OL Oncology Letters 1792-1074 1792-1082 D.A. Spandidos 10.3892/ol.2018.8306 OL-0-0-8306 Articles Distinct prognostic value of mRNA expression of guanylate-binding protein genes in skin cutaneous melanoma WangQiaoqi 1 WangXiangkun 2 LiangQian 1 WangShijun 3 XiwenLiao 2 PanFuqiang 1 ChenHongyang 1 LiDong 1 Cosmetic and Plastic Center, The Second Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi 530000, P.R. China Department of Hepatobiliary Surgery, The First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi 530021, P.R. China Department of Colorectal and Anal Surgery, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan 450000, P.R. China Correspondence to: Professor Dong Li, Cosmetic and Plastic Center, The Second Affiliated Hospital of Guangxi Medical University, 116 Daxue East Road, Nanning, Guangxi 530000, P.R. China, E-mail: ld_gxykdx@hotmail.com 05 2018 20 03 2018 15 5 7914 7922 21092017 14032018 Copyright: © Wang et al. 2018 This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

The purpose of the present study was to assess if guanylate-binding protein (GBP) mRNAs could be prognostic biomarkers for patients with skin cutaneous melanoma (SKCM). The prognostic value of GBP mRNA expression in patients with SKCM was investigated by analyzing gene expression data in 459 SKCM patients. The data were extracted from the OncoLnc database of The Cancer Genome Atlas. A high expression of GBP1, GBP2, GBP3, GBP4 and GBP5 were correlated with favorable overall survival (OS) in the SKCM patients followed for over 30 years. In addition, a high expression of GBP6 mRNA was not correlated with OS in the SKCM patients. A joint effects analysis showed that the co-incidence of the high expression of GBP1-5 was correlated with favorable overall survival in SKCM patients. Our findings suggest that GBP1-5 mRNAs in SKCM are associated with favorable prognosis and may be potential prognostic biomarkers. The combination of GBP1-5 could improve the sensitivity for predicting OS in SKCM patients.

 guanylate-binding protein mRNA correlation prognosis Introduction

Skin cutaneous melanoma (SKCM) is one of the most aggressive malignancies; tumors of millimeters are lethal. SKCM accounts for 91% of new cases of skin cancers and results in 74% of skin-related deaths (1). The incidence of SKCM has continued to increase in recent years, and it tends to affect younger people (1,2). The 5-year and 10-year relative survival rates for persons with SKCM are 92 and 89%, respectively. The primary treatment for SKCM is surgery combined with chemotherapy, immunotherapy and radiation (3). Effective prognosis markers may aid in therapeutic treatment for SKCM patients (4). Previous studies have shown that many genes, including AURKB, CCNE1, CDCA8, CDK4, CENPO, GINS2, H2AFZ, LIG1, PKMYT1, PLK1, PTTG1, SKA1, TUBA1B, TUBA1C, TYMS (5), and EZH2 (6), are associated with poor prognosis in SKCM. However, the association between GBP genes and the prognosis of SKCM has not been reported.

Guanylate-binding protein (GBP) belongs to the superfamily of INF-inducible guanosine triphosphate hydrolases (GTPases) (7,8). Up to now, seven human GBP genes, including guanylate-binding protein 1 (GBP1), guanylate-binding protein 2 (GBP2), guanylate-binding protein 3 (GBP3), guanylate-binding protein 4 (GBP4), guanylate-binding protein 5 (GBP5), guanylate-binding protein family member 6 (GBP6) and guanylate-binding protein 7 (GBP7), have been reported (9–11). GBPs, such as GBP1 and GBP2, have antiviral and antimicrobial activities in host defense (12) and could act as protective factors in host defense, controlling infection and autoimmunity (13).

The roles of GBP genes in cancers are complicated. Studies showed that some GBP family members were expressed in colorectal cancer (CRC) (14–17), breast cancer (18,19), oral squamous cell carcinoma (OSCC) (20), esophageal squamous cell carcinomas (SCC) (21), cutaneous T-cell lymphoma (22), prostate cancer (23), and Kaposi's sarcoma (24,25). GBP1 was upregulated in CRC (15) and OSCC (20), modulated the migration and invasion of OSCC cell in vitro (20), and inhibited the growth of highly malignant TS/A mammary carcinoma cells (19) and CRC tumors (14,15,17) in vivo. The high expression of GBP1 was associated with high overall pathological stage in OSCC tissue. GBP2 was related to T-cell infiltration in breast cancer (18).

The expression of GBP mRNAs is highly induced by interferon-γ (IFN-γ) in many cells including fibroblasts, B cells, T cells, and some tumor cells (15,24,26). GBP was also associated with the prognosis of many cancers (14,16–21,23,25). In addition, GBP1 plays dual roles in different tumor cells. Upregulated GBP1 mediated the anti-tumorigenic effects of IFN-γ and correlated with better OS in CRC (15,17). However, overexpressed GBP1 was significantly associated with poorer prognosis in OSCC patients (20). A high expression of GBP2 with a favorable prognosis was found in patients with node-negative breast carcinomas (18). However, the prognostic value of individual GBP in SKCM remains elusive. The present study investigated the prognostic value of individual GBP mRNA and made a joint effects analysis in 459 SKCM patients using OncoLnc data generated from The Cancer Genome Atlas (TCGA; https://cancergenome.nih.gov/, accessed March 1, 2017) database (27). Our results indicated that a high mRNA expression of individual GBP1-5 genes and a high co-expression of these gene mRNAs were correlated with high OS, suggesting that these genes may be potential prognosis biomarkers in SKCM patients.

 Materials and methods <sec> <title>Data preparation

TCGA survival data of SKCM was extracted from OncoLnc (http://www.oncolnc.org/, accessed March 3, 2017) (27), including the patients' ID in TCGA, sex, age at diagnosis, events, median survival, survival time, death status, and GBP members' mRNA expression regarding 459 SKCM patients. Briefly, 7 GBP sub-members (GBP1, GBP2, GBP3, GBP4, GBP5, GBP6, and GBP7) were entered into the database (http://www.oncolnc.org/, accessed by March 3, 2017). The patients were sorted into a percentile of 50:50 by the expression of every GBP sub-member, and then SKCM patients' survival data information was obtained.

The Metabolic gEne RApid Visualizer (MERAV: http://merav.wi.mit.edu/SearchByGenes.html, accessed March 1, 2017) (28) was used to make a boxplot of GBP sub-members' expression levels in normal tissue and primary tumors of skin cancer. After GBP genes and the selected tissue type were submitted on the website, boxplots were made and displayed. The unit for mRNA expression is counted in downloaded TCGA data.

 Correlation and bioinformatics analysis

The Pearson correlation coefficient was used to assess the co-expression of GBP genes. The relative expression levels of GBP genes in multiple normal tissues were determined with the GTEx Portal (http://www.gtexportal.org/home/, accessed April 25, 2017) (29). A gene function prediction website (GeneMANIA: http://genemania.org/, accessed March 15, 2017) (30) was also used to construct the gene-gene interaction networks. The Database for Annotation, Visualization, and Integrated Discovery (DAVID) v.6.7 (https://david.ncifcrf.gov/tools.jsp, accessed April 3, 2017) (31,32) was used to annotate input genes, classify gene functions, identify gene conversions, and carry out Gene Ontology (GO) term analysis (32). P<0.05 and a false discovery rate (FDR) <0.05 were considered to indicate a statistically significant difference.

 Survival analysis

A Kaplan-Meier estimator with a log-rank test was used to evaluate the correlation of six mRNAs with patient survival. Hazard ratios (HR) and 95% confidence intervals (CI) were used to assess the relative risk of SKCM survival.

 Joint effects analysis

A joint effects analysis was performed based on the survival analysis results. Patients were regrouped based on the combined GBP mRNA expression and OS scores, which were calculated by summarizing all of the points given to GBP1-5 in a patient when 1 point was assigned to genes of high expression with favorable OS and 0 points were assigned to genes of low expression with poor OS.

 Statistical analysis

Statistical analyses were carried out using SPSS v.22.0 software (IBM, Chicago, IL, USA). P<0.05 was considered to indicate a statistically significant difference.

 Results <sec> <title>mRNA expression of GBP genes in human normal skin and skin cancer tissues

The GBP family is composed of seven members. Among the seven GBP genes, only GBP7 was not found in www.oncolnc.org, likely due to its low expression (33). In human skin tissue, GBP2 and GBP6 were expressed at high levels, whereas the remaining GBP genes (GBP1, GBP3, GBP4 and GBP5) were expressed at low levels (Fig. 1A-F; Fig. 1A, GBP1; Fig. 1B, GBP2; Fig. 1C, GBP3; Fig. 1D, GBP4; Fig. 1E, GBP5; Fig. 1F, GBP6).

The boxplots of the GBP family generated from MERAV shows differences in the expression levels of GBP genes between normal skin tissue and primary skin tumor. The expressions of GBP1, GBP4 and GBP5 in normal skin tissue were higher than skin cancer. Moreover, the expressions of GBP2, GBP3 and GBP6 in skin cancer were higher than in normal skin tissue (Fig. 2).

 Functions and correlation of the mRNA expression of GBP genes in human tissues

A co-expression analysis (Fig. 3) showed that GBP1, GBP2, GBP3, GBP4 and GBP5 were co-expressed in human tissues. GBP3 was in the NFATC2 pathway, GBP2 was in the IFI35, IRF9 and XAF1 pathways, and GBP2 was predicted in the IRF1 pathway. The correlation of individual GBP family gene mRNA expression was tested using the Pearson correlation coefficient (Table I). With the exception of GBP6, the mRNA expression of all other GBP family genes was significantly (R=0.550–0.842, P<0.001) positively correlated (Table I). A GO term analysis using DAVID revealed that GBP genes were significantly associated with the biological process of immune response, as well as the molecular functions of GTPases activity and guanosine triphosphate (GTP) binding (Table II).

 Survival analysis

The prognostic value of the GBP family gene mRNA expression was assessed with SPSS. A high expression of GBP1-5 was significantly (P<0.001) associated with a favorable OS in SKCM patients (Fig. 4A-E). GBP6 expression did not show a significant correlation with OS in SKCM patients (P=0.401925, HR=1.121, 95% CI=0.8580–1.465) (Fig. 4F).

 Joint effects analysis

A joint effects analysis was used to determine the combined effect of the GBP gene mRNA co-expression on the OS of SKCM patients. Patients were divided into 6 groups: Group 1 (0 points group, n=136), group 2 (1 point group, n=60), group 3 (2 points group, n=32), group 4 (3 points group, n=39), group 5 (4 points group, n=47) and group 6 (5 points group, n=141) (detailed grouping information is shown in Table III). Kaplan-Meier estimator with a log-rank test was used to evaluate the prognostic value of these 6 groups. The co-overexpression of GBP1-5 in Group 6 (141 from 455) was found to be more highly correlated with a favorable OS than the co-overexpression of fewer GBP genes in other groups (P<0.0001). In contrast, the expression of GBPs was homogeneously low in Group 1 (136 from 455), which was found to be more highly correlated with poor OS than the other groups (P<0.0001) (Fig. 5).

 Discussion

In the present study, the data for the GBP gene mRNA expression and survival of SKCM patients were extracted from OncoLnc, analyzed to predict the function of GBP genes, and assessed for the potential of the mRNA expression of GBP genes to be used as prognosis biomarkers. Our analysis revealed that GBPs may be responsible for host defense, GTP binding and GTP hydrolysis. The correlation between the GBP gene mRNA levels and OS suggested that GBP mRNA may be good prognosis biomarkers for SKCM patients.

Our bioinformatics analysis revealed that the most meaningful molecular functions of GBP were GTPase activity, GTP binding, guanylate nucleotide binding, guanylate ribonucleotide binding, ribonucleotide binding, purine ribonucleotide binding, purine nucleotide binding, and nucleotide binding, which is in agreement with the observations that GBP belongs to the superfamily of INF-inducible GTPases including four sub-families: GBPs, immunity-related GTPases, very large inducible GTPase and myxovirus resistance proteins (7,8). The probable involvement of GBPs in the biological process of immunity deduced by our analysis is in agreement with the observations that GBPs, such as GBP1 and GBP2, have antiviral and antimicrobial activities in host defense (12) and that GBPs could act as protective factors in host defense, controlling infection and autoimmunity (13). The predicted immunity roles of GBP genes are also in agreement with the finding that upregulated GBP1 in CRC inhibits tumor growth (14,15,17) and that GBP2 was associated with T-cell infiltration in breast cancer (18).

In the present study, the Kaplan-Meier curves show that a high expression of GBP1-5 was found to be correlated with favorable OS in all SKCM patients. The correlation between a high expression of GBP1 and favorable OS in SKCM patients is in accordance with its correlation with high survival in CRC (17) but contrary to its correlation with poor survival in OSCC (20). These results indicated that GBP1 could play different roles in different cancers. The correlation of a high expression of GBP2 with a favorable OS in SKCM observed in the present study and in node-negative breast carcinomas (18) suggests that GBP2 may share the same mechanism in both SKCM and node-negative breast carcinomas. Though the expression of GBP6 in skin tumor tissue was more than its expression in normal tissue, no correlation between prognosis value with high expression of GBP6 or low expression of GBP6 was found. It is unclear why the high expression of downregulated GBP1, GBP4 and GBP5, as well as upregulated GBP2 and GBP3 showed the same correlation with a favorable OS in skin cancer. No survival information on GBP7 in SKCM patients is available, likely due to its low expression in normal skin tissue and SKCM, which makes it difficult to assess its correlation with prognosis outcomes.

The joint effects analysis showed that the co-expression of GBP1-5 all at high levels was correlated with a favorable OS in SKCM patients. In contrast, the co-expression of GBP1-5 at low levels was correlated with poor OS in SKCM patients. There was a tendency for GBP genes with higher expressions to be more highly correlated with favorable OS. The induced high co-expression of GBP1-5 in cells by IFN-γ (10,34) suggests that it may be possible to increase patients' favorable OS through the induction of a higher co-expression of GBP genes with IFN-γ. This hypothesis needs to be further investigated and experimentally proved. The combination of GBP1-5 may improve the sensitivity of predicting OS in SKCM patients.

There were limitations to the present study that should be recognized. First, since the data from the TCGA database and OncoLnc was not comprehensive, the present study evaluated the association between gene expression level and OS based on a log-rank test in Kaplan-Meier analysis. Second, the patients in the present study were exclusively from a single source, which meant that a multivariate analysis could not be used to validate the results. It is necessary to validate the prognostic value of these genes in patients with SKCM using independent external validation datasets containing complete clinical information. Despite these limitations, our current study was the first to report that the upregulation of the GBP genes (GBP1, GBP2, GBP3, GBP4 and GBP5) in SKCM was associated with a favorable prognosis. GBP1-5 may be used as prognostic biomarkers for SKCM patients.

In conclusion, a high expression of 5 GBP genes (GBP1, GBP2, GBP3, GBP4 and GBP5) was individually and coincidentally related to a favorable prognosis for SKCM. GBP1-5 may be used as potential prognostic biomarkers for SKCM patients. These results need to be confirmed in further studies.

 Acknowledgements

The authors would like to thank the contributors of OncoLnc for sharing the SKCM survival data on open access.

 Abbreviations GBP

guanylate-binding protein

 OS

overall survival

 SKCM

skin cutaneous melanoma

 References SiegelRLMillerKDJemalACancer statistics, 2016CA Cancer J Clin66730201610.3322/caac.2133226742998 SiegelRMaJZouZJemalACancer statistics, 2014CA Cancer J Clin64929201410.3322/caac.2120824399786 MillerKDSiegelRLLinCCMariottoABKramerJLRowlandJHSteinKDAlteriRJemalACancer treatment and survivorship statistics, 2016CA Cancer J Clin66271289201610.3322/caac.2134927253694 WeissSAHannifordDHernandoEOsmanIRevisiting determinants of prognosis in cutaneous melanomaCancer12141084123201510.1002/cncr.2963426308244 ZainulabadeenAYaoPZareHUnderexpression of specific interferon genes is associated with poor prognosis of melanomaPLoS One12e0170025201710.1371/journal.pone.017002528114321 TiffenJWilsonSGallagherSJHerseyPFilippFVSomatic copy number amplification and hyperactivating somatic mutations of EZH2 correlate with DNA methylation and drive epigenetic silencing of genes involved in tumor suppression and immune responses in melanomaNeoplasia18121132201610.1016/j.neo.2016.01.00326936398 KimBHShenoyARKumarPBradfieldCJMacMickingJDIFN-inducible GTPases in host cell defenseCell Host Microbe12432444201210.1016/j.chom.2012.09.00723084913 HotterDSauterDKirchhoffFGuanylate binding protein 5: Impairing virion infectivity by targeting retroviral envelope glycoproteinsSmall GTPases83137201710.1080/21541248.2016.118999027275775 GooYHSonSHYechoorVKPaulATranscriptional profiling of foam cells reveals induction of guanylate-binding proteins following western diet acceleration of atherosclerosis in the absence of global changes in inflammationJ Am Heart Assoc5e002663201610.1161/JAHA.115.00266327091181 TripalPBauerMNaschbergerEMörtingerTHohenadlCCornaliEThurauMStürzlMUnique features of different members of the human guanylate-binding protein familyJ Interferon Cytokine Res274452200710.1089/jir.2007.008617266443 DegrandiDKonermannCBeuter-GuniaCKresseAWürthnerJKurigSBeerSPfefferKExtensive characterization of IFN-induced GTPases mGBP1 to mGBP10 involved in host defenseJ Immunol17977297740200710.4049/jimmunol.179.11.772918025219 VestalDJJeyaratnamJAThe guanylate-binding proteins: Emerging insights into the biochemical properties and functions of this family of large interferon-induced guanosine triphosphataseJ Interferon Cytokine Res318997201110.1089/jir.2010.010221142871 ShenoyARWellingtonDAKumarPKassaHBoothCJCresswellPMacMickingJDGBP5 promotes NLRP3 inflammasome assembly and immunity in mammalsScience336481485201210.1126/science.121714122461501 FriedmanKBrodskyASLuSWoodSGillAJLombardoKYangDResnickMBMedullary carcinoma of the colon: A distinct morphology reveals a distinctive immunoregulatory microenvironmentMod Pathol29528541201610.1038/modpathol.2016.5426965581 Britzen-LaurentNLipnikKOckerMNaschbergerESchellererVSCronerRSViethMWaldnerMSteinbergPHohenadlCStürzlMGBP-1 acts as a tumor suppressor in colorectal cancer cellsCarcinogenesis34153162201310.1093/carcin/bgs31023042300 Britzen-LaurentNHerrmannCNaschbergerECronerRSStürzlMPathophysiological role of guanylate-binding proteins in gastrointestinal diseasesWorld J Gastroenterol2264346443201610.3748/wjg.v22.i28.643427605879 NaschbergerECronerRSMerkelSDimmlerATripalPAmannKUKremmerEBruecklWMPapadopoulosTHohenadlCAngiostatic immune reaction in colorectal carcinoma: Impact on survival and perspectives for antiangiogenic therapyInt J Cancer12321202129200810.1002/ijc.2376418697200 GodoyPCadenasCHellwigBMarchanRStewartJReifRLohrMGehrmannMRahnenführerJSchmidtMHengstlerJGInterferon-inducible guanylate binding protein (GBP2) is associated with better prognosis in breast cancer and indicates an efficient T cell responseBreast Cancer21491499201410.1007/s12282-012-0404-823001506 LipnikKNaschbergerEGonin-LaurentNKodajovaPPetznekHRungaldierSAstigianoSFerriniSStürzlMHohenadlCInterferon gamma-induced human guanylate binding protein 1 inhibits mammary tumor growth in miceMol Med16177187201010.2119/molmed.2009.0017220454519 YuCJChangKPChangYJHsuCWLiangYYuJSChiLMChangYSWuCCIdentification of guanylate-binding protein 1 as a potential oral cancer marker involved in cell invasion using omics-based analysisJ Proteome Res1037783788201110.1021/pr200413321714544 GuimarãesDPOliveiraIMde MoraesEPaivaGRSouzaDMBarnasCOlmedoDBPintoCEFariaPADe Moura GalloCVInterferon-inducible guanylate binding protein (GBP)-2: A novel p53-regulated tumor marker in esophageal squamous cell carcinomasInt J Cancer124272279200910.1002/ijc.2394419003964 FellenbergFHartmannTBDummerRUsenerDSchadendorfDEichmüllerSGBP-5 splicing variants: New guanylate-binding proteins with tumor-associated expression and antigenicityJ Invest Dermatol12215101517200410.1111/j.0022-202X.2004.22613.x15175044 PersanoLMoserleLEspositoGBronteVBarbieriVIafrateMGardimanMPLargheroPPfefferUNaschbergerEInterferon-alpha counteracts the angiogenic switch and reduces tumor cell proliferation in a spontaneous model of prostatic cancerCarcinogenesis30851860200910.1093/carcin/bgp05219237608 Lubeseder-MartellatoCGuenziEJörgATöpoltKNaschbergerEKremmerEZietzCTschachlerEHutzlerPSchwemmleMGuanylate-binding protein-1 expression is selectively induced by inflammatory cytokines and is an activation marker of endothelial cells during inflammatory diseasesAm J Pathol16117491759200210.1016/S0002-9440(10)64452-512414522 GuenziETöpoltKCornaliELubeseder-MartellatoCJörgAMatzenKZietzCKremmerENappiFSchwemmleMThe helical domain of GBP-1 mediates the inhibition of endothelial cell proliferation by inflammatory cytokinesEMBO J2055685577200110.1093/emboj/20.20.556811598000 ChengYSColonnoRJYinFHInterferon induction of fibroblast proteins with guanylate binding activityJ Biol Chem2587746775019836305951 AnayaJOncoLnc: Linking TCGA survival data to mRNAs, miRNAs, and lncRNAsPeer J Comput Sci2e67201610.7717/peerj-cs.67 ShaulYDYuanBThiruPNutter-UphamAMcCallumSLanzkronCBellGWSabatiniDMMERAV: A tool for comparing gene expression across human tissues and cell typesNucleic Acids Res44D560D566201610.1093/nar/gkv133726626150 CarithersLJArdlieKBarcusMBrantonPABrittonABuiaSAComptonCCDeLucaDSPeter-DemchokJGelfandETA novel approach to high-quality postmortem tissue procurement: The GTEx projectBiopreserv Biobank13311319201510.1089/bio.2015.003226484571 Warde-FarleyDDonaldsonSLComesOZuberiKBadrawiRChaoPFranzMGrouiosCKaziFLopesCTThe GeneMANIA prediction server: Biological network integration for gene prioritization and predicting gene functionNucleic Acids Res38(Web Server Issue)W214W220201010.1093/nar/gkq53720576703 Huang daWShermanBTLempickiRABioinformatics enrichment tools: Paths toward the comprehensive functional analysis of large gene listsNucleic Acids Res37113200910.1093/nar/gkn92319033363 Huang daWShermanBTLempickiRASystematic and integrative analysis of large gene lists using DAVID bioinformatics resourcesNat Protoc44457200910.1038/nprot.2008.21119131956 OlszewskiMAGrayJVestalDJIn silico genomic analysis of the human and murine guanylate-binding protein (GBP) gene clustersJ Interferon Cytokine Res26328352200610.1089/jir.2006.26.32816689661 Britzen-LaurentNBauerMBertonVFischerNSygudaAReipschlägerSNaschbergerEHerrmannCStürzlMIntracellular trafficking of guanylate-binding proteins is regulated by heterodimerization in a hierarchical mannerPLoS One5e14246201010.1371/journal.pone.001424621151871

GBP genes in multiple normal tissues in the GTEx Portal database. The expression of GBP genes in normal skin tissue was highlighted in red. (A) GBP1 gene expression in multiple normal tissues; (B) GBP2 gene expression in multiple normal tissues; (C) GBP3 gene expression in multiple normal tissues. GBP, guanylate-binding protein. (D) GBP4 gene expression in multiple normal tissues; (E) GBP5 gene expression in multiple normal tissues; (F) GBP6 gene expression in multiple normal tissues. GBP, guanylate-binding protein.

The MERAV boxplots of GBP family expression in skin normal tissue and primary tumor. (A) Boxplot for GBP1 expression; (B) boxplot for GBP2 expression; (C) boxplot for GBP3 expression; (D) boxplot for GBP4 expression; (E) boxplot for GBP5 expression; (F) boxplot for GBP6 expression. MERAV, Metabolic gEne RApid Visualizer; GBP, guanylate-binding protein.

Co-expression/pathway/predication analysis of GBP1, GBP2, GBP3, GBP4 and GBP5 according to human expression data in GeneMANIA. GBP, guanylate-binding protein.

The prognostic value of GBP expression. (A) survival curves are plotted for all SKCM patients of GBP1 (n=458); (B) survival curves are plotted for all SKCM patients of GBP2 (n=458); (C) survival curves are plotted for all SKCM patients of GBP3 (n=458); (D) survival curves are plotted for all SKCM patients of GBP4 (n=458); (E) survival curves are plotted for all SKCM patients of GBP5 (n=458); (F) survival curves are plotted for all SKCM patients of GBP6 (n=458). Data were analyzed using SPSS. GBP, guanylate-binding protein; SKCM, skin cutaneous melanoma.

The result of the joint effects analysis. OS stratified by 5 GBP genes expression levels. Group 1 (0 points group, n=136), Group 2 (1 point group, n=60), Group 3 (2 points group, n=32), Group 4 (3 points group, n=39), Group 5 (4 points group, n=47) and Group 6 (5 points group, n=141). GBP, guanylate-binding protein.

Co-expression of GBP family at mRNA level.

GBP1 GBP2 GBP3 GBP4 GBP5 GBP6
Genes R P-value R P-value R P-value R P-value R P-value R P-value
GBP1 – – 0.842 <0.001 0.743 <0.001 0.670 <0.001 0.702 <0.001 0.000 0.995
GBP2 0.842 <0.001 – – 0.685 <0.001 0.643 <0.001 0.654 <0.001 0.047 0.313
GBP3 0.743 <0.001 0.685 <0.001 – – 0.550 <0.001 0.594 <0.001 0.022 0.636
GBP4 0.670 <0.001 0.643 <0.001 0.550 <0.001 – – 0.768 <0.001 −0.013 0.776
GBP5 0.702 <0.001 0.654 <0.001 0.594 <0.001 0.768 <0.001 – – 0.001 0.990
GBP6 0.000 0.995 0.047 0.313 0.022 0.636 −0.013 0.776 0.001 0.990 – –

The correlations of gene mRNA expression in GBP families were tested using the Pearson correlation coefficient. R, Pearson correlation coefficient; GBP, guanylate-binding protein.

Analysis of enriched GO terms for GBP genes carried out using DAVID.

Category Term Genes P-value FDR
Molecular function GTPase activity 6 1.08×10−09 5.58×10−07
GTP binding 6 1.88×10−08 9.70×10−06
Guanylate nucleotide binding 6 2.15×10−08 1.11×10−05
Guanylate ribonucleotide binding 6 2.15×10−08 1.11×10−05
Ribonucleotide binding 6 5.63×10−05 0.029018
Purine ribonucleotide binding 6 5.63×10−05 0.029018
Purine nucleotide binding 6 7.01×10−05 0.03611
Nucleotide binding 6 1.54×10−04 0.079382
Biological process Immune response 6 3.40×10−07 7.69×10−05
Cellular component Internal side of plasma membrane 3 0.001798 –
Plasma membrane part 3 0.078854 –
Plasma membrane 3 0.210327 –

GBP, guanylate-binding protein; GO Gene Ontology; DAVID, Database for Annotation, Visualization, and Integrated Discovery; GTPases, guanosine triphosphates. GTP, guanosine triphosphate; FDR, false discovery rate.

Grouping information for the combination among GBP genes.

Group Points Composition
1 0 Low GBP1+Low GBP2+Low GBP3+Low GBP4+Low GBP5
2 1 High GBP1+Low GBP2+Low GBP3+Low GBP4+Low GBP5
2 1 Low GBP1+High GBP2+Low GBP3+Low GBP4+Low GBP5
2 1 Low GBP1+Low GBP2+High GBP3+Low GBP4+Low GBP5
2 1 Low GBP1+Low GBP2+Low GBP3+High GBP4+Low GBP5
2 1 Low GBP1+Low GBP2+Low GBP3+Low GBP4+High GBP5
3 2 High GBP1+High GBP2+Low GBP3+Low GBP4+Low GBP5
3 2 High GBP1+Low GBP2+High GBP3+Low GBP4+Low GBP5
3 2 High GBP1+Low GBP2+Low GBP3+High GBP4+Low GBP5
3 2 High GBP1+Low GBP2+Low GBP3+Low GBP4+High GBP5
3 2 Low GBP1+High GBP2+High GBP3+Low GBP4+Low GBP5
3 2 Low GBP1+High GBP2+Low GBP3+High GBP4+Low GBP5
3 2 Low GBP1+High GBP3+Low GBP3+Low GBP4 +High GBP5
3 2 Low GBP1+Low GBP2+High GBP3+High GBP4+Low GBP5
3 2 Low GBP1+Low GBP2+High GBP3+Low GBP4+High GBP5
3 2 Low GBP1+Low GBP2+Low GBP3+High GBP4+High GBP5
4 3 High GBP1+High GBP2+High GBP3+Low GBP4+Low GBP5
4 3 High GBP1+High GBP2+Low GBP3+High GBP4+Low GBP5
4 3 High GBP1+High GBP2+Low GBP3+Low GBP4+High GBP5
4 3 High GBP1+Low GBP2+High GBP3+High GBP4+Low GBP5
4 3 High GBP1+Low GBP2+High GBP3+Low GBP4+High GBP5
4 3 High GBP1+Low GBP2+Low GBP3+High GBP4+High GBP5
4 3 Low GBP1+High GBP2+High GBP3+High GBP4+Low GBP5
4 3 Low GBP1+High GBP2+High GBP3+Low GBP4+High GBP5
4 3 Low GBP1+High GBP2+Low GBP3+High GBP4+High GBP5
4 3 Low GBP1+Low GBP2+High GBP3+High GBP4+High GBP5
5 4 High GBP1+High GBP2+High GBP3+High GBP4+Low GBP5
5 4 High GBP1+High GBP2+High GBP3+Low GBP4+High GBP5
5 4 High GBP1+High GBP2+Low GBP3+High GBP4+High GBP5
5 4 High GBP1+Low GBP2+High GBP3+High GBP4+High GBP5
5 4 Low GBP1+High GBP2+High GBP3+High GBP4+High GBP5
6 5 High GBP1+High GBP2+High GBP3+High GBP4+High GBP5

With the median value of the gene expression as cutoff, the patients were designated as high expression or low expression for every member of GBP family, and grouped based on the combination of the gene expression levels. Group 1 (all low expression genes, 0 points group, n=136), group 2 (1 high expression gene, 1 point group, n=60), group 3 (2 high expression genes, 2 points group, n=32), group 4 (3 high expression genes, 3 points group, n=39), group 5 (4 high expression genes, 4 points group, n=47), group 6 (all high expression genes, 5 points group, n=141). GBP, guanylate-binding protein.