing the adrenocortical tumor from normal adrenal glands. Finally, the mRNA-level expression of hub genes between ACC and normal tissues was compared on the basis of Oncomine database(23) (https://www.oncomine.org/).

Relationship between gene expression and TME

We performed ssGSEA(24) to assess the infiltration of various immune cell in the TME using R package “GSVA”(25). The characteristic genes for each immune cell were obtained from other research(26). Moreover, Estimation of STromal and Immune cells in MAlignant Tumors using Expression data (ESTIMATE) algorithm(27) was applied to value stromal and immune microenvironment infiltration. ESTIMATE score is defined as tumor purity, which is the sum of stromal and immune score.

Gene Set Enrichment Analysis (GSEA)

To reveal the underlying function of potential prognostic biomarkers for ACC, we performed gene set enrichment analysis (GSEA). We divide ACCs into high gene expression group and low gene expression group. In our study, “DSigDB.ALL.gmt” obtained from Molecular Signatures Database (MSigDB) and “c2.cp.kegg.v6.0.symbols.gmt” were selected as the reference sets. Gene size ≥ 100, FDR < 25%, and |enrichment score (ES)|> 0.60 were considered as the cut-off criteria for significantly enriched KEGG pathways.

Figure 1 provided the research process of this study

Co-expression network construction to identify clinically significant modules

TCGA_PK_A4HP and TCGA_OR_A5JB were excluded from the subsequent study (Fig. 2(a)). Seventy-seven samples with were included in the analysis (Fig. 2(b)). In this research, the power of β = 4 (scale-free R^2 = 0.92) was determined as the soft thresholding to construct a scale-free network (Fig. 3).

We identified 23 modules (Fig. 4(a)) by using hierarchical average linkage clustering. Then we screened out modules related to the prognosis of ACC. Firstly, blue module was correlated with OS time (r = 0.35, p = 0.002, Fig. 4(b)) and OS (r = -0.61, p = 1e-08, Fig. 4(b)). Meanwhile, brown module was related with OS (r = 0.51, p = 4e-06, Fig. 4(b)) and OS time (r =-0.52, p = 3e-06, Fig. 4(b)). In addition, the MS (of OS time) of brown module ranked first among all modules, and the MS (of OS) of blue module was the highest in all modules (Fig. 4(d)). The relationship between MM and GS with OS time (cor = 0.77, p = 1.5e-180)/ OS (cor = 0.69, p = 1.9e-130) in brown module were significant too(Fig. 4(c)). The same results occurred in blue module (OS time: cor = 0.51, p = 9.7e-53; OS: cor = 0.71, p < 1e-200) (Fig. 4(c)). Consequently, we chose the blue and brown modules for further analysis.

GO Biological Processes and KEGG Analysis

We conducted GO and KEGG pathway enrichment analysis to explore the underlying function of genes in prognosis-related modules. On the basis of GO analysis, genes in the blue module were enriched in four BPs, including pattern specification process, regulation of membrane potential, embryonic organ morphogenesis and embryonic skeletal system development (Fig. 5(a)). As for the brown module, genes were enriched in 209 BPs (Supplementary Table S1). The top 10 terms were organelle fission, nuclear division, chromosome segregation, nuclear chromosome segregation, mitotic nuclear division, DNA replication, DNA conformation change, sister chromatid segregation, mitotic sister chromatid segregation and regulation of chromosome segregation (Fig. 5(c)). While in KEGG pathway enrichment analysis, genes in blue module were enriched in two KEGG pathways, including neuroactive ligand-receptor interaction and axon guidance (Fig. 5(b)). Meanwhile, genes in the brown module were enriched in six pathways, including cell cycle, oocyte meiosis, fanconi anemia pathway, progesterone-mediated oocyte maturation, DNA replication and homologous recombination (Fig. 5(d)).

Identification of hub genes

First, 1770 genes that reached the cutoff criterion (|GS|≥0.2) for both OS time and OS were selected (Fig. 6(a)). Then we got 16 genes from the blue module based on the screening criterion (|GS|≥0.2, |MM|≥0.8) (Fig. 6(b)). In the same way, we got 118 genes from the brown module (Fig. 6(c)). In all, 134 hub genes in key modules were screened out. Moreover, we tried to find hub genes associated with immunity. Therefore, two genes including BIRC5 (baculoviral IAP repeat containing 5) and PRLR (prolactin receptor), which were overlapping in WGCNA and IRGs gene lists, were eventually identified as immune-related prognosis biomarkers for further validation (Fig. 6(d)).

Hub Gene Validation

After screening out two biomarkers, we performed internal validation. At the beginning, according to the GEPIA, higher BIRC5 expression was significantly correlated with worse OS (Hazard ratio [HR] = 6.5, p = 6.5e-06, Fig. 7(a)) and DFS (HR = 3.3, p = 0.00054, Fig. 7(b)). Furthermore, we found that patients in high PRLR expression group had better OS (HR = 0.19, p = 7.2e-05, Fig. 7(e)) and DFS (HR = 0.29, p = 0.00039, Fig. 7(f)). Then we made a comparison of the mRNA expression of selected genes in ACC tumor tissue with normal tissue, the results revealed higher expression of BIRC5 (p < 0.05, Fig. 7(c)) in ACC samples compared with normal samples. It should be pointed out that although it was not significantly that PRLR expression in normal adrenal glands was higher than that in ACC tissues, there was still such a trend that the median expression of PRLR in normal adrenal glands was higher than that in ACC samples (Fig. 7(g)). Furthermore, the results also showed that patients with high BIRC5 expression (F = 6.06, p = 0.000981; Fig. 7(d)) had higher tumor stage. On the contrary, low expression of PRLR (F = 3.53, p = 0.0191; Fig. 7(h)) was significantly correlated to higher tumor stage.

On the other hand, we performed serious external validation. The results suggested that higher expression of BIRC5 was significantly related with worse OS (GSE10927: p = 0.033, Fig. 8(a); GSE19750: p = 0.0035, Fig. 8(b)) and EFS (Event-free survival) (GSE76019: p = 0.012, Fig. 8(c); GSE76021: p = 0.00012, Fig. 8(d)). Meanwhile, higher PRLR expression was significantly related with better OS (GSE10927: p = 0.039, Fig. 8(e); GSE19750: p = 0.039, Fig. 8(f)) and EFS (GSE76019: p = 0.031, Fig. 8(g); GSE76021: p = 0.041, Fig. 8(h)). On the basis of the Oncomine database, BIRC5 expression in ACC tissues is higher than that in normal adrenal glands (p = 0.01, Fig. 9(a)), and PRLR expression is higher in normal adrenal glands than that in ACC tissues (p = 1.59e-4, Fig. 9(b)). As shown in Fig. 10, BIRC5 was significantly correlated with ACC prognosis in one-way ANOVA (TCGA-ACC: p = 0.000003, F = 25.053; GSE76021: p = 0.000003, F = 35.021; GSE76019: p = 0.001, F = 13.624; Fig. 10(a)), spearman correlation analysis (TCGA-ACC: p = 0.000007, F = 0.483; GSE76021: p = 0.000001, F = 0.769; GSE76019: p = 0.001, F = 0.533; Fig. 10(b)) and distance correlation (TCGA-ACC: p = 2.2e-16, F = 0.508; GSE76021: p = 2.2e-16, F = 0.77848; GSE76019: p = 2.16e-11, F = 0.5595; GSE10927: p = 2.2e-16, F = 0.50755, Fig. 10(c)). PRLR was significantly correlated with ACC prognosis in one-way ANOVA (TCGA-ACC: p = 0.000003, F = 25.531; GSE76021: p = 0.028, Fig. 10(a)), spearman correlation analysis (TCGA-ACC: p = 0.000005, F = -0.49; Fig. 10(b)) and distance correlation (TCGA-ACC: p = 2.2e-16, F = 0.504; GSE76021: p = 0.000951, F = 0.45833; GSE76019: p = 0.0464, F = 0.33021; GSE10927: p = 2.2e-16, F = 0.50379, Fig. 10(c)). The results of ROC analysis indicated that BIRC5 (TCGA-ACC: AUC = 0.79, Fig. 11(a); GSE10927: AUC = 0.71, Fig. 11(b); GSE76019: AUC = 0.82, Fig. 11(c); GSE76021: AUC = 0.96, Fig. 11(d)) and PRLR (TCGA-ACC: AUC = 0.8, Fig. 11E; GSE10927: AUC = 0.71, Fig. 11(f); GSE76019: AUC = 0.69, Fig. 11(g); GSE76021: AUC = 0.72, Fig. 11(h)) both could significantly distinguish ACC samples from normal samples.

Relationship Between Gene Expression And TME

Varieties of immune cells in TME play an vital role in regulating anti-tumor response. Therefore, we conducted ssGSEA to determine the degree of alteration in the particular component of tumor infiltrating lymphocytes (TILs). Our study involved 28 immunophenotypes, including B cells, T cells, dendritic cells (DC) and other tumor-infiltrating lymphocytes. The results revealed that activated CD4 T cell and type 2 T helper cell were positively related with the expression levels of BIRC5 (Fig. 12(a)). And the level of CD56dim natural killer cell and type 17 T helper cell was positively related with PRLR expression (Fig. 12(c)). Pearson's correlation analysis suggested that the abundance of anti-tumor immune cells and pro-tumor immune cells was positively correlated in local environment (BIRC5: R = 0.9008, p < 0.001, Fig. 12(b); PRLR: R = 0.9008, p < 0.001, Fig. 12(d)).

Gene Set Enrichment Analysis

GSEA showed that BIRC5 was significantly enriched in KEGG pathways including cell cycle (nominal p < 0.001, n = 122, FDR < 25%, ES = -0.72, Fig. 13(a)). Meanwhile, we found that PRLR was significantly related to neuroactive ligand receptor interaction (nominal p = 0.002, n = 245, FDR = 19.1%, ES = -0.51, Fig. 13(b)).

On the other hand, we also used DSigDB to explore the drug pathways enriched in BIRC5 high expression samples. BIRC5 were mainly enriched in “resveratrol_mcf7_down” (size = 100, ES = -0.830, nominal p ≤ 0.001, FDR = 0.02340), “lucanthone_ctd_00006227” (size = 202, ES = -0.806, nominal p ≤ 0.001, FDR = 0.0447), “8-azaguanine_pc3_down” (size = 192, ES = -0.672, nominal p ≤ 0.001, FDR = 0.0356), “thioguanosine _mcf7_down” (size = 145, ES = -0.646, nominal p ≤ 0.001, FDR = 0.02331), “dasatinib_ctd_00004330” (size = 474, ES = -0.636, nominal p = 0.002, FDR = 0.04433), “monobenzone_pc3_down” (size = 196, ES = -0.627, nominal p ≤ 0.001, FDR = 0.0516) (Fig. 13(c)).

Adrenocortical carcinoma is prone to local invasion and metastasis, with poor prognosis and limited treatment(28). Currently, the preferred treatment for ACC is radical resection of the solid tumor(29). However, almost half of the patients present disseminated metastasis at the first diagnosis, and locoregional metastases or recurrence occur in approximately one-third of patients after surgery(30). For these patients, mitotane is the only approved treatment by the Food and Drug Administration and is generally poorly tolerated with limited efficacy(31). Recently, the development of immunotherapy has changed the treatment model of various cancers(32), and increasing researches have focused on immune-related molecules. However, some studies had shown that current immunotherapy did not show the expected effect on ACC. Fiorentiniv C et al pointed out that in numerous cancers including ACC, the preliminary clinical findings about anti-PD-L1 drugs seemed to be unsatisfactory(33). ACC was even considered by some scholars to be a tumor with low immune response(34). Therefore, it is urgently to explore novel immune-related prognosis biomarkers for ACC.

In this study, we constructed a weight gene co-expression network by using TCGA-ACC data. Tow clinical features (OS time, OS) were considered as important prognostic indicators. We also tried to find the key modules which were most related to the OS Time or OS of ACC. Among 23 modules, the blue and brown modules were the most associated with OS Time and OS. One hundred and thirty-four genes which reached the cutoff were considered as hub genes in key modules. To determine the immune-related genes among the hub genes, we used the ImmPort database. Finally, BIRC5 and PRLR were identified as novel immune-related prognosis biomarkers for ACC.

We used two databases to perform the validations of hub genes. GEPIA analysis confirmed that PRLR expression was lower in ACC than that in normal adrenal glands based on the TCGA-ACC data. In addition, Oncomine analysis showed that PRLR expression was higher in normal adrenal glands than in ACC samples. And BIRC5 showed extremely opposite results. Furthermore, whether on the basis of GEPIA analysis, GSE10927, GSE76019, or GSE76021 database, PRLR and BIRC5 were significantly related with the prognosis of ACC, demonstrating that PRLR and BIRC5 were powerful immune-related prognostic biomarkers in ACC.

Since the level of immune infiltration has a strong correlation with the prognosis of tumors(35), we used the ESTIMATE algorithm and ssGSEA to study the relationship between the hub gene expression and the level of immune cell infiltration. BIRC5 expression was significantly positively correlated with activated CD4 T cell and type 2 T helper cell. And PRLR expression was significantly positively related to CD56dim natural killer cell and type 17 T helper cell. Thus, BIRC5 and PRLR might be potential therapeutic targets for immunotherapy in ACC. The analysis of immune cell that associated with hub genes further demonstrated the possible action pathway of immune-related genes. BIRC5 and PRLR may influence the prognosis of ACC by different immune pathways. According to DSigDB, we screened six potential drug targets for BIRC5 with GSEA analysis, which may become the targets of immunotherapy.

BIRC5 (also called survivin) is one of the inhibitor of apoptosis protein(36), which can inhibit apoptosis-related signaling pathways and promote cell proliferation(37). It has been previously identified as one of the prognostic biomarkers for malignant tumor (38–41). Functionally, BIRC5 can stabilize the mitotic apparatus, ensure proper segregation of chromosomes and maintain microtubule integrity. It contributes to complete cell division stably and effectively (42–46), which is indispensable in the tumor cell growth. In addition to being a contributing factor of almost all cancer type, increasing data indicates a critical role of BIRC5 in regulating the proliferation of non-malignant, physiological adult T cells(40, 47, 48). This is in line with our study, in which the expression of activated CD4 T cell and type 2 T helper cell was significantly higher in high expression BIRC5 samples than that in low expression BIRC5 samples. Although BIRC5 is considered as a vital cancer gene, only limited progress has been made in finding novel treatments targeting this pathway so far (49). Under these circumstances, scholars have studied the survivin inhibitor YM155, which is expected to shut down the transcription of the BIRC5 gene(50). But further study of YM155 has been less successful(51). Therefore, whether BIRC5 could be immune-related therapeutic target still requires more researches.

PRLR (prolactin receptor) is one of the cytokine receptor and its role in the genesis and progression of tumors is still controversial. Prolactin (PRL) is not only expressed on the surface of breast epithelial cells, but also on the surface of many other cells(52). Knockout of the PRLR locus suppresses the progression of breast tumors(53). In humans, PRLR expression higher in malignant epithelium than adjacent normal tissue(54). But various research on the relationship between PRLR and PRL and breast tumors remains inconclusive(55). On the other hand, the expression of prolactin in prostate cancer is correlated with high tumor grade and aggressive course(56, 57). However, the relationship between PRLR and adrenal tissue is uncertain. Glasow’s research showed that PRL inhibits the proliferation of adrenal cells and PRLR is expressed in adrenocortical tumors but not in pheochromocytoma(58). Our results actually suggested that PRLR contributed to a favorable outcome in the prognosis of ACC, which is in accordance with Glasow’s results. These differences may be due to differences in research methods, or differences in the biological characteristics of PRLR in different tissues.

There are several limitations of our study. First, there are no experimental data (in vivo and in vitro validation) to support these findings at present. Secondly, the relationship between the two hub genes has not been further explored. Finally, the mechanism of the influence of the two biomarkers on the prognosis of ACC needs to be further studied. Therefore, more molecular biology experiments are supposed to be carried out to study the functions of the hub genes.

In conclusion, by using various bioinformatics methods, we identified two immune-related prognostic biomarkers (BIRC5, PRLR) of ACC. Furthermore, the underlying function of genes related to prognosis were predicted. These biomarkers have great potential in translational medicine and may become immunotherapy targets for ACC in the future..

ACC: Adrenocortical carcinoma

ImmPort: The Immunology Database and Analysis Portal

GSEA: Gene set enrichment analysis

DSigDB: Drug Signatures Database

TME: tumor microenvironment

BIRC5: baculoviral IAP repeat containing 5

PRLR: prolactin receptor

ssGSEA: single-sample gene-set enrichment analysis

TCGA: The Cancer Genome Atlas

GEO: Gene Expression Omnibus

IRGs: immune-related genes

WGCNA: weighted gene co-expression network analysis

GO: Gene Ontology

KEGG: Kyoto Encyclopedia of Genes and Genomes

GEPIA: Gene Expression Profiling Interactive Analysis

ANOVA: analysis of variance

ROC: receiver operating characteristic curve

AUC: area under the curve

MSigDB: Molecular Signatures Database

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and material

All analyzed data related to this paper are included in this paper.

Competing interests

The authors declare that they have no competing interests.

Funding

This study was supported by the medical science and technology innovation platform support project of Zhongnan Hospital of Wuhan University (PTXM2019006).

Author’s contributions

T. L., X. B. and X. Y. designed the research, X. B. and X. Y. conducted the study procedures. X. B., X. Y., G. D., Z. C., K. T., H. W. and T. L. analyzed the results. T. L. and X. B. provided analysis tools. X. B. and X. Y. wrote the manuscript. All authors reviewed the manuscript.

Acknowledgements

We would like to acknowledge the TCGA and GEO database developed by the National Institutes of Health (NIH).

Else T, Kim AC, Sabolch A, Raymond VM, Kandathil A, Caoili EM, et al. Adrenocortical carcinoma. Endocr Rev. 2014;35(2):282-326.
Sharma E, Dahal S, Sharma P, Bhandari A, Gupta V, Amgai B, et al. The Characteristics and Trends in Adrenocortical Carcinoma: A United States Population Based Study. J Clin Med Res. 2018;10(8):636-40.
Fay AP, Elfiky A, Telo GH, McKay RR, Kaymakcalan M, Nguyen PL, et al. Adrenocortical carcinoma: the management of metastatic disease. Crit Rev Oncol Hematol. 2014;92(2):123-32.
Allolio B, Fassnacht M. Clinical review: Adrenocortical carcinoma: clinical update. J Clin Endocrinol Metab. 2006;91(6):2027-37.
Binnewies M, Roberts EW, Kersten K, Chan V, Fearon DF, Merad M, et al. Understanding the tumor immune microenvironment (TIME) for effective therapy. Nat Med. 2018;24(5):541-50.
Gajewski TF, Schreiber H, Fu YX. Innate and adaptive immune cells in the tumor microenvironment. Nat Immunol. 2013;14(10):1014-22.
Papewalis C, Fassnacht M, Willenberg HS, Domberg J, Fenk R, Rohr UP, et al. Dendritic cells as potential adjuvant for immunotherapy in adrenocortical carcinoma. Clin Endocrinol (Oxf). 2006;65(2):215-22.
Kanczkowski W, Tymoszuk P, Ehrhart-Bornstein M, Wirth MP, Zacharowski K, Bornstein SR. Abrogation of TLR4 and CD14 expression and signaling in human adrenocortical tumors. J Clin Endocrinol Metab. 2010;95(12):E421-9.
Parise IZS, Parise GA, Noronha L, Surakhy M, Woiski TD, Silva DB, et al. The Prognostic Role of CD8(+) T Lymphocytes in Childhood Adrenocortical Carcinomas Compared to Ki-67, PD-1, PD-L1, and the Weiss Score. Cancers (Basel). 2019;11(11).
Yoo M, Shin J, Kim J, Ryall KA, Lee K, Lee S, et al. DSigDB: drug signatures database for gene set analysis. Bioinformatics. 2015;31(18):3069-71.
Giordano TJ, Kuick R, Else T, Gauger PG, Vinco M, Bauersfeld J, et al. Molecular classification and prognostication of adrenocortical tumors by transcriptome profiling. Clin Cancer Res. 2009;15(2):668-76.
Legendre CR, Demeure MJ, Whitsett TG, Gooden GC, Bussey KJ, Jung S, et al. Pathway Implications of Aberrant Global Methylation in Adrenocortical Cancer. PLoS One. 2016;11(3):e0150629.
Surakhy M, Wallace M, Bond E, Grochola LF, Perez H, Di Giovannantonio M, et al. A common polymorphism in the retinoic acid pathway modifies adrenocortical carcinoma age-dependent incidence. Br J Cancer. 2020;122(8):1231-41.
Pinto EM, Rodriguez-Galindo C, Choi JK, Pounds S, Liu Z, Neale G, et al. Prognostic Significance of Major Histocompatibility Complex Class II Expression in Pediatric Adrenocortical Tumors: A St. Jude and Children's Oncology Group Study. Clin Cancer Res. 2016;22(24):6247-55.
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550.
Gautier L, Cope L, Bolstad BM, Irizarry RA. affy--analysis of Affymetrix GeneChip data at the probe level. Bioinformatics. 2004;20(3):307-15.
Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.
Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284-7.
Patil I. Visualizations with statistical details: The ‘ggstatsplot’ approach. PsyArXiv. 2021.
Rizzo ML, Szekely GJ. Energy: E-statistics: Multivariate inference via the energy of data. R Package Version 17-2. 2017.
Therneau TM. survival: Survival Analysis. Technometrics. 2015;46(1):111-2.
Sachs MC. plotROC: A Tool for Plotting ROC Curves. J Stat Softw. 2017;79.
Rhodes DR, Yu J, Shanker K, Deshpande N, Varambally R, Ghosh D, et al. ONCOMINE: A Cancer Microarray Database and Integrated Data-Mining Platform. Neoplasia. 2004;6(1):1-6.
Barbie DA, Tamayo P, Boehm JS, Kim SY, Moody SE, Dunn IF, et al. Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature. 2009;462(7269):108-12.
Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC bioinformatics. 2013;14:7.
Charoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep. 2017;18(1):248-62.
Yoshihara K, Shahmoradgoli M, Martinez E, Vegesna R, Kim H, Torres-Garcia W, et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat Commun. 2013;4:2612.
Fassnacht M, Libe R, Kroiss M, Allolio B. Adrenocortical carcinoma: a clinician's update. Nat Rev Endocrinol. 2011;7(6):323-35.
Fassnacht M, Kroiss M, Allolio B. Update in adrenocortical carcinoma. J Clin Endocrinol Metab. 2013;98(12):4551-64.
Else T, Williams AR, Sabolch A, Jolly S, Miller BS, Hammer GD. Adjuvant therapies and patient and tumor characteristics associated with survival of adult patients with adrenocortical carcinoma. J Clin Endocrinol Metab. 2014;99(2):455-61.
Head L, Kiseljak-Vassiliades K, Clark TJ, Somerset H, King J, Raeburn C, et al. Response to Immunotherapy in Combination With Mitotane in Patients With Metastatic Adrenocortical Cancer. J Endocr Soc. 2019;3(12):2295-304.
Drake CG, Lipson EJ, Brahmer JR. Breathing new life into immunotherapy: review of melanoma, lung and kidney cancer. Nat Rev Clin Oncol. 2014;11(1):24-37.
Fiorentini C, Grisanti S, Cosentini D, Abate A, Rossini E, Berruti A, et al. Molecular Drivers of Potential Immunotherapy Failure in Adrenocortical Carcinoma. J Oncol. 2019;2019:6072863.
Zheng S, Cherniack AD, Dewal N, Moffitt RA, Danilova L, Murray BA, et al. Comprehensive Pan-Genomic Characterization of Adrenocortical Carcinoma. Cancer Cell. 2016;29(5):723-36.
Xiong Y, Wang K, Zhou H, Peng L, You W, Fu Z. Profiles of immune infiltration in colorectal cancer and their clinical significant: A gene expression-based study. Cancer Med. 2018;7(9):4496-508.
Altieri DC. Survivin - The inconvenient IAP. Semin Cell Dev Biol. 2015;39:91-6.
Ungvari I, Hadadi E, Virag V, Bikov A, Nagy A, Semsei AF, et al. Implication of BIRC5 in asthma pathogenesis. Int Immunol. 2012;24(5):293-301.
Kim YH, Kim SM, Kim YK, Hong SP, Kim MJ, Myoung H. Evaluation of survivin as a prognostic marker in oral squamous cell carcinoma. J Oral Pathol Med. 2010;39(5):368-75.
Rödel F, Sprenger T, Kaina B, Liersch T, Rödel C, Fulda S, et al. Survivin as a prognostic/predictive marker and molecular target in cancer therapy. Curr Med Chem. 2012;19(22):3679-88.
Shariat SF, Karakiewicz PI, Godoy G, Karam JA, Ashfaq R, Fradet Y, et al. Survivin as a prognostic marker for urothelial carcinoma of the bladder: a multicenter external validation study. Clin Cancer Res. 2009;15(22):7012-9.
Xie Y, Ma X, Gu L, Li H, Chen L, Li X, et al. Prognostic and Clinicopathological Significance of Survivin Expression in Renal Cell Carcinoma: A Systematic Review and Meta-Analysis. Sci Rep. 2016;6:29794.
Ruchaud S, Carmena M, Earnshaw WC. The chromosomal passenger complex: one for all and all for one. Cell. 2007;131(2):230-1.
Kitagawa M, Lee SH. The chromosomal passenger complex (CPC) as a key orchestrator of orderly mitotic exit and cytokinesis. Front Cell Dev Biol. 2015;3:14.
Niedzialkowska E, Wang F, Porebski PJ, Minor W, Higgins JM, Stukenberg PT. Molecular basis for phosphospecific recognition of histone H3 tails by Survivin paralogues at inner centromeres. Mol Biol Cell. 2012;23(8):1457-66.
Altieri DC. Survivin and IAP proteins in cell-death mechanisms. Biochem J. 2010;430(2):199-205.
Arora V, Cheung HH, Plenchette S, Micali OC, Liston P, Korneluk RG. Degradation of survivin by the X-linked inhibitor of apoptosis (XIAP)-XAF1 complex. J Biol Chem. 2007;282(36):26202-9.
Andersson KME, Brisslert M, Cavallini NF, Svensson MND, Welin A, Erlandsson MC, et al. Survivin co-ordinates formation of follicular T-cells acting in synergy with Bcl-6. Oncotarget. 2015;6(24):20043-57.
Sharief MK, Semra YK. Heightened expression of survivin in activated T lymphocytes from patients with multiple sclerosis. J Neuroimmunol. 2001;119(2):358-64.
Wheatley SP, Altieri DC. Survivin at a glance. J Cell Sci. 2019;132(7).
Rauch A, Hennig D, Schafer C, Wirth M, Marx C, Heinzel T, et al. Survivin and YM155: how faithful is the liaison? Biochim Biophys Acta. 2014;1845(2):202-20.
Kudchadkar R, Ernst S, Chmielowski B, Redman BG, Steinberg J, Keating A, et al. A phase 2, multicenter, open-label study of sepantronium bromide (YM155) plus docetaxel in patients with stage III (unresectable) or stage IV melanoma. Cancer Med. 2015;4(5):643-50.
Damiano JS, Wasserman E. Molecular pathways: blockade of the PRLR signaling pathway as a novel antihormonal approach for the treatment of breast and prostate cancer. Clin Cancer Res. 2013;19(7):1644-50.
Oakes SR, Robertson FG, Kench JG, Gardiner-Garden M, Wand MP, Green JE, et al. Loss of mammary epithelial prolactin receptor delays tumor formation by reducing cell proliferation in low-grade preinvasive lesions. Oncogene. 2007;26(4):543-53.
Touraine P, Martini JF, Zafrani B, Durand JC, Labaille F, Malet C, et al. Increased expression of prolactin receptor gene assessed by quantitative polymerase chain reaction in human breast tumors versus normal breast tissues. The Journal of clinical endocrinology and metabolism. 1998;83(2):667-74.
Bernard V, Young J, Chanson P, Binart N. New insights in prolactin: pathological implications. Nat Rev Endocrinol. 2015;11(5):265-75.
Li H, Ahonen TJ, Alanen K, Xie J, LeBaron MJ, Pretlow TG, et al. Activation of signal transducer and activator of transcription 5 in human prostate cancer is associated with high histological grade. Cancer Res. 2004;64(14):4774-82.
Li H, Zhang Y, Glass A, Zellweger T, Gehan E, Bubendorf L, et al. Activation of signal transducer and activator of transcription-5 in prostate cancer predicts early recurrence. Clin Cancer Res. 2005;11(16):5863-8.
Glasow A, Haidan A, Gillespie J, Kelly PA, Chrousos GP, Bornstein SR. Differential expression of prolactin receptor (PRLR) in normal and tumorous adrenal tissues: separation of cellular endocrine compartments by laser capture microdissection (LCM). Endocrine Research. 1998;24(3-4):857-62.

TableS1.xlsx
Supplementary Table S1: GO biological processes of genes in brown module.

Download PDF

Version 1

posted

You are reading this latest preprint version

BIRC5 and PRLR Might Be Novel Immune-Related Prognostic Biomarkers in Adrenocortical Carcinoma (ACC)

Status:

Version 1

Abstract

Figures

Background

Materials And Methods

Data collection

Data Preprocessing

Co-expression network construction

Identification of Clinically Significant Modules

Functional and Pathway Enrichment Analysis

Identification of hub genes

Validation of hub genes

Relationship between gene expression and TME

Gene Set Enrichment Analysis (GSEA)

Results

Co-expression network construction to identify clinically significant modules

GO Biological Processes and KEGG Analysis

Identification of hub genes

Hub Gene Validation

Relationship Between Gene Expression And TME

Gene Set Enrichment Analysis

Discussion

Conclusions

Abbreviations

Declarations

Ethics approval and consent to participate

Consent for publication

Availability of data and material

Competing interests

Funding

Author’s contributions

Acknowledgements

References

Supplementary Files

Status:

Version 1