Immune regulation and prognosis indicating ability of a newly constructed multi-genes containing signature in clear cell renal cell carcinoma

Ziwei Gui; Juan Du; Nan Wu; Ningning Shen; Zhiqing Yang; Huijun Yang; Xuzhi Wang; Na Zhao; Zixin Zeng; Rong Wei; Wenxia Ma; Chen Wang

doi:10.1186/s12885-023-11150-4

Immune regulation and prognosis indicating ability of a newly constructed multi-genes containing signature in clear cell renal cell carcinoma

Gui, Ziwei; Du, Juan; Wu, Nan; Shen, Ningning; Yang, Zhiqing; Yang, Huijun; Wang, Xuzhi; Zhao, Na; Zeng, Zixin; Wei, Rong; Ma, Wenxia; Wang, Chen 2023-07-12 00:00:00 Background Clear cell renal cell carcinoma (ccRCC) is the most common renal malignancy, although newly develop- ing targeted therapy and immunotherapy have been showing promising effects in clinical treatment, the effective biomarkers for immune response prediction are still lacking. The study is to construct a gene signature according to ccRCC immune cells infiltration landscape, thus aiding clinical prediction of patients response to immunotherapy. Methods Firstly, ccRCC transcriptome expression profiles from Gene Expression Omnibus (GEO) database as well as immune related genes information from IMMPORT database were combine applied to identify the differently expressed meanwhile immune related candidate genes in ccRCC comparing to normal control samples. Then, based on protein–protein interaction network (PPI) and following module analysis of the candidate genes, a hub gene cluster was further identified for survival analysis. Further, LASSO analysis was applied to construct a signature which was in succession assessed with Kaplan–Meier survival, Cox regression and ROC curve analysis. Moreover, ccRCC patients were divided as high and low-risk groups based on the gene signature followed by the difference estimation of immune treatment response and exploration of related immune cells infiltration by TIDE and Cibersort analysis respectively among the two groups of patients. Results Based on GEO and IMMPORT databases, a total of 269 differently expressed meanwhile immune related genes in ccRCC were identified, further PPI network and module analysis of the 269 genes highlighted a 46 genes cluster. Next step, Kaplan–Meier and Cox regression analysis of the 46 genes identified 4 genes that were supported to be independent prognosis indicators, and a gene signature was constructed based on the 4 genes. Furthermore, after assessing its prognosis indicating ability by both Kaplan–Meier and Cox regression analysis, immune relation of the signature was evaluated including its association with environment immune score, Immune checkpoint inhibi- tors expression as well as immune cells infiltration. Together, immune predicting ability of the signature was prelimi- nary explored. Ziwei Gui, Juan Du and Nan Wu contributed equally to this work. *Correspondence: Wenxia Ma mawenxia@sxmu.edu.cn Chen Wang wangchen@sxmu.edu.cn Full list of author information is available at the end of the article © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Gui et al. BMC Cancer (2023) 23:649 Page 2 of 24 Conclusions Based on ccRCC genes expression profiles and multiple bioinformatic analysis, a 4 genes containing signature was constructed and the immune regulation of the signature was preliminary explored. Although more detailed experiments and clinical trials are needed before potential clinical use of the signature, the results shall pro- vide meaningful insight into further ccRCC immune researches. Keywords Clear cell renal cell carcinoma (ccRCC), Immune response, LASSO analysis, Gene signature, Prediction biomarker Background yet. For example, clinical trial Checkmate025 showed Renal cell carcinoma has been the most common kidney that advanced ccRCC patients benefited from immuno - malignancy which comprises ccRCC and non clear cell therapy regardless of PD-L1 expression [18, 19]. Check- Renal Cell Carcinoma (nccRCC), and ccRCC accounts mate214 revealed that the objective remission rate (ORR) for approximately 70% ~ 75% of all the cases [1, 2]. Attrib- in immunotherapy received group was higher than that uting to the rapid development of molecular pathology, in sunitinib treatment group no matter PD-L1 expres- the genome mechanism behind ccRCC occurence has sion was higher or less than 1% [20, 21]. Meanwhile, Key- been gradual clear, of which short arm of chromosome note426 results also revealed that both PD-L1 negative 3 (3p) genes variations were showing defining character - and positive ccRCC patients benefited from pablizumab istic roles involving most importantly VHL gene known and axitinib combination therapy [22, 23]. by symbolic “double hit”. The aberrant change of VHL Meanwhile, as for the use of TMB in renal malignancy, including gene mutation and promoter methylation Robert M Samstein.et al. reported an pan-cancer MSK- causes the “first hit”, followed by "second hit", namely the IMPACT genome sequencing analysis based on 1600 3p chromosome deletion leading to tumor occurrence cases of cancer samples, of which 151 cases were RCC [3–5]. Besides VHL, other 3p gene variations were also samples, and the results showed that TMB was non sig- reported in ccRCC, for instance SETD2 [6], BAP1 [7] and nificant related with ccRCC overall survival [24]. And PBRM1 [8], which were reported to be survival related. since microsatellite instability (MSI-H and dMMR) is Regarding the clinical treatment, over the past two dec- very rare in RCC patients, MSI is not supported by evi- ades, molecular targeted therapies and immune therapies dence-based medicine yet to be an effective immune bio - have been showing great potential. Currently, at least 13 marker [14, 25]. drugs in 6 categories have been approved for metastatic Above all, ICIs has been an promising treatment in ccRCC, including VEGFR, mTORC1, c-Met, FGFR inhi- clinical ccRCC, but the effective biomarkers for immune bition, cytokines, and most recently anti PD-1/PD-L1 response prediction are still lacking, it is of great impor- immune checkpoint inhibitors (ICIs) which has been tance to keep exploring ccRCC genome and identifying a promising pillar of nowadays clinical treatment [4, 9]. new potential immune response biomarkers thus aiding Multi clinical trials most notably KEYNOTE-564 results more precise understanding of the disease and shed- supported ccRCC as immune sensitive and demonstrated ding promising light on further clinical immunotherapy the efficiency of ICIs in advanced patients clinical treat - application. ment using independent ICI therapies or ICI + TKI com- In the study, multi online ccRCC transcriptome pro- bination therapies [10]. files, our local hospital patients samples as well as vari - Although evidence-based medicine supported RCC ous bioinformatic analysis tools were combine used to as one of the malignancies that could benefit from neo- explore ccRCC genome data, identifying the survival adjuvant PD-1/PD-L1 blockade [11–13], the immune related meanwhile immune regulation associated genes response obviously vary among individuals indicating the and constructing a potential immune signature, fur- importance of usable and effective immune biomarkers ther necessary signaling mechanism was preliminary for selecting the potential patients that were most likely explored. The results shall provide meaningful insights to be able to benefit from the treatment [14]. Currently, to the unearth of potential new immune biomarkers PD-L1 expression, MSI and tumor mutation burden and shed promising light on further ccRCC immune (TMB) were three most widely used biomarkers in other researches. types of cancers, however, their use in ccRCC are still in dispute. Materials and methods As for PD-L1 expression, although it has been an effec - Data source: ccRCC transcriptome data from GEO database tive biomarker in other cancers for predicting immune From GEO online database, we widely screened ccRCC response [15–17], it’s function in ccRCC is inconclusive related profiles for exploring the changed genome Gui et al. BMC Cancer (2023) 23:649 Page 3 of 24 information in cancer comparing to normal renal sam- is short for Search Tool for the Retrieval of Interact- ples. The selection criteria of GEO profiles were set as: ing Genes was applied to construct the PPI network 1. the information of profiles were based on human tis - of above selected genes for observing the interaction sues (not animal models of any species); 2. the samples between individual genes. Further, based on PPI net- type was solid tissue (not tumor cell lines); 3. the contain- work, Molecular Complex Detection (MCODE) function ing data were mRNA/ cDNA/ transcriptome sequencing of Cytoscape3.6.0 software [33] was used to analyze the data; 4. covering both ccRCC cancer and normal renal promising function modules (gene clusters sharing simi- control samples; 5. each profile should contain at least 40 lar function) from the gene nest. or more samples. Further, Gene ontology analysis (GO) and Kyoto Ency- Based on above selection criteria, a total of five clopedia of Genes and Genomes (KEGG) [34] were ccRCC cDNA expression profiles were selected, includ - used to annotate basic biological attributes of the list of ing four profiles namely GSE53000 [26], GSE53757 [27], genes in each module including their main cellular loca- GSE68417 [28] and GSE71963 [29] containing a total of tion, involved biological processes, molecular functions 186 cases of ccRCC samples and 108 normal kidney sam- and the signaling pathways they mainly enriched in. The ples were selected for further potential genes selection. module that was predicted to be most related with tumor Meanwhile, another datasets GSE22541 which includes immune modulation and possess the highest module 68 cases of ccRCC samples was applied as validating data score would be highly focused and identified as a poten - source. (Table S1 for detailed information of the profiles tial gene cluster for further analysis. including samples amount, contributors and accessed online website). Univariate survival combine with Cox regression analysis of immune related gene cluster for hub genes Data processing: identify the differently expressed Following the identification of the immune related gene meanwhile immune related genes in ccRCC comparing cluster, each gene in the module would be in succession to normal control brought for firstly univariate survival analysis by UAL - The GEO transcriptome data were used to 1. explore CAN [35] and GEPIA [36], which have been two effec - the differently expressed genes in ccRCC comparing to tive online services for survival analysis. Then, the genes normal kidney samples; 2. combine with IMMPORT that were supported by both univariate analysis methods immune database [30] for collaborate identify the differ - to be statistical significantly associated with ccRCC sur - ently expressed meanwhile immune regulation related vival would be processed for multivariate COX regres- genes. sion analysis based on TCGA ccRCC data using SPSS19.0 To reveal the aberrant differently expressed genes in analysis. Any gene that was indicated by all three analysis ccRCC comparing to normal control samples, four GEO to be associating with patients survival would be identi- profiles GSE53000, GSE53757, GSE68417 and GSE71963 fied as credible prognostic indicating hub genes and pro - were in succession analyzed with GEO2R which was pro- cessed for next step interpretation. vided pared with each GEO profile. The analysis criteria was set as adjusted P value < 0.05 meanwhile |log2FC|< 1, Estimation of hub genes’ physicochemical properties 1 ≤|log2FC|< 2, 2 ≤|log2FC|< 3 and |log2FC|≥ 3 respec- For understanding the basic information of the selected tively, thus the genes expression change distribution hub genes, ProtParam [37], ProtScale [38], and Human namely the genes that were < twofold, 2 ~ fourfold, Protein Atlas [39] were combine used. As for the physico- 4 ~ eightfold and > eightfold different in cancer versus chemical properties of genes, ProtParam and ProtScale normal control would be preliminary understood. were applied to understand the basic information of the Then, Venn diagram [31] would be used to identify the genes encoding proteins including the aminoacid compo- immune related genes from all the high level differently sition, estimated molecular weight and protein half life, expressed genes based on IMMPORT immune genes list, computed protein instability index and theoretical isoe- therefore, the genes that were preliminary supported to lectric point, as well as hydrophobicity and hydrophilicity be both aberrant changed expressed and immune regula- of proteins. tion related were selected as candidate genes for further Besides ProtParam and ProtScale, Human Protein Atlas analysis. is also an effective and well used online service for inter - preting certain proteins information, in the study, it was Protein–protein interaction (PPI) network construction applied to predict the cellular location of the selected hub and function module analysis of the candidate genes genes for the convenience of further clinical test. After identifying the differently expressed meanwhile Additionally, UALCAN as well as GEPIA, which immune related candidate genes, STRING [32], which have been two resourceful web services constructed Gui et al. BMC Cancer (2023) 23:649 Page 4 of 24 Association between the selected hub genes expression based on TCGA and GTEx programs were in succes- and ccRCC clinical pathological features sion accessed to observe the expression difference of Ualcan has been a widely used integrated data-mining hub genes in broad-spectrum human cancers compar- platform for analyzing cancer transcriptome data, besides ing to corresponding normal control samples, espe- previous analysis of the expression difference of hub cially in ccRCC versus normal renal tissues. genes in broad-spectrum human cancers comparing to corresponding normal control samples, UALCAN was in addition applied to analyze the association between hub Quantitative real‑time PCR (QPCR) experiment genes’ expression and ccRCC clinical parameters includ- for evaluating the expression change of selected hub ing patients age, gender, tumor grade, stage and lymph genes in cancer vs normal tissues node metastasis, aiming to aiding the better understand- Besides above UALCAN as well as GEPIA online anal- ing of the potential biological roles of selected hub genes ysis, 30 pairs of ccRCC cancer tissues and adjacent in ccRCC. paracancerous normal renal tissues which were all col- lected from our local hospital were used for validating the expression changes of selected hub genes. All the patients tissues were collected from surgeries at local Other genetic alterations of the selected hub genes as well hospital General Surgery Department and sent for as their potential related signaling pathways analysis pathology examination then being long-term stored Besides the mRNA expression difference as well as at Pathology Department Biobank. The Informed con- association with clinical pathological parameters, sent from the patients as well as the approval by the the PPI networks which were centered on each of the Hospital Institutional Board were both obtained (Sec- four selected hub genes were constructed in succes- ond Hospital of ShanXi Medical University, China). sion, for the purpose of aiding better understanding The mRNA of 30 pairs of ccRCC cancer and adjacent of hub genes’ potential biological roles as well as their normal renal tissues were extracted using RNAiso-Plus related signaling pathways by KEGG analysis in ccRCC (TAKARA, DaLian, China). And then1 μg extracted development [40, 41]. mRNA was used for cDNA synthesis using cDNA Moreover, other types of variations of the selected key synthesis kit (TAKARA, DaLian, China) following genes including mutation ratio, copy number variation, operating instruction. Further, qPCR was performed amplification and deletion ratio were explored based on on Roche Light Cycler z 480 and the primers of the cBioPortal database which has been an effective cancer tested hub genes used during the process were listed genomics data website covering more than 2,8000 cancer as below: samples. After logging into the cBioPortal website, the MMP9: “cancer types summary” module of “quick search” section Former: AGA CCT GGG CAG ATT CCA AAC was used for exploring the genetic alteration character- Reverse: CGG CAA GTC TTC CGA GTA GT istics of previous selected hub genes in various cancer NFKB1: types, especially ccRCC. Former: AGC ACG ACA ACA TCT CAT T Reverse: CAG GCA CAA CTC CTT CAT IRF7: Construction and clinical features analysis of an prognosis Former: CCC ACG CTA TAC CAT CTA CCT related immune gene signature Reverse: GAT GTC GTC ATA GAG GCT GTTG To maximum the clinical utilization of hub gene indi- HMOX1: cators that were selected based on above processes, an Former: TGC CAG TGC CAC CAA GTT CAAG prognosis-related immune signature was constructed Reverse: TGT TGA GCA GGA ACG CAG TCTTG using LASSO algorithm performed with glmnet R pack- GAPDH: age based on TCGA ccRCC data, thus an unique regres- Former: AGA AGG CTG GGG CTC ATT TG sion coefficient was assigned to each gene indicator Reverse: AGG GGC CAT CCA CAG TCT TC which multiplies the gene expression. Based on the final The PCR cycling condition was set as: 95 °C 10 min score of each case calculated according to the gene signa- for 1 cycle; 95 °C 10 s, 58 °C 30 s, and 72 °C 34 s for ture, ccRCC patients were classified as high-risk and low- 35 cycles followed by the melting curve stage. And the risk groups (median score was set as cut off value). relative gene expression in each sample was recorded as Further, the clinical features of high-risk and low-risk the average 2^ − ΔΔCT calculation result of three rep- groups of patients were analyzed based on TCGA data licates. Further, T-test was used for detailed statistical which contains resourceful ccRCC samples, despite the analysis. P < 0.05 was considered statistically significant. censoring data, an effective pool of over 533 cases of Gui et al. BMC Cancer (2023) 23:649 Page 5 of 24 Association between the gene signature and ccRCC ccRCC patients information were applied for preliminary estimated environment immune score observing the clinical features association of the con- ESTIMATE, which is short for Estimation of Stromal structed signature. and Immune cells in Malignant Tumors using Expression data and has been a well accepted cancer immune evalu- Preliminary prognosis validation of the gene signature ation tool in multiple cancers was applied to estimate the To validate the survival relationship of last step con- immune score of ccRCC samples based on TCGA genes structed gene signature, a series of methods were com- expression data. The correlation between the signature bine used to analyze the TCGA ccRCC data including and ESTIMATE algorithm based ccRCC immune score, firstly Kaplan–Meier survival which was used to com - stromal score as well as tumor purity were evaluated pare the survival difference between high-risk and low- using R package for aiding the validation of immune rela- risk groups of patients, then AUC curve was performed tionship of the constructed gene signature. to observe the 1, 3 and 5 years survival prediction ability of the signature. Further, univariate as well as multivari- Correlation between gene signature and the expression ate Cox regression analysis were applied for testing the of immune checkpoint inhibitors independence survival prediction ability of the signa- Immune checkpoints have been showing inspiring drug ture together with other well accepted prognosis related targeting effects in multiple cancers by reversing the clinical parameters. Furthermore, a nomogram combin- tumor immuno suppressive microenvironment, and ing the signature and these validated clinical parameters expression of immune checkpoints especially PD-L1, was constructed for together evaluating clinical ccRCC CTLA4, TIGIT, TIM-3 and LAG-3 have been well patients prognosis, all based on the TCGA over 533 cases accepted as clinical biomarkers for selecting potential of ccRCC patients information. cancer patients that were most likely to benefit from Moreover, an independent GEO profile different from immunotherapy. Therefore, the association between the the four profiles that were used to identify the hub genes gene signature and expression level of these immune and construct the gene signature, namely GSE22541 was checkpoints in TCGA ccRCC samples were assessed in additionally applied to perform the Kaplan–Meier sur- the study, as well as the comparison of expression differ - vival as well as ROC curve analysis, for the purpose of ence of these immune checkpoints between high-risk and validating the prognosis correlation of the gene signature. low-risk groups of ccRCC patients. Gene set enrichment analysis (GSEA) of two ccRCC patients Evaluation of relationships between gene signature and 22 subgroups based on gene signature tumor infiltrating immune cells (TICs) After preliminary prognosis relationship analysis, the For characterizing the microenvironment immune land- gene signature was next step processed for further scape between high-risk and low-risk groups of ccRCC immune association evaluating. Based on the constructed patients, CIBERSORT algorithm [44] was performed gene signature, TCGA ccRCC patients were divided as to calculate the relative contents of 22 TICs based on high-risk and low-risk subgroups, for determining how TCGA profiles data, followed by analyzing the relation - the immunological pathways and corresponding immune ship between the 22 TICs and gene signature. Further, genes differ between the two ccRCC subgroups, GSEA the survival analysis of 22 TICs especially the ones that [42] was performed for signaling enrichment analysis, relates with gene signature were conducted for identi- and the threshold was set as P < 0.05. fying the specific immune cell infiltration that impacts patients prognosis. Difference of immunogenic cell death (ICD) between high‑risk and low‑risk groups of ccRCC patients Statistical analysis ICD has been gradually accepted as a form of regulated Statistical analysis was performed using SPSS 26.0. biological cell death meanwhile supported by evidence- For the enumeration data including QPCR experiment based medicine to be able to trigger cellular adaptive revealing the expression of different genes in ccRCC can - immune response through the emission of damage asso- cers vs normal control samples, the data were analyzed ciated molecular patterns (DAMPs), thus potentially using t-test. As for the measurement data for instance contributing to clinical immunotherapy. In the study, the association between constructed gene signature and the expression distribution of 32 ICD related genes [43] ccRCC clinical parameters, the data were analyzed by which were identified based on literature studies were χ test. And for the correlation analysis, for instance the explored in high-risk and low-risk groups of ccRCC correlation between gene signature and immune check- patients for preliminary evaluating the immune status points expression, the data were analyzed by Spearman difference between the two groups patients. Gui et al. BMC Cancer (2023) 23:649 Page 6 of 24 analysis. Meanwhile, for the survival analysis, Kaplan– promising gene clusters. The first gene cluster posses Meier were performed. p < 0.05 was considered statisti- the highest computed module score and contains 46 cally significant. genes with a big portion of them predicted to be related with immune system modulation (Fig. 2B). Meanwhile, Results the second and third gene modules contain 25 and 29 ccRCC transcriptome data identified 269 high level genes respectively, and genes were mostly related with differently expressed meanwhile immune related genes CXCR4, PI3K, EGF and mTOR related signaling path- in cancer versus normal renal samples ways (Fig. 2C, D). Four GEO profiles GSE53000, GSE53757, GSE68417 and Given the first module gene cluster possess the high - GSE71963 were combine applied to explore the aber- est score and a big percentage of containing genes were rant differently expressed genes in ccRCC comparing to immune system related which shows more potential for normal renal samples. And in GSE53000, a total of 5559 further clinical immune indicators selection, the 46 genes genes were identified to be differently expressed includ - in the gene cluster were mainly focused for next step ing 4270 genes with the expression change ≤ twofold, analysis. 1028 genes that were 2 ~ fourfold, 180 genes that were 4 ~ eightfold and 81 genes whose expression were > eight- Kaplan–Meier combine with Cox regression analysis fold in ccRCC comparing to normal renal samples of cluster genes identified 4 ccRCC prognosis related hub (Fig. 1A). And in GSE53757, a total of 28021 genes were genes identified, and the number was 21367, 4857, 1195 and Genes should be of more potential if they were both 602 in ≤ twofold, 2 ~ fourfold, 4 ~ eightfold and > eight- immune regulation and survival related, and promising fold groups respectively (Fig. 1B). In GSE68417, a total of immune biomarkers should also be prognosis related for 10150 genes were identified, and the number was 8276, potential clinical medical drug targeting use. To further 1425, 290 and 159 in each group (Fig. 1C). Meanwhile, in analyze survival relationship of the 46 selected immune GSE71963, a total of 10744 genes were identified, and the related candidate genes, univariate survival analysis number was 6266, 3120, 841 and 517 genes in each group including UALCAN and GEPIA, as well as multivariate respectively (Fig. 1D, Table S2). Cox regression analysis were in succession performed, Considering the feasibility of further clinical medical and the results supported four genes, namely MMP9 use, we mainly focused on the high differently expressed (Fig. 3A), NFKB1 (Fig. 3B), IRF7 (Fig. 3C) and HMOX1 genes (at least > fourfold in cancer vs. normal). As a (Fig. 3D) as independent prognostic indicators in ccRCC, result, besides the genes that were shared in multiple all four genes not only relate with patients overall sur- profiles, the analysis of 4 GEO profiles indicated a total vival but also progress free survival indicating their high of 3095 genes that were high level changed expressed in potential in clinical medical use (Table 1). ccRCC comparing to normal control samples (Fig. 1E). Further, the immune related genes list was obtained from Basic physicochemical properties of the 4 selected hub IMMPORT immune database, and Venn diagram analy- genes sis result identified 269 genes from the 3095 genes that Basic physiochemical properties of MMP9, NFKB1, IRF7 were both high level expression changed and immune and HMOX1 were preliminary interpreted before deeper regulation related for next step analysis (Fig. 1F, detailed scientific use of them (Table 2). As for MMP9, which is 269 genes information was listed in Table S3). a member of matrix metalloproteinase (MMP) family, locates in 20q13.12 and encodes a protein composed of PPI network of 269 genes highlighted a 46 707 amino acids with an estimated molecular weight of genes‑containing immune relating gene cluster 78.5KD. The theoretical isoelectric point of the protein The PPI network of 269 differently expressed mean - is estimated to be 5.69 and instability index to be 41.10, while potentially immune related genes was constructed meanwhile, the grand average of hydrophobic value of (Fig. 2A), and based on the network we identified three the protein is -0.394 indicating MMP9 works as a cellular (See figure on next page.) Fig. 1 Differential expression genes in ccRCC comparing to normal renal samples identified from GEO profiles. Four GEO profiles (A) GSE53000, (B) GSE53757, (C) GSE68417 and (D) GSE71963 were accessed to identify differently expressed genes in ccRCC vs normal renal samples, and based on these profiles, the up-regulated (right side) and down-regulated (left side) differential expression genes in ccRCC were identified. The genes were then divided into four groups based on the expression difference level as: < twofold genes (orange-colored spots), 2 ~ fourfold genes (red-colored spots), 4 ~ eightfold genes (green-colored spots) and > eightfold genes (black-colored spots). E The intersection of the genes in four GEO profiles for revealing the genes that were shared in different profiles. F The intersection of differential expressed genes revealed by GEO profiles and immune related genes from IMMPORT database, thus the genes that were both differential expressed and immune related were revealed Gui et al. BMC Cancer (2023) 23:649 Page 7 of 24 Fig. 1 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 8 of 24 unstable and hydrophilic protein which locates in cellular associated with the development of heme oxygenase1 cytoplasm or to be secreted in the extracellular region, deficiency and pulmonary disease, as well as chronic and the related signaling pathways include collagen chain obstructive (Fig. 3H). trimerization and apoptotic pathways in synovial fibro - blasts (Fig. 3E). Validation of the changed expression of selected hub NFKB1, which is short for Nuclear Factor Kappa B genes in ccRCC versus normal renal tissues Subunit 1 locates in 4q24 and encodes a protein com- Although the four selected hub genes were obtained from posed of 968 amino acids and estimated to be weighting the differently expressed gene clusters analyzed based on 105KD with computed theoretical isoelectric point as GEO data from the beginning, after preliminary inter- 5.20 and instability index as 38.15. Meanwhile, the grand pretation of the basic physicochemical information, it’s average of hydrophobic value of protein is -0.339 indicat- necessary to validate each of the gene’s aberrant changed ing NFKB1 to be cellular stable and hydrophilic. NFKB1 expression in ccRCC comparing to normal renal samples is predicted to locates in nucleoplasm and cytoplasm, it individually. In the study, both online analysis as well has been reported as a transcription regulator that could local hospital samples were used for detecting the genes be activated by various cellular stimuli such as cytokines, expression level. ultraviolet irradiation, and bacterial or viral products. Firstly, two analysis databases including UALCAN Activated NFKB1 translocates into cell nucleus and stim- and GEPIA were used, and the results revealed that as ulates the expression of genes involved in various biologi- for MMP9 and IRF7 genes, they gain of expression in cal functions (Fig. 3F). most of human cancers (Fig. 4A, D). And as for NFKB1 IRF7 is short for Interferon Regulatory Factor 7 and it’s and HMOX1, their expression vary in different cancers a member of the interferon regulatory factor (IRF) family, (Fig. 4G, J), although in ccRCC, all four genes were indi- locating in 11p15.5 and encoding a protein composed of cated to be statistical significantly up regulated in cancers 503 amino acids including 56 negatively charged amino comparing to normal renal tissues (Fig. 4B, E, H, K). acid residues (ASP + Glu) and 49 positively charged Then, the result of qRT-PCR experiment which was amino acid residues (Arg + Lys). The estimated protein conducted using 30 local hospital ccRCC and paired molecular weight is 54.2KD with theoretical isoelectric normal renal tissues also supported the aberrant gain point computed to be 5.89. Meanwhile, the estimated of expression of all four genes including MMP9, IRF7, instability index of the protein is 63.17 and grand aver- NFKB1 and HMOX1 in ccRCC (Fig. 4C, F, I, L). age of hydrophobic value is -0.367, the cellular location of the gene is predicted to be in nucleoplasm or cytoplasm Correlation analysis between the selected hub genes (Fig. 3G). and ccRCC clinical parameters Meanwhile, HMOX1 is short for Heme Oxygenase1, For analyzing the association between MMP9, IRF7, and locates in 22q12.3, encoding a protein composed of NFKB1 and HMOX1 expression and ccRCC clinical 288 amino acids including 35 negatively charged amino parameters, UALCAN online platform was used. And acid residues (ASP + Glu) and 36 positively charged the result revealed an inspiring fact that although all four amino acid residues (Arg + Lys). The estimated protein genes were indicated to express higher in cancer compar- molecular weight is 32.8KD with theoretical isoelec- ing to normal samples, the four genes tend to play oppo- tric point computed as 7.89. Moreover, the estimated site roles in cancer development. To be more specific, the instability index of the protein is 60.81 and grand aver- expression of MMP9 and IRF7 genes which were sup- age of hydrophobic value is -0.427 indicating HMOX1 ported by previous survival analysis to be related with works as a cellular unstable and hydrophilic protein worse patients prognosis tend to be higher as the cancer which is consistent with the ProtScale analysis result stage and grade advancing, more obviously, both of the of HMOX1 structure showing that the protein harbors genes express higher in patients with node metastasis more hydrophilic regions than hydrophobicity regions. (Fig. 5A-J). HMOX1 is predicted to locates in cellular Golgi appa- Meanwhile, as for NFKB1 and HMOX1 genes which ratus and plasma membrane, it has been reported to be were indicated to relate with better patients prognosis, (See figure on next page.) Fig. 2 PPI network construction of 269 differential expressed meanwhile immune related genes in ccRCC and function modules analysis. (A) Based on four GEO profiles as well as IMMPORT datasets, 269 differential expression meanwhile immune related genes in ccRCC were identified, and the PPI network of these 269 genes was constructed. And based on the PPI network, (B) the first, (C) second and (C) third genes function modules were analyzed, each module was shown with a diagrammatic sketch (left diagram) and the detailed module information (right table) including the computed module score, module description and detailed involving genes. (*The first module with the highest module score meanwhile immune regulation related was focused for further analysis) Gui et al. BMC Cancer (2023) 23:649 Page 9 of 24 Fig. 2 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 10 of 24 Fig. 3 Survival analysis and basic physicochemical properties exploration of four selected signature comprised genes. The overall survival (left) and disease free survival (right) analysis in ccRCC patients, including (A) MMP9 gene, (B) NFKB1 gene, (C) IRF7 gene and (D) HMOX1 gene. The predicted cellular location (left) and computed hydrophility/hydrophobicity property of (E) MMP9 protein, (F) NFKB1 protein, (G) IRF7 protein and (H) HMOX1 protein the expression tend to keep decreasing as the cancer the difference was not statistical significant presumably stage and grade advancing, and their expression were due to the limited patients cases number in N1 group lower in patients with lymph node metastasis, although (Fig. 5K-T). Gui et al. BMC Cancer (2023) 23:649 Page 11 of 24 Table 1 Univariate combine with multivariate Cox Regression analysis result of the 4 hub genes used for signature construction OSCC parameters P Value B value HR 95% CI Univariate analysis Multivariate analysis UALCAN GEPIA MMP9 0.001 0.016 0.002 0.256 1.292 1.009–1.519 NFKB1 0.013 2.9E-05 0.004 -0.216 0.805 0.696–0.932 IRF7 < 0.0001 9E-04 0.022 0.077 1.080 1.011–1.154 HMOX1 0.034 0.00062 < 0.001 -0.222 0.801 0.708–0.906 Table 2 Basic physicochemical properties of the 4 hub genes used for gene signature construction Gene Property MMP9 NFKB1 IRF7 HMOX1 Formula C H N O S C H N O S C H N O S C H N O S 3517 5298 958 1035 28 4643 7343 1271 1458 33 2418 3740 678 710 19 1475 2323 405 427 8 Molecular Weight 78.46KD 105.36KD 54.28KD 32.82KD Number of amino acids 707AA 968AA 503AA 288AA Theoretical pI 5.69 5.20 5.89 7.89 Aliphatic index 65.13 84.74 72.07 83.02 Hydrophobic value -0.394 -0.339 -0.367 -0.427 Estimated protein half life 30 h 30 h 30 h 30 h Instability index 42.10 38.15 63.17 60.81 Selected hub genes mainly enriched signaling pathways Besides mRNA expression and potential related signal- and other types of genetic alteration analysis ing pathways, other genetic alterations including mutation For preliminary exploring the mechanism behind the dif- ratio, protein structure variant and copy number varia- ferent roles of the four genes in ccRCC, the PPI network tion of the four genes were preliminary explored based on that were centered on each of the four genes including cBioPortal database. However, limited inspiration could MMP9 (Fig. 6A), IRF7 (Fig. 6C), NFKB1 (Fig. 6E) and be achieved based on this part of analysis considering the HMOX1 (Fig. 6G) were constructed, followed by GO/ fact that only few ccRCC samples were included in cBio- KEGG analyzing the biological functions as well as sign- Portal database, although various types of MMP9 (Figure aling pathways the fours genes and their surrounding S1A), IRF7 (Figure S1B), NFKB1 (Figure S1C) and HMOX1 partner genes most enriched. (Figure S1D) genes alteration were observed in different And the results revealed that the four genes potentially human cancers indicating the potential different functions evolved in different signaling in cancer development, as these genes play in human cancers. for MMP9, it mainly evolves in proteoglycan syndecan as well as integrin mediated signaling pathways (Fig. 6B), and Construction of a 4 genes containing ccRCC prognosis IRF gene was indicated to be relate with immune and inter- related immune gene signature and clinical features feron related biological functions (Fig. 6D). Meanwhile, analysis NFKB1 gene was related with the canonical NF-kB signal- To maximum the clinical prediction value of the four ing as well as Aurora A, CD40/40L and endogenous TLR selected genes, Cox-proportional hazards analysis signaling pathways (Fig. 6F). And HMOX1 was indicated to based on LASSO algorithm was applied to construct a be associated with HIF-1a gene related cancer hypoxia and MMP9, NFKB1, IRF7 and HMOX1 four genes contain- oxygen homeostasis regulation (Fig. 6H). ing signature which weights the normalized expression (See figure on next page.) Fig. 4 The differential expression of four selected signature comprised genes in ccRCC included human cancers. UALCAN prediction of (A) MMP9 gene, (D) IRF7 gene, (G) NFKB1 gene and (J) HMOX1 gene expression in broad spectrum human cancers. GEPIA analysis of (B) MMP9 gene, (E) IRF7 gene, (H) NFKB1 gene and (K) HMOX1 gene in ccRCC comparing to normal renal samples. QPCR experiment using local hospital ccRCC samples for validating the changed expression of (C) MMP9 gene, (F) IRF7 gene, (I) NFKB1 gene and (L) HMOX1 gene in ccRCC comparing to normal renal samples Gui et al. BMC Cancer (2023) 23:649 Page 12 of 24 Fig. 4 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 13 of 24 Fig. 5 The association between four selected hub genes expression and ccRCC clinical parameters. The association between MMP9 expression and ccRCC (A) patients gender, (B) age, (C) cancer stage, (D) cancer grade and (E) lymph node metastasis. The association between IRF7 expression and ccRCC (F) patients gender, (G) age, (H) cancer stage, (I) cancer grade and (J) lymph node metastasis. The association between NFKB1 expression and ccRCC (K) patients gender, (L) age, (M) cancer stage, (N) cancer grade and (O) lymph node metastasis. The association between HMOX1 expression and ccRCC (P) patients gender, (Q) age, (R) cancer stage, (S) cancer grade and (T) lymph node metastasis. (* p < 0.05, **p < 0.01, ***p < 0.001. The first layer * which is right above the error bar representing comparison to normal group, and the above layers * which were above a secondary line represent the comparison between corresponding groups that were covered by the line) Gui et al. BMC Cancer (2023) 23:649 Page 14 of 24 level of each gene to the regression coefficient of multi - a point scale was assigned for each variable, the sum of variate Cox regression analysis. And the result revealed all the variables points equal to the final score of each a formula: Risk Score = 0.256 * expression (MMP9)— patient, and the survival could be predicted by drawing 0.222 * expression (HMOX1)—0.216 * expression a vertical line from the total point axis downward to the (NFKB1) + 0.077 * expression (IRF7) as the best signature outcome axis (Fig. 7F). for combining the four differently expressed meanwhile Secondly, besides above TCGA data, an independent immune related hub genes (Fig. 7A, B). GEO ccRCC cDNA expression profile GSE22541 was Based on the signature, the risk score for each patient also included for validating the prognosis correlation of was calculated followed by the patients being categorized the gene signature. After observing the detailed expres- into high-risk or low-risk groups according to the median sion of all the four genes in the GSE22541 datasets (Fig- risk score which was set as the cut off point for the signa - ure S2E), the patients samples were divided as high and ture (Fig. 7C). Further, the association between the gene low risk groups based on the constructed gene signature signature and ccRCC clinical features was preliminary score (Figure S2A, S2B). Afterwards, Kaplan–Meier sur- analyzed, which result revealed that higher risk score vival (Figure S2C) as well as ROC curve (Figure S2D) was positively related with older age (> 45 years old) and were also conducted, and the results supported the high more advanced T, N and M stage, meanwhile, the low risk group of patients had a statistical significantly worse risk group of patients were tend to be younger (≤ 45 years prognosis than low risk group patients. old) female with lower TNM stage (Table 3). High‑risk group of ccRCC cases were more enriched High risk score based on the gene signature indicated in immune related phenotype worse ccRCC patients prognosis After preliminary demonstrating the association between All four genes used to construct the gene signature constructed gene signature and ccRCC prognosis, the were previously supported to be high level differently influence of gene signature on cancer immune profiles expressed in ccRCC comparing to normal renal samples, was to be investigated. And in the first step, GSEA was but changed gene expression doesn’t equal to survival utilized to analyze the immune-related biological pro- association. To validate the survival relationship of the cesses linked to the signature, and the analysis result gene signature, two independent analysis methods were showed that the high-risk group cases were significantly performed. enriched in multiple biological processes, of which Firstly, a series of analysis were applied using TCGA 4 immune-related processes were identified includ - ccRCC data including at first Kaplan–Meier survival ing HUMORAL_IMMUNE_RESPONSE (NES = 1,733, analysis, which result revealed that the high risk group of Nominal p value = 0.0), REGUL ATION_OF_T_CELL_ patients had a statistical significantly worse overall sur - MIGRATION (NES = 1.762, Nominal p value = 0.0), REG- vival than their low risk counterparts (Fig. 7D). Then the UL ATION_OF_LYMPHOCYTE_CELL_MIGR ATION ROC curve showed that the area under the ROC curve (NES = 1.743, Nominal p value = 0.003), POSITIVE_REG- (AUC) of gene signature for overall survival was 0.747 ULATION_OF_IMMUNOGLOBULIN_PRODUCTION at 1 year, 0.696 at 3 year and 0.705 at 5 years (Fig. 7E). (NES = 1.754,Nominal p value = 0.0). Meanwhile, the Meanwhile, univariate Kapkan-Meier survival as well low-risk group cases were not indicated to be enriched in as multivariate Cox regression analysis were applied for any immune-related biological processes (Fig. 7G). testing the survival prediction ability of the signature, and the results supported the risk score calculated based High‑risk and low‑risk groups of ccRCC patients revealed on the gene signature works as an independent prog- disparate ICD expression levels nosis indicator for ccRCC patients together with some Besides GSEA immune phenotype enrichment analysis, other well accepted prognosis related clinical param- given the significant roles of ICD in antitumor immuno - eters including patient T and M stage (Table 4). Further, logical responses, the connection between gene signature a nomogram was constructed and and in the nomogram, and ICD related genes were evaluated for additionally (See figure on next page.) Fig. 6 PPI network centered on four selected hub genes and GO/KEGG analysis of their enriched biological pathways. (A) The PPI network which is centered on MMP9 gene for analyzing (B) the main biological signaling pathways MMP9 and its connected genes mainly participated in. (C) The PPI network which is centered on IRF7 gene for analyzing (D) the main biological signaling pathwaysIRF7 and its connected genes mainly participated in. (E) The PPI network which is centered on NFKB1 gene for analyzing (F) the main biological signaling pathways NFKB1 and its connected genes mainly participated in. (G) The PPI network which is centered on MMP9 gene for analyzing (H) the main biological signaling pathways MMP9 and its connected genes mainly participated in. (KEGG software analysis permitted by Kanehisa laboratory) Gui et al. BMC Cancer (2023) 23:649 Page 15 of 24 Fig. 6 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 16 of 24 Fig. 7 Construction of a four genes containing meanwhile immune and prognosis related ccRCC gene signature. LASSO analysis to calculate (A) the coefficient and (B) the likelihood deviance for constructing a suitable immune meanwhile prognosis related signature which was comprised of strictly calculated four genes. (C) TCGA ccRCC patients were divided into high-risk and low-risk groups based on the calculated signature score (the cut off value was set as the median signature score in all samples). (D) Survival analysis of the high-risk and low-risk groups of ccRCC patients. (E) ROC curve of the gene signature to predict ccRCC patients survival of 1 year, 3 years and 5 years respectively. (F) ccRCC patients prognosis prediction nomogram constructed based on genes signature and clinical parameters which were supported by Cox Regression to be independently related with patients survival. (G) Significant enrichment of immune-related phenotype including immune response and immune cells migration in high-risk group of ccRCC patients compared with that in low-risk group patients Gui et al. BMC Cancer (2023) 23:649 Page 17 of 24 Table 3 Association between the gene signature and ccRCC exploring the immune status in high-risk and low-risk clinical features groups of ccRCC patients. And the results revealed that the expression of a large portion of the 32 ICD related Parameters Gene signature P Value genes were statistical significantly different between the Low‑risk group High‑risk group two groups of patients indicating the diverse immune status in the microenvironment of two groups of patients Gender (Fig. 8A). female 112 (59.6%) 76 (40.4%) 0.001 male 154 (44.6%) 191 (55.4%) Age Risk score calculated based on the gene signature ≤ 45 38 (64.4%) 21 (35.6%) 0.018 associated with ccRCC estimated environment immune > 45 228 (48.1%) 246 (51.9%) score Race For further validating the immune association of the gene White 234 (50.6%) 228 (49.4%) 0.697 signature, ESTIMATE was performed to evaluate the Yellow 4 (50.0%) 4 (50.0%) immune, stromal score and tumor purity of ccRCC sam- Black 25 (44.6%) 31 (55.4%) ples. And the result revealed that although no significant T stage correlation was found between the gene signature and T1 164 (60.01%) 109 (39.9%) < 0.001 ccRCC stromal score, a mediate correlation was revealed T2 31 (44.9%) 38 (55.1%) between the risk score which was calculated based on the T3 69 (38.3%) 111 (61.7%) gene signature and tumor immune score as well as tumor T4 2 (18.2%) 9 (81.8%) purity (Fig. 8B). Meanwhile, the high risk-group patients N stage were tend to posses higher immune score and lower N0 115 (47.9%) 125 (52.1%) 0.023 tumor purity, indicating the immune targeting potential N1 3 (18.8%) 13 (81.3%) of the group of patients (Fig. 8C). M stage M0 231 (54.7%) 191 (45.3%) < 0.001 M1 22 (27.8%) 57 (72.2%) Multi immune checkpoints expressed higher in high risk group of ccRCC patients based on the gene signature Besides estimation of immune score, the association Table 4 Survival prediction value of the gene signature included between the gene signature and clinical promising ccRCC clinical parameters immune checkpoints including PD-L1, CTLA4, TIGIT, Clinical parameters P Value Exp (B) TIM-3 and LAG-3 were evaluated (Fig. 8D). And a median association was revealed between the gene sig- Univariate Multivariate analysis analysis nature and CTLA4 expression. Moreover, mild correla- tion was indicated between gene signature and two of Age < 0.001 0.026 1.627 (1.058–2.500) the immune checkpoints including LAG3 and TIGIT, Gender 0.693 - - meanwhile, no significant relation was found between Race 0.719 - - the signature and PD-L1 or TIM-3 expression (Fig. 8E). T stage < 0.001 0.018 1.360 (1.054–1.755) An inspiring fact was that all CTLA4, LAG3 and TIGIT N stage < 0.001 0.480 1.290 (0.636–2.615) tend to express higher in high-risk group of patients M stage < 0.001 < 0.001 2.794 (1.728–4.517) which was categorized based on the gene signature Signature risk score < 0.001 0.002 1.871 (1.299–2.693) (Fig. 8F), and the distribution shall be an additional sup- port besides above ESMINATE immune score evaluation result for indicating the immune targeting potential for this group of ccRCC patients. (See figure on next page.) Fig. 8 Correlation between gene signature and ccRCC immune microenvironment landscape. (A) Relative expression of ICD related genes in high-risk and low-risk groups of ccRCC patients. (B) Correlation between gene signature and ccRCC computed immune score, stromal score and tumor purity calculated using ESTIMATE algorithm. (C) Estimated immune score, stromal score and tumor purity distribution in high-risk and low-risk ccRCC groups respectively. (D) Association between gene signature and immune checkpoints expression. (E) Correlation between gene signature and PD-L1, LAG-3, TIGIT and CALT-4 expression respectively. (F) Relative expression of five immune checkpoints including PD-L1, LAG-3, TIGIT, TIM3 and CALT-4 expression in high-risk and low-risk ccRCC groups respectively Gui et al. BMC Cancer (2023) 23:649 Page 18 of 24 Fig. 8 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 19 of 24 Evaluation of relationships between gene signature and 22 and stratify patients into different groups for increasing tumor infiltrating immune cells (TICs) the potential effectiveness of ICIs therapy. Considering Previous analysis supported that the constructed gene most biological cellular functions were performed by signature was related to immunity, so we carried out different types of cell proteins, in the study, we mainly analyses on 22 TICs whose distribution profiles were focused on the protein encoding genes that were aber- draw based on CIBERSORT algorithm to further study rant differently expressed meanwhile prognosis as well the interaction between the gene signature and ccRCC as immune related for series of analysis. Thus, we mainly immune microenvironment. And the correlation analy- focused on GEO transcriptome profiles for identifing sis result found four types of TICs to be related with the potential immune regulation related gene candidates. gene signature including plasma cells, activated CD4( +) Based on four different ccRCC cDNA expression pro - T memory cells, activated dendritic cells and resting files which were all selected by strict criteria as see in the mast cells (Fig. 9A-C). Materials and Methods part, we identified the differen - Further, the prognostic abilities of the 22 TICs were tial expression genes in ccRCC cancer vs. normal renal tested and the results revealed that of the four signature tissues and then divided them into 4 groups according to related TICs, CD4( +) T cell and resting mast cell were the difference level as < twofold, 2 ~ fourfold, 4 ~ eightfold able to predict ccRCC patients prognosis. The resting and > eightfold genes considering the potential unique and activated CD4( +) T memory cells played opposite functions and clinical use of each group, for example, roles in patients survival, namely the activated CD4( +) an interesting phenomenon has been discovered that T memory cells were related with worse patients survival the more genes expression difference are, the more their (Fig. 9D), meanwhile, the resting CD4( +) T memory cells cellular location tend to be far away from cell nuclear predicting better patients survival (Fig. 9E). Also, the [51–53]. For the feasibility of further clinical medical use, resting mast cells were correlated with positive patients in the selection process of candidate genes, we mainly prognosis (Fig. 9F). focused on the high level differently expressed namely Combining the analysis results, one inspiring deduc- at least > fourfold genes that were more convenient to tion could be draw that CD4( +) T memory cells and be tested by IHC experiment which has been a common resting mast cells not only are significantly related to the method in clinical pathology diagnosis, considering that gene signature but also predict ccRCC patients progno- the genes shall harbor more chance to be translated into sis, indicating these immune cells may play important clinical use if they are suitable to be tested by IHC. Fur- roles in the immune regulation of the gene signature in ther, the intersection between GEO selected high level ccRCC microenvironment. aberrant differently expressed genes and immune related gene list from IMMPORT database indicated 269 genes Discussion that were both high level expression changed in ccRCC ICIs has been an increasing rising up clinical method and and immune related as candidate genes for next step holds great promise for treating ccRCC [9], but the effec - analysis. tive biomarkers for predicting immune response are still As for the construction of multi genes containing sig- lacking, the well accepted immune prediction biomark- nature, LASSO algorithm has been widely accepted as ers in other cancers for instance PD-L1 expression, MSI an effective tool that is suitable to construct gene mod - status and TMB haven’t been supported thoroughly by els basing on large numbers of correlated covariate. evidence-based medicine to be effective in ccRCC [45, But instead of constructing an gene signature directly 46]. For the clinical benefit from ICI therapy, it is of great from the 269 candidate genes, we further in succession importance to keep exploring ccRCC genome and iden- performed module analysis as well as multiple survival tifying new potential biomarkers thus benefiting further analysis to scale down the candidate genes and identify clinical application of immunotherapy in the cancer. the promising “unique key genes” during ccRCC devel- In recent years, multiple genes containing signatures opment, and only used LASSO for estimating the coef- representative of caner immune status have been iden- ficient of signature genes. This is for the considering tified in several cancers besides ccRCC, including gene that the signature and consist genes should be of more signatures that were comprised of LncRNAs, miRNAs clinical potential if they were not only immune regula- or immune regulation related genes, and they have been tion but also survival related for clinical medical drug showing inspiring clinical effects [47–50]. Based on targeting use. Therefore, the PPI network of the 269 these reports, it’s of clinical feasibility to explore ccRCC genes was constructed followed by genes module analy- genome information and develop meaningful immune sis which highlighted a 46 genes containing cluster, and prediction models which were also prognosis related to further survival analysis including GEPIA and UAL- evaluate the immune status of ccRCC microenvironment CAN univariate survival as well as multivariate Cox Gui et al. BMC Cancer (2023) 23:649 Page 20 of 24 Fig. 9 Correlation between gene signature and 22 immune cells infiltration in ccRCC. (A) Relative distribution of 22 immune cells in high-risk and low-risk groups of ccRCC patients. (B, C) Correlation between gene signature and various immune cells infiltration in ccRCC. Association between (D) T cells CD4 memory activated, (E) T cells CD4 memory resting and (F) Mast cells resting microenvironment infiltration and ccRCC patients survival Gui et al. BMC Cancer (2023) 23:649 Page 21 of 24 Regression analysis of each of the 46 genes supported Although survival relation was an important part, four genes: MMP9, NFKB1, IRF7 and HMOX1 to be the main aim of the signature was for potential associated with patients survival and worked as inde- immune prediction. Immune escape has been one of pendent prognostic indicators in ccRCC development. the major characteristics in malignant tumors involv- Interestingly, no direct relationship has yet been dis- ing multiple probable mechanisms [54, 55], for exam- covered among the four genes. MMP9 is s a member ple the increasing immune suppressive cells including of matrix metalloproteinase (MMP) family which is Treg cells and tumour-associated macrophages (TAM) well known to be involved in the breakdown of extra- in tumor microenvironment [47], and the up-regu- cellular matrix during multiple normal physiological lated expression of immunosuppressive molecules and diseases processes, and MMP9 has been reported for instance cytotoxic T lymphocyte associated anti- to be able to degrades type IV and V collagen which gen-4 (CTLA-4), also decreasing expression of can- are important microenvironment elements. NFKB1 is a cer antigens which results the inactivation of tumor transcription regulator that could be activated by cel- killing CD8 + T cells [56–59]. Therefore, we explored lular stimuli such as cytokines, ultraviolet irradiation, the probable relation between the gene signature and bacterial and viral products, and inappropriate acti- immune suppressive mechanisms. And the results vation of the gene has been known to associate with revealed that the difference expression of ICD related a number of inflammatory diseases, while persistent genes in high-risk and low-risk groups of patients inhibition of NFKB1 leads to inappropriate immune cell which were categorized based on the constructed gene development or delayed cell growth. IFR7 is predicted signature, as well as the statistical significant corre- to locates in nucleoplasm and cytoplasm and it has lation between the signature and ESTIATE immune been reported to play roles in innate immune response score supported the signature was immune modula- against DNA and RNA viruses. Meanwhile, HMOX1 tion associated. Further, we investigated the expres- is predicted to locates in cellular Golgi apparatus and sion of immune checkpoints including PD-L1, CTLA4, plasma membrane and it has been reported to be TIGIT, TIM-3 and LAG-3 between the high-risk and associated with the development of heme oxygenase1 low-risk groups of patients, and the results showed the deficiency and pulmonary disease, as well as chronic high-risk patients had higher expression of CTLA4, obstructive. Moreover, the PPI network centered on the LAG3 and TIGIT than the low-risk patients indicat- four genes and following KEGG analysis also supported ing the immune targeting potential for this group of the independent roles of these genes in cancer devel- ccRCC patients. opment. The together identification of the four genes As the results of the relation analysis between the and a gene signature combine all of them indicating the signature and 22 TICs indicated that the signature was elaborate collaboration network of various genes in cel- significantly related with CD4( +) T memory cells and lular activities, opening up further cancer researches of resting mast cells infiltration, and not only the two unlimited possibilities. immune cells were related with the signature, but also Based on the selected four genes and coefficient calcu - they were able to predict ccRCC patients prognosis. lated with LASSO algorithm for each gene, an immune CD4( +) T memory cells have already been reported meanwhile prognosis related gene signature was con- to confer vital functions on malignancy immune reg- structed. Survival relationship validation including ulation, including participating in the activation of Kaplan–Meier survival, Cox proportional-hazards model CD8 + T and NK killing cells, involving in the tumour and ROC curve based on both TCGA data and an inde- immunological reactions [60, 61]. And mast cells were pendent GEO profile all supported the signature worked reported to be able to not only influence tumor expan - as an prognostic factor after combining the four genes in sion via inducing angiogenesis and changing tumor one equation, proving the effectiveness of apply the gene extracellular matrix composition, but also could influ - signature in ccRCC prognosis prediction. Since many ence the infiltration and activity of dendritic cells, clinical parameters especially tumor TNM stage as been tumor-associated macrophages and lymphocytes, pro- well known as critical survival related aspects, we pro- moting pro-inflammatory reactions in tumor micro - posed a nomogram assessment that combines the signa- environment [47, 62, 63]. The association between the ture and other clinical features. Although current result signature and immune checkpoints expression as well has not supported the signature to be a better prognosis as different immune cells infiltration, suggesting the factor than TNM stage, the construction of the nomo- stronger immunosuppressive environment in high-risk gram shall work as a complementary perspective on indi- groups of patients comparing to low-risk group, high- vidual tumour and aiding the comprehensive evaluation lighting the potential of this group to benefit from fur - of clinical ccRCC patients prognosis. ther clinical immunotherapy. Gui et al. BMC Cancer (2023) 23:649 Page 22 of 24 manuscript. As the corresponding author, both WM and CW had full access to Conclusion all data of the manuscript, CW made the final decision to submit the article for The present study defined a four genes containing signa - publication. All authors read and approved the final manuscript. ture based on ccRCC genes expression information, the Funding signature was not only closely associated with patients The work was supported by the China central government funds for guiding survival, but also immune regulation related. Multiple local scientific and technological development (YDZJSX2021A042), the fund in vitro experiments data analysis supported the asso- of Shanxi Medical key scientific research project (2021XM34) and the fund of Natural Science Foundation of ShanXi Province in China (201901D211498). ciation between signature and ccRCC microenviron- ment immune aspects including immune checkpoints Availability of data and materials expression and various types of immune cells infiltration. Publicly available datasets were analyzed in this study. The data can be found here: GSE53000:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 Although the current result is not yet enough to sup- 000. GSE53757:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 port the application of the signature in clinical medical 757. GSE68417:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE68 immunotherapy, rigorous prospective studies performed 421. GSE71963:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE71 on animal models as well as clinical trials are still needed, GSE22541:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE22 541. the results shall provide meaningful insight into better All data generated or analyzed based on the online datasets and other experi- understanding of the disease and shed lighting on further ments during this study are included in this published article. ccRCC immune regulation researches. Declarations Abbreviations Ethnic approval and consent to participate ccRCC Clear cell renal cell carcinoma All of the local hospital ccRCC patients samples used for qPCR experiment GEO G ene Expression Omnibus were stored at the hospital biobank (Second Hospital of ShanXi Medical Uni- PPI Protein–protein interaction network versity). All the sample donors have been informed of the possibility that their ICIs I mmune checkpoint inhibitors donated tissues might be used for certain scientific research. The Informed TMB Tumor mutation burden consent from the patients were all obtained, and the paper agreement signed ICD Immunogenic cell death with the donors’ signatures were kept by the biobank committee. And the TIC Tumor infiltrating immune cells biobank ccRCC samples being used in this research was approval by both TAM Tumour-associated macrophages the biobank committee and Hospital institutional ethics committee (Second Hospital of ShanXi Medical University, ShanXi Province, China). And all experi- ment methods were carried out in accordance with relevant guidelines and Supplementary Information regulations or declaration of Helsinki. The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s12885- 023- 11150-4. Consent for publication Not applicable. Additional file 1: Supplementary Figure 1. Genetic alterations of four Competing interests hub genes based on cBioPortal dataset. Different types of (A) MMP9, (B) All of the authors agreed the publication of the paper and declare no conflicts IRF7, (C) NFKB1 and (D) HMOX1 variations including gene amplification, of interests. deletion, mutation and structural variants in various human cancers revealed by cBioPortal dataset. Author details Additional file 2: Supplementary Figure 2. GEO profiles validating the 1 Department of Pathology, Second Clinical Medical College of ShanXi Medical prognosis correlation of the constructed gene signature. (A) GSE22541 2 University, Tai Yuan City, ShanXi Province, China. Department of Anesthesiol- patients were divided into high-risk and low-risk groups based on the ogy, Second Hospital of ShanXi Medical University, Tai Yuan, ShanXi Province, calculated signature score. (B) The survival status of all the GSE22541 3 China. Department of Pathology, Second Hospital of ShanXi Medical Univer- patients samples. (C) Survival analysis of the high-risk and low-risk groups sity, No.382 Wuyi Road, Tai Yuan, ShanXi Province, 030000, China. of GSE22541 patients. (D) ROC curve of the gene signature to predict GSE22541 patients survival. (E) Relative expression of MMP9, IRF7, NFKB1 Received: 15 December 2022 Accepted: 4 July 2023 and HMOX1 genes in GSE22541 samples displayed in a heatmap. Additional file 3: Supplementary Table 1. Detailed information of the GEO datasets used for identifying differently expressed genes in ccRCC vs normal renal tissues. Supplementary Table 2. Number of the four levels differently expressed genes in ccRCC analyzed based on four GEO profiles. References Supplementary Table 3. Detailed information of the 269 differently 1. Wu J, Xu WH, Wei Y, Qu YY, Zhang HL, Ye DW. An integrated score and expressed meanwhile immune modulation related genes in ccRCC. nomogram combining clinical and immunohistochemistry factors to pre- dict high isup grade clear cell renal cell carcinoma. Front Oncol. 2018;8:634. 2. Browning L, Colling R, Verrill C. WHO/ISUP grading of clear cell renal cell Acknowledgements carcinoma and papillary renal cell carcinoma; validation of grading on We sincerely appreciate the researchers for providing their GEO databases the digital pathology platform and perspectives on reproducibility of information online which were important data resources for the study, it is our grade. Diagn Pathol. 2021;16(1):75. pleasure to acknowledge their contributions. 3. Stone L. Kidney cancer: activation of oncogenes driven by VHL loss in ccRCC. Nat Rev Urol. 2017;14(11):637. Authors’ contributions 4. Expert Group of G, Urological Pathology CSoP. [Recent advances and ZG, JD and NW designed the study and together drafting the manuscript, expert consensus on molecular pathology of renal cell carcinomas (2020 contributed equally to the whole study. NS, ZY and XW performed the data version)]. Zhonghua Bing Li Xue Za Zhi. 2020;49(12):1232–41. collecting and analysis. HY, NZ and ZZ participated in the data interpretation 5. Williamson SR, Gill AJ, Argani P, Chen YB, Egevad L, Kristiansen G, and study design, RW were involved in the drafting and critical revision of the Grignon DJ, Hes O. Report from the International Society of Urological Gui et al. BMC Cancer (2023) 23:649 Page 23 of 24 Pathology (ISUP) consultation conference on molecular pathology 23. Bedke J, Stuhler V. KEYNOTE-426: axitinib plus pembrolizumab vs suni- of urogenital cancers: III: molecular pathology of kidney cancer. Am J tinib in the first-line treatment of metastatic renal cell carcinoma. Urologe Surg Pathol. 2020;44(7):e47–65. A. 2020;59(7):841–2. 6. Ibragimova I, Maradeo ME, Dulaimi E, Cairns P. Aberrant promoter 24. Samstein RM, Lee CH, Shoushtari AN, Hellmann MD, Shen R, Janjigian hypermethylation of PBRM1, BAP1, SETD2, KDM6A and other chro- YY, Barron DA, Zehir A, Jordan EJ, Omuro A, et al. Tumor mutational load matin-modifying genes is absent or rare in clear cell RCC. Epigenetics. predicts survival after immunotherapy across multiple cancer types. Nat 2013;8(5):486–93. Genet. 2019;51(2):202–6. 7. Wi YC, Moon A, Jung MJ, Kim Y, Bang SS, Jang K, Paik SS, Shin SJ. Loss 25. Raimondi A, Sepe P, Zattarin E, Mennitto A, Stellato M, Claps M, Gua- of nuclear BAP1 expression is associated with high WHO/ISUP grade in dalupi V, Verzoni E, de Braud F, Procopio G. Predictive biomarkers of clear cell renal cell carcinoma. J Pathol Transl Med. 2018;52(6):378–85. response to immunotherapy in metastatic renal cell cancer. Front Oncol. 8. Wang Y, Guo X, Bray MJ, Ding Z, Zhao Z. An integrative genomics 2020;10:1644. approach for identifying novel functional consequences of PBRM1 26. GSE53000: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 truncated mutations in clear cell renal cell carcinoma (ccRCC). BMC 000. Accessed 1 Jan 2022. Genomics. 2016;17(Suppl 7):515. 27. GSE53757: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 9. Srivastava A, Doppalapudi SK, Patel HV, Srinivasan R, Singer EA. The 757. Accessed 2 Jan 2022. roaring 2020s: a new decade of systemic therapy for renal cell carci- 28. GSE68417: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE68 noma. Curr Opin Oncol. 2022;34(3):234–42. 417. Accessed 2 Jan 2022. 10. Pandey A. KEYNOTE-564: Adjuvant immunotherapy for renal cell carci- 29. GSE71963: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE71 noma. Indian J Urol. 2022;38(1):75–6. 963. Accessed 3 Jan 2022. 11. Doppalapudi SK, Leopold ZR, Thaper A, Kaldany A, Chua K, Patel HV, 30. IMMPORT database: https:// www. immpo rt. org/ home. Accessed 7 Feb Srivastava A, Singer EA. Clearing up clear cell: clarifying the immuno- 2022. oncology treatment landscape for metastatic clear cell RCC. Cancers 31. VENNY: https:// bioin fogp. cnb. csic. es/ tools/ venny/. . Accessed 9 Jan 2022. (Basel). 2021;13(16):4140. 32. STRING: https:// string- db. org/. Accessed 15 Feb 2022. 12. Rini BI, Pal SK, Escudier BJ, Atkins MB, Hutson TE, Porta C, Verzoni E, 33. Cytoscape:http:// www. softp edia. com/ get/ Scien ce- CAD/ Cytos cape. Needle MN, McDermott DF. Tivozanib versus sorafenib in patients shtml. Accessed 22 Feb 2022. with advanced renal cell carcinoma ( TIVO-3): a phase 3, multi- 34. FunRich3.1.3: http:// www. funri ch. org/. Accessed 27 Feb 2022. centre, randomised, controlled, open-label study. Lancet Oncol. 35. UALCAL: http:// ualcan. path. uab. edu/ analy sis. html. Accessed 7 Mar 2022. 2020;21(1):95–104. 36. GEPIA: http:// gepia. cancer- pku. cn/. Accessed 2 Mar 2022. 13. Schmidt AL, Tabakin AL, Singer EA, Choueiri TK, McKay RR. Next steps: 37. ProtParam: https:// web. expasy. org/ protp aram/. Accessed 13 Mar 2022. sequencing therapies in metastatic kidney cancer in the contemporary 38. ProtScale: https:// web. expasy. org/ prots cale/. Accessed 13 Mar 2022. era. Am Soc Clin Oncol Educ Book. 2021;41:1–11. 39. HumanProteinAtlas: https:// www. prote inatl as. org/. Accessed 14 Mar 14. Zhu J, Armstrong AJ, Friedlander T W, Kim W, Pal SK, George DJ, Zhang 2022. T. Biomarkers of immunotherapy in urothelial and renal cell carcinoma: 40. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. PD-L1, tumor mutational burden, and beyond. J Immunother Cancer. Nucleic Acids Res. 2000;28(1):27–30. 2018;6(1):4. 41. Kanehisa M, Furumichi M, Sato Y, Kawashima M, Ishiguro-Watanabe M. 15 Dumitru A, Dobrica EC, Croitoru A, Cretoiu SM, Gaspar BS. Focus on KEGG for taxonomy-based analysis of pathways and genomes. Nucleic PD-1/PD-L1 as a therapeutic target in ovarian cancer. Int J Mol Sci. Acids Res. 2023;51(D1):D587–92. 2022;23(20):12067. 42. GEPIA: https:// www. gsea- msigdb. org/ gsea/ index. jsp. Accessed 07 Jun 16. Nunez Abad M, Calabuig-Farinas S, de Lobo Mena M, Torres-Martinez 2022. S, Garcia Gonzalez C, Garcia Garcia JA, Iranzo Gonzalez-Cruz V, Camps 43. Wang X, Wu S, Liu F, Ke D, Wang X, Pan D, Xu W, Zhou L, He W. An immu- Herrero C. Programmed death-ligand 1 (PD-L1) as immunotherapy nogenic cell death-related classification predicts prognosis and response biomarker in breast cancer. Cancers (Basel). 2022;14(2):307. to immunotherapy in head and neck squamous cell carcinoma. Front 17. Deng H, Zhao Y, Cai X, Chen H, Cheng B, Zhong R, Li F, Xiong S, Li J, Liu Immunol. 2021;12:781466. J, et al. PD-L1 expression and Tumor mutation burden as Pathological 44. Cibersort: https:// ciber sortx. stanf ord. edu/. Accessed 16 May 2022. response biomarkers of Neoadjuvant immunotherapy for early-stage 45. Labriola MK, Zhu J, Gupta RT, McCall S, Jackson J, Kong EF, White JR, non-small cell lung cancer: a systematic review and meta-analysis. Crit Cerqueira G, Gerding K, Simmons JK, et al. Characterization of tumor Rev Oncol Hematol. 2022;170:103582. mutation burden, PD-L1 and DNA repair genes to assess relationship to 18. Motzer RJ, Escudier B, George S, Hammers HJ, Srinivas S, Tykodi SS, Sos- immune checkpoint inhibitors response in metastatic renal cell carci- man JA, Plimack ER, Procopio G, McDermott DF, et al. Nivolumab versus noma. J Immunother Cancer. 2020;8(1):e000319. everolimus in patients with advanced renal cell carcinoma: updated 46. Kotecha RR, Motzer RJ, Voss MH. Towards individualized therapy for meta- results with long-term follow-up of the randomized, open-label, phase static renal cell carcinoma. Nat Rev Clin Oncol. 2019;16(10):621–33. 3 checkmate 025 trial. Cancer. 2020;126(18):4156–67. 47. Long J, Wang A, Bai Y, Lin J, Yang X, Wang D, Yang X, Jiang Y, Zhao H. 19. Tomita Y, Fukasawa S, Shinohara N, Kitamura H, Oya M, Eto M, Tanabe Development and validation of a TP53-associated immune prognostic K, Saito M, Kimura G, Yonese J, et al. Nivolumab versus everolimus in model for hepatocellular carcinoma. EBioMedicine. 2019;42:363–74. advanced renal cell carcinoma: Japanese subgroup 3-year follow-up 48. Mei W, Jin L, Zhang B, Sun X, Yang G, Li S, Ye L. Computer classification analysis from the phase III checkmate 025 study. Jpn J Clin Oncol. and construction of a novel prognostic signature based on moonlighting 2019;49(6):506–14. genes in prostate cancer. Front Oncol. 2022;12:982267. 20. Motzer RJ, Choueiri TK, McDermott DF, Powles T, Vano YA, Gupta S, Yao 49 Yin Y, Li S, Liang X, Li K, Xie M, Hu B. Construction and validation of an J, Han C, Ammar R, Papillon-Cavanagh S, et al. Biomarker analysis from oxaliplatin-resistant gene signature in colorectal cancer patients who CheckMate 214: nivolumab plus ipilimumab versus sunitinib in renal underwent chemotherapy. Pharmaceuticals (Basel). 2022;15(9):1139. cell carcinoma. J Immunother Cancer. 2022;10(3):e004316. https:// doi. 50. Gao J, Huo S, Zhang Y, Zhao Z, Pan H, Liu X. Construction of ovarian org/ 10. 1136/ jitc- 2021- 004316. metastasis-related immune signature predicting prognosis of gastric 21. Jena R. CheckMate 214 trial: Immune checkpoint regulators for cancer patients. Cancer Med. 2022;12(1):913–29. advanced renal cell carcinoma. Indian J Urol. 2020;36(3):231–3. 51. Zhang X, Wang Z, Zeng Z, Shen N, Wang B, Zhang Y, Shen H, Lu W, Wei 22. Powles T, Plimack ER, Soulieres D, Waddell T, Stus V, Gafanov R, Nosov R, Ma W, et al. Bioinformatic analysis identifying FGF1 gene as a new D, Pouliot F, Melichar B, Vynnychenko I, et al. Pembrolizumab plus prognostic indicator in clear cell renal cell carcinoma. Cancer Cell Int. axitinib versus sunitinib monotherapy as first-line treatment of 2021;21(1):222. advanced renal cell carcinoma (KEYNOTE-426): extended follow- 52. Wei R, Qi G, Zeng Z, Shen N, Wang Z, Shen H, Gao L, Song C, Ma W, Wang up from a randomised, open-label, phase 3 trial. Lancet Oncol. C. IMUP and GPRC5A: two newly identified risk score indicators in pancre - 2020;21(12):1563–73. atic ductal adenocarcinoma. Cancer Cell Int. 2021;21(1):620. Gui et al. BMC Cancer (2023) 23:649 Page 24 of 24 53. Ma W, Wang B, Zhang Y, Wang Z, Niu D, Chen S, Zhang Z, Shen N, Han W, Zhang X, et al. Prognostic significance of TOP2A in non-small cell lung cancer revealed by bioinformatic analysis. Cancer Cell Int. 2019;19:239. 54. Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer immunoed- iting: from immunosurveillance to tumor escape. Nat Immunol. 2002;3(11):991–8. 55. Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrat- ing immunity’s roles in cancer suppression and promotion. Science. 2011;331(6024):1565–70. 56. Li X, Shao C, Shi Y, Han W. Lessons learned from the blockade of immune checkpoints in cancer immunotherapy. J Hematol Oncol. 2018;11(1):31. 57 Li Y, Li F, Jiang F, Lv X, Zhang R, Lu A, Zhang G. A mini-review for cancer immunotherapy: molecular understanding of PD-1/PD-L1 path- way & translational blockade of immune checkpoints. Int J Mol Sci. 2016;17(7):1151. 58. Pardoll DM. The blockade of immune checkpoints in cancer immuno- therapy. Nat Rev Cancer. 2012;12(4):252–64. 59. Zou W, Chen L. Inhibitory B7-family molecules in the tumour microenvi- ronment. Nat Rev Immunol. 2008;8(6):467–77. 60. Khan M, Arooj S, Wang H. Soluble B7-CD28 family inhibitory immune checkpoint proteins and anti-cancer immunotherapy. Front Immunol. 2021;12:651634. 61. Schildberg FA, Klein SR, Freeman GJ, Sharpe AH. Coinhibitory pathways in the B7-CD28 ligand-receptor family. Immunity. 2016;44(5):955–72. 62. Kalkusova K, Smite S, Darras E, Taborska P, Stakheev D, Vannucci L, Bartunkova J, Smrz D. Mast cells and dendritic cells as cellular immune checkpoints in immunotherapy of solid tumors. Int J Mol Sci. 2022;23(19):11080. 63. Honda T, Keith YH. Novel insights into the immune-regulatory func- tions of mast cells in the cutaneous immune response. Front Immunol. 2022;13:898419. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your ﬁeld rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Cancer Springer Journals http://www.deepdyve.com/lp/springer-journals/immune-regulation-and-prognosis-indicating-ability-of-a-newly-VJ5GiOTUm9

Loading next page...

References (78)

F. Schildberg, Sarah Klein, G. Freeman, A. Sharpe (2016)
Coinhibitory Pathways in the B7-CD28 Ligand-Receptor Family.
Immunity, 44 5
(ProtParam: https://web.expasy.org/protparam/. Accessed 13 Mar 2022.)
ProtParam: https://web.expasy.org/protparam/. Accessed 13 Mar 2022.
ProtParam: https://web.expasy.org/protparam/. Accessed 13 Mar 2022., ProtParam: https://web.expasy.org/protparam/. Accessed 13 Mar 2022.
S. Doppalapudi, Zev Leopold, Akshay Thaper, A. Kaldany, Kevin Chua, Hiren Patel, A. Srivastava, E. Singer (2021)
Clearing up Clear Cell: Clarifying the Immuno-Oncology Treatment Landscape for Metastatic Clear Cell RCC
Cancers, 13
R. Schreiber, L. Old, M. Smyth (2011)
Cancer Immunoediting: Integrating Immunity’s Roles in Cancer Suppression and Promotion
Science, 331
(2022)
VENNY
(GSE53000: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53000. Accessed 1 Jan 2022.)
GSE53000: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53000. Accessed 1 Jan 2022.
GSE53000: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53000. Accessed 1 Jan 2022., GSE53000: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53000. Accessed 1 Jan 2022.
J. Long, Anqiang Wang, Yi Bai, Jianzhen Lin, Xu Yang, Dongxu Wang, Xiaobo Yang, Yan Jiang, Haitao Zhao (2019)
Development and validation of a TP53-associated immune prognostic model for hepatocellular carcinoma
EBioMedicine, 42
(2022)
UALCAL
(2022)
Cytoscape:http
A. Srivastava, S. Doppalapudi, Hiren Patel, R. Srinivasan, E. Singer (2022)
The roaring 2020s: a new decade of systemic therapy for renal cell carcinoma
Current Opinion in Oncology, 34
(Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science. 2011;331(6024):1565–70.)
Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science. 2011;331(6024):1565–70.
Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science. 2011;331(6024):1565–70., Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science. 2011;331(6024):1565–70.
FunRich 3 . 1
Yuanyuan Wang, Xingyi Guo, M. Bray, Z. Ding, Zhongming Zhao (2016)
An integrative genomics approach for identifying novel functional consequences of PBRM1 truncated mutations in clear cell renal cell carcinoma (ccRCC)
BMC Genomics, 17
Muhammad Khan, Sumbal Arooj, Hua Wang (2021)
Soluble B7-CD28 Family Inhibitory Immune Checkpoint Proteins and Anti-Cancer Immunotherapy
Frontiers in Immunology, 12
(ProtScale: https://web.expasy.org/protscale/. Accessed 13 Mar 2022.)
ProtScale: https://web.expasy.org/protscale/. Accessed 13 Mar 2022.
ProtScale: https://web.expasy.org/protscale/. Accessed 13 Mar 2022., ProtScale: https://web.expasy.org/protscale/. Accessed 13 Mar 2022.
L. Browning, R. Colling, C. Verrill (2021)
WHO/ISUP grading of clear cell renal cell carcinoma and papillary renal cell carcinoma; validation of grading on the digital pathology platform and perspectives on reproducibility of grade
Diagnostic Pathology, 16
R. Wei, G. Qi, Zixin Zeng, N. Shen, Ziyue Wang, Honghong Shen, Lifang Gao, Chengkun Song, Wen-xia Ma, Chen Wang (2021)
IMUP and GPRC5A: two newly identified risk score indicators in pancreatic ductal adenocarcinoma
Cancer Cell International, 21
Xinwen Wang, Shouwu Wu, Feng Liu, Dianshan Ke, Xinwu Wang, Dinglong Pan, Weifeng Xu, Ling Zhou, Weidong He (2021)
An Immunogenic Cell Death-Related Classification Predicts Prognosis and Response to Immunotherapy in Head and Neck Squamous Cell Carcinoma
Frontiers in Immunology, 12
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations
Xiaolei Li, C. Shao, Yufang Shi, W. Han (2018)
Lessons learned from the blockade of immune checkpoints in cancer immunotherapy
Journal of Hematology & Oncology, 11
Jason Zhu, A. Armstrong, T. Friedlander, W. Kim, S. Pal, D. George, Tian Zhang (2018)
Biomarkers of immunotherapy in urothelial and renal cell carcinoma: PD-L1, tumor mutational burden, and beyond
Journal for Immunotherapy of Cancer, 6
A. Raimondi, P. Sepe, E. Zattarin, A. Mennitto, M. Stellato, M. Claps, V. Guadalupi, E. Verzoni, F. Braud, G. Procopio (2020)
Predictive Biomarkers of Response to Immunotherapy in Metastatic Renal Cell Cancer
Frontiers in Oncology, 10
Matthew Labriola, Jason Zhu, Rajan Gupta, S. McCall, J. Jackson, E. Kong, J. White, G. Cerqueira, Kelly Gerding, J. Simmons, D. George, Tian Zhang (2020)
Characterization of tumor mutation burden, PD-L1 and DNA repair genes to assess relationship to immune checkpoint inhibitors response in metastatic renal cell carcinoma
Journal for Immunotherapy of Cancer, 8
(2022)
HumanProteinAtlas
R. Jena (2020)
CheckMate 214 trial: Immune checkpoint regulators for advanced renal cell carcinoma
Indian Journal of Urology : IJU : Journal of the Urological Society of India, 36
Xiaoqin Zhang, Ziyue Wang, Zixin Zeng, N. Shen, Bin Wang, Yaping Zhang, Honghong Shen, Wei Lu, R. Wei, Wenxia Ma, Chen Wang (2020)
Bioinformatic analysis identifying FGF1 gene as a new prognostic indicator in clear cell Renal Cell Carcinoma
Cancer Cell International, 21
A. Dumitru, E. Dobrica, A. Croitoru, S. Crețoiu, B. Gaspar (2022)
Focus on PD-1/PD-L1 as a Therapeutic Target in Ovarian Cancer
International Journal of Molecular Sciences, 23
R. Kotecha, R. Motzer, M. Voss (2019)
Towards individualized therapy for metastatic renal cell carcinoma
Nature Reviews Clinical Oncology
(GSE71963: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE71963. Accessed 3 Jan 2022.)
GSE71963: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE71963. Accessed 3 Jan 2022.
GSE71963: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE71963. Accessed 3 Jan 2022., GSE71963: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE71963. Accessed 3 Jan 2022.
(GEPIA: http://gepia.cancer-pku.cn/. Accessed 2 Mar 2022.)
GEPIA: http://gepia.cancer-pku.cn/. Accessed 2 Mar 2022.
GEPIA: http://gepia.cancer-pku.cn/. Accessed 2 Mar 2022., GEPIA: http://gepia.cancer-pku.cn/. Accessed 2 Mar 2022.
R. Motzer, B. Escudier, S. George, H. Hammers, S. Srinivas, S. Tykodi, J. Sosman, E. Plimack, G. Procopio, D. McDermott, D. Castellano, T. Choueiri, F. Donskov, H. Gurney, S. Oudard, M. Richardet, K. Peltola, A. Alva, M. Carducci, J. Wagstaff, C. Chevreau, S. Fukasawa, Y. Tomita, T. Gauler, C. Kollmannsberger, F. Schutz, J. Larkin, D. Cella, M. McHenry, S. Saggi, N. Tannir (2020)
Nivolumab versus everolimus in patients with advanced renal cell carcinoma: Updated results with long-term follow-up of the randomized, open-label, phase 3 CheckMate 025 trial.
Cancer
B. Rini, S. Pal, B. Escudier, M. Atkins, T. Hutson, C. Porta, E. Verzoni, M. Needle, D. McDermott (2019)
Tivozanib versus sorafenib in patients with advanced renal cell carcinoma (TIVO-3): a phase 3, multicentre, randomised, controlled, open-label study.
The Lancet. Oncology
(GSE68417: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE68417. Accessed 2 Jan 2022.)
GSE68417: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE68417. Accessed 2 Jan 2022.
GSE68417: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE68417. Accessed 2 Jan 2022., GSE68417: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE68417. Accessed 2 Jan 2022.
(HumanProteinAtlas: https://www.proteinatlas.org/. Accessed 14 Mar 2022.)
HumanProteinAtlas: https://www.proteinatlas.org/. Accessed 14 Mar 2022.
HumanProteinAtlas: https://www.proteinatlas.org/. Accessed 14 Mar 2022., HumanProteinAtlas: https://www.proteinatlas.org/. Accessed 14 Mar 2022.
Katerina Kalkusova, Sindija Šmite, Elea Darras, P. Taborska, D. Stakheev, L. Vannucci, J. Bartunkova, D. Smrz (2022)
Mast Cells and Dendritic Cells as Cellular Immune Checkpoints in Immunotherapy of Solid Tumors
International Journal of Molecular Sciences, 23
R. Motzer, T. Choueiri, D. McDermott, T. Powles, Y. Vano, Saurabh Gupta, Jin Yao, Celine Han, R. Ammar, Simon Papillon-Cavanagh, S. Saggi, M. McHenry, P. Ross-Macdonald, M. Wind-Rotolo (2022)
Biomarker analysis from CheckMate 214: nivolumab plus ipilimumab versus sunitinib in renal cell carcinoma
Journal for Immunotherapy of Cancer, 10
(Cytoscape:http://www.softpedia.com/get/Science-CAD/Cytoscape.shtml. Accessed 22 Feb 2022.)
Cytoscape:http://www.softpedia.com/get/Science-CAD/Cytoscape.shtml. Accessed 22 Feb 2022.
Cytoscape:http://www.softpedia.com/get/Science-CAD/Cytoscape.shtml. Accessed 22 Feb 2022., Cytoscape:http://www.softpedia.com/get/Science-CAD/Cytoscape.shtml. Accessed 22 Feb 2022.
A. Schmidt, Alexandra Tabakin, E. Singer, T. Choueiri, R. McKay (2021)
Next Steps: Sequencing Therapies in Metastatic Kidney Cancer in the Contemporary Era.
American Society of Clinical Oncology educational book. American Society of Clinical Oncology. Annual Meeting, 41
Y. Wi, A. Moon, M. Jung, Yeseul Kim, S. Bang, Kiseok Jang, S. Paik, Su-Jin Shin (2018)
Loss of Nuclear BAP1 Expression Is Associated with High WHO/ISUP Grade in Clear Cell Renal Cell Carcinoma
Journal of Pathology and Translational Medicine, 52
R. Samstein, Chung-Han Lee, A. Shoushtari, M. Hellmann, R. Shen, Y. Janjigian, D. Barron, A. Zehir, E. Jordan, A. Omuro, T. Kaley, S. Kendall, R. Motzer, A. Hakimi, M. Voss, P. Russo, J. Rosenberg, G. Iyer, B. Bochner, D. Bajorin, H. Al-Ahmadie, J. Chaft, C. Rudin, Gregory Riely, S. Baxi, A. Ho, R. Wong, D. Pfister, J. Wolchok, C. Barker, P. Gutin, C. Brennan, V. Tabar, I. Mellinghoff, L. Deangelis, C. Ariyan, N. Lee, W. Tap, M. Gounder, S. D’Angelo, L. Saltz, Z. Stadler, H. Scher, J. Baselga, P. Razavi, C. Klebanoff, R. Yaeger, N. Segal, G. Ku, R. DeMatteo, M. Ladanyi, N. Rizvi, M. Berger, N. Riaz, D. Solit, T. Chan, L. Morris (2019)
Tumor mutational load predicts survival after immunotherapy across multiple cancer types
Nature Genetics, 51
J. Bedke, V. Stühler (2020)
KEYNOTE-426
Der Urologe, 59
H. Togt (2003)
Publisher's Note
J. Netw. Comput. Appl., 26
(GEPIA: https://www.gsea-msigdb.org/gsea/index.jsp. Accessed 07 Jun 2022.)
GEPIA: https://www.gsea-msigdb.org/gsea/index.jsp. Accessed 07 Jun 2022.
GEPIA: https://www.gsea-msigdb.org/gsea/index.jsp. Accessed 07 Jun 2022., GEPIA: https://www.gsea-msigdb.org/gsea/index.jsp. Accessed 07 Jun 2022.
S. Williamson, A. Gill, P. Argani, Ying-bei Chen, L. Egevad, G. Kristiansen, D. Grignon, O. Hes (2020)
Report From the International Society of Urological Pathology (ISUP) Consultation Conference on Molecular Pathology of Urogenital Cancers: III: Molecular Pathology of Kidney Cancer.
American Journal of Surgical Pathology
Junlong Wu, Wenhao Xu, Yu Wei, Y. Qu, Hai-Liang Zhang, D. Ye (2018)
An Integrated Score and Nomogram Combining Clinical and Immunohistochemistry Factors to Predict High ISUP Grade Clear Cell Renal Cell Carcinoma
Frontiers in Oncology, 8
(2022)
IMMPORT database
H. Deng, Yi Zhao, Xiuyu Cai, Hualin Chen, B. Cheng, R. Zhong, Feng Li, S. Xiong, Jianfu Li, Jun Liu, J. He, W. Liang (2022)
PD-L1 expression and Tumor mutation burden as Pathological response biomarkers of Neoadjuvant immunotherapy for Early-stage Non-small cell lung cancer: A systematic review and meta-analysis.
Critical reviews in oncology/hematology
(Cibersort: https://cibersortx.stanford.edu/. Accessed 16 May 2022.)
Cibersort: https://cibersortx.stanford.edu/. Accessed 16 May 2022.
Cibersort: https://cibersortx.stanford.edu/. Accessed 16 May 2022., Cibersort: https://cibersortx.stanford.edu/. Accessed 16 May 2022.
(FunRich3.1.3: http://www.funrich.org/. Accessed 27 Feb 2022.)
FunRich3.1.3: http://www.funrich.org/. Accessed 27 Feb 2022.
FunRich3.1.3: http://www.funrich.org/. Accessed 27 Feb 2022., FunRich3.1.3: http://www.funrich.org/. Accessed 27 Feb 2022.
Wangli Mei, L. Jin, Bihui Zhang, Xianchao Sun, Guosheng Yang, Sheng Li, L. Ye (2022)
Computer classification and construction of a novel prognostic signature based on moonlighting genes in prostate cancer
Frontiers in Oncology, 12
(IMMPORT database: https://www.immport.org/home. Accessed 7 Feb 2022.)
IMMPORT database: https://www.immport.org/home. Accessed 7 Feb 2022.
IMMPORT database: https://www.immport.org/home. Accessed 7 Feb 2022., IMMPORT database: https://www.immport.org/home. Accessed 7 Feb 2022.
M. Kanehisa, S. Goto (2000)
KEGG: Kyoto Encyclopedia of Genes and Genomes
Nucleic acids research, 28 1
(GSE53757: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53757. Accessed 2 Jan 2022.)
GSE53757: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53757. Accessed 2 Jan 2022.
GSE53757: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53757. Accessed 2 Jan 2022., GSE53757: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53757. Accessed 2 Jan 2022.
(2022)
GEPIA
(VENNY: https://bioinfogp.cnb.csic.es/tools/venny/. . Accessed 9 Jan 2022.)
VENNY: https://bioinfogp.cnb.csic.es/tools/venny/. . Accessed 9 Jan 2022.
VENNY: https://bioinfogp.cnb.csic.es/tools/venny/. . Accessed 9 Jan 2022., VENNY: https://bioinfogp.cnb.csic.es/tools/venny/. . Accessed 9 Jan 2022.
M. Kanehisa, Miho Furumichi, Yoko Sato, Masayuki Kawashima, Mari Ishiguro-Watanabe (2022)
KEGG for taxonomy-based analysis of pathways and genomes
Nucleic Acids Research, 51
(Motzer RJ, Choueiri TK, McDermott DF, Powles T, Vano YA, Gupta S, Yao J, Han C, Ammar R, Papillon-Cavanagh S, et al. Biomarker analysis from CheckMate 214: nivolumab plus ipilimumab versus sunitinib in renal cell carcinoma. J Immunother Cancer. 2022;10(3):e004316. 10.1136/jitc-2021-004316.)
Motzer RJ, Choueiri TK, McDermott DF, Powles T, Vano YA, Gupta S, Yao J, Han C, Ammar R, Papillon-Cavanagh S, et al. Biomarker analysis from CheckMate 214: nivolumab plus ipilimumab versus sunitinib in renal cell carcinoma. J Immunother Cancer. 2022;10(3):e004316. 10.1136/jitc-2021-004316.
Motzer RJ, Choueiri TK, McDermott DF, Powles T, Vano YA, Gupta S, Yao J, Han C, Ammar R, Papillon-Cavanagh S, et al. Biomarker analysis from CheckMate 214: nivolumab plus ipilimumab versus sunitinib in renal cell carcinoma. J Immunother Cancer. 2022;10(3):e004316. 10.1136/jitc-2021-004316., Motzer RJ, Choueiri TK, McDermott DF, Powles T, Vano YA, Gupta S, Yao J, Han C, Ammar R, Papillon-Cavanagh S, et al. Biomarker analysis from CheckMate 214: nivolumab plus ipilimumab versus sunitinib in renal cell carcinoma. J Immunother Cancer. 2022;10(3):e004316. 10.1136/jitc-2021-004316.
Ilsiya Ibragimova, Marie Maradeo, Essel Dulaimi, P. Cairns (2013)
Aberrant promoter hypermethylation of PBRM1, BAP1, SETD2, KDM6A and other chromatin-modifying genes is absent or rare in clear cell RCC
Epigenetics, 8
S. Canete-Portillo, E. Velazquez, G. Kristiansen, L. Egevad, D. Grignon, A. Chaux, A. Cubilla (2020)
Report From the International Society of Urological Pathology (ISUP) Consultation Conference on Molecular Pathology of Urogenital Cancers V: Recommendations on the Use of Immunohistochemical and Molecular Biomarkers in Penile Cancer.
American Journal of Surgical Pathology
Wenxia Ma, Bin Wang, Yaping Zhang, Ziyue Wang, D. Niu, Siyu Chen, Zhirong Zhang, N. Shen, Weixia Han, Xiaoqin Zhang, R. Wei, Chen Wang (2019)
Prognostic significance of TOP2A in non-small cell lung cancer revealed by bioinformatic analysis
Cancer Cell International, 19
(STRING: https://string-db.org/. Accessed 15 Feb 2022.)
STRING: https://string-db.org/. Accessed 15 Feb 2022.
STRING: https://string-db.org/. Accessed 15 Feb 2022., STRING: https://string-db.org/. Accessed 15 Feb 2022.
(2020)
Urological Pathology CSoP. [Recent advances and expert consensus on molecular pathology of renal cell carcinomas
Yixin Yin, Siqi Li, Xin-qiang Liang, Ke-zhi Li, Ming-zhi Xie, Bang-li Hu (2022)
Construction and Validation of an Oxaliplatin-Resistant Gene Signature in Colorectal Cancer Patients Who Underwent Chemotherapy
Pharmaceuticals, 15
T. Powles, E. Plimack, D. Soulières, T. Waddell, V. Stus, R. Gafanov, D. Nosov, F. Pouliot, B. Melichar, I. Vynnychenko, S. Azevedo, D. Borchiellini, Raymond McDermott, J. Bedke, S. Tamada, L. Yin, Mei Chen, L. Molife, Michael Atkins, B. Rini (2020)
Pembrolizumab plus axitinib versus sunitinib monotherapy as first-line treatment of advanced renal cell carcinoma (KEYNOTE-426): extended follow-up from a randomised, open-label, phase 3 trial.
The Lancet. Oncology
Expert Group of G, Urological Pathology CSoP
D. Pardoll (2012)
The blockade of immune checkpoints in cancer immunotherapy
Nature Reviews Cancer, 12
G. Dunn, A. Bruce, H. Ikeda, L. Old, R. Schreiber (2002)
Cancer immunoediting: from immunosurveillance to tumor escape
Nature Immunology, 3
Yongshu Li, Fangfei Li, Feng Jiang, Xiaoqing Lv, Rongjia Zhang, Aiping Lu, Ge Zhang (2016)
A Mini-Review for Cancer Immunotherapy: Molecular Understanding of PD-1/PD-L1 Pathway & Translational Blockade of Immune Checkpoints
International Journal of Molecular Sciences, 17
(UALCAL: http://ualcan.path.uab.edu/analysis.html. Accessed 7 Mar 2022.)
UALCAL: http://ualcan.path.uab.edu/analysis.html. Accessed 7 Mar 2022.
UALCAL: http://ualcan.path.uab.edu/analysis.html. Accessed 7 Mar 2022., UALCAL: http://ualcan.path.uab.edu/analysis.html. Accessed 7 Mar 2022.
T. Honda, Y. Keith (2022)
Novel Insights Into the Immune-Regulatory Functions of Mast Cells in the Cutaneous Immune Response
Frontiers in Immunology, 13
(2022)
https
L. Stone (2017)
Kidney cancer: Activation of oncogenes driven by VHL loss in ccRCC
Nature Reviews Urology, 14
(2020)
[Recent advances and expert consensus on molecular pathology of renal cell carcinomas (2020 version)]
Zhonghua Bing Li Xue Za Zhi, 49
Jianpeng Gao, Shiying Huo, Yu Zhang, Zhenxiong Zhao, Hongda Pan, Xiao-wen Liu (2022)
Construction of ovarian metastasis‐related immune signature predicting prognosis of gastric cancer patients
Cancer Medicine, 12
Y. Tomita, S. Fukasawa, N. Shinohara, H. Kitamura, M. Oya, M. Eto, K. Tanabe, M. Saito*, G. Kimura, J. Yonese, M. Yao, H. Uemura (2019)
Nivolumab versus everolimus in advanced renal cell carcinoma: Japanese subgroup 3-year follow-up analysis from the Phase III CheckMate 025 study.
Japanese journal of clinical oncology, 49 6
A. Pandey (2022)
KEYNOTE-564: Adjuvant immunotherapy for renal cell carcinoma
Indian Journal of Urology : IJU : Journal of the Urological Society of India, 38
M. Abad, S. Calabuig-Fariñas, M. Mena, S. Torres-Martínez, C. González, José García, V. González-Cruz, C. Herrero (2022)
Programmed Death-Ligand 1 (PD-L1) as Immunotherapy Biomarker in Breast Cancer
Cancers, 14
W. Zou, Lieping Chen (2008)
Inhibitory B7-family molecules in the tumour microenvironment
Nature Reviews Immunology, 8

Publisher: Springer Journals
Copyright: Copyright © The Author(s) 2023
eISSN: 1471-2407
DOI: 10.1186/s12885-023-11150-4
Publisher site: See Article on Publisher Site

Abstract

Background Clear cell renal cell carcinoma (ccRCC) is the most common renal malignancy, although newly develop- ing targeted therapy and immunotherapy have been showing promising effects in clinical treatment, the effective biomarkers for immune response prediction are still lacking. The study is to construct a gene signature according to ccRCC immune cells infiltration landscape, thus aiding clinical prediction of patients response to immunotherapy. Methods Firstly, ccRCC transcriptome expression profiles from Gene Expression Omnibus (GEO) database as well as immune related genes information from IMMPORT database were combine applied to identify the differently expressed meanwhile immune related candidate genes in ccRCC comparing to normal control samples. Then, based on protein–protein interaction network (PPI) and following module analysis of the candidate genes, a hub gene cluster was further identified for survival analysis. Further, LASSO analysis was applied to construct a signature which was in succession assessed with Kaplan–Meier survival, Cox regression and ROC curve analysis. Moreover, ccRCC patients were divided as high and low-risk groups based on the gene signature followed by the difference estimation of immune treatment response and exploration of related immune cells infiltration by TIDE and Cibersort analysis respectively among the two groups of patients. Results Based on GEO and IMMPORT databases, a total of 269 differently expressed meanwhile immune related genes in ccRCC were identified, further PPI network and module analysis of the 269 genes highlighted a 46 genes cluster. Next step, Kaplan–Meier and Cox regression analysis of the 46 genes identified 4 genes that were supported to be independent prognosis indicators, and a gene signature was constructed based on the 4 genes. Furthermore, after assessing its prognosis indicating ability by both Kaplan–Meier and Cox regression analysis, immune relation of the signature was evaluated including its association with environment immune score, Immune checkpoint inhibi- tors expression as well as immune cells infiltration. Together, immune predicting ability of the signature was prelimi- nary explored. Ziwei Gui, Juan Du and Nan Wu contributed equally to this work. *Correspondence: Wenxia Ma mawenxia@sxmu.edu.cn Chen Wang wangchen@sxmu.edu.cn Full list of author information is available at the end of the article © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Gui et al. BMC Cancer (2023) 23:649 Page 2 of 24 Conclusions Based on ccRCC genes expression profiles and multiple bioinformatic analysis, a 4 genes containing signature was constructed and the immune regulation of the signature was preliminary explored. Although more detailed experiments and clinical trials are needed before potential clinical use of the signature, the results shall pro- vide meaningful insight into further ccRCC immune researches. Keywords Clear cell renal cell carcinoma (ccRCC), Immune response, LASSO analysis, Gene signature, Prediction biomarker Background yet. For example, clinical trial Checkmate025 showed Renal cell carcinoma has been the most common kidney that advanced ccRCC patients benefited from immuno - malignancy which comprises ccRCC and non clear cell therapy regardless of PD-L1 expression [18, 19]. Check- Renal Cell Carcinoma (nccRCC), and ccRCC accounts mate214 revealed that the objective remission rate (ORR) for approximately 70% ~ 75% of all the cases [1, 2]. Attrib- in immunotherapy received group was higher than that uting to the rapid development of molecular pathology, in sunitinib treatment group no matter PD-L1 expres- the genome mechanism behind ccRCC occurence has sion was higher or less than 1% [20, 21]. Meanwhile, Key- been gradual clear, of which short arm of chromosome note426 results also revealed that both PD-L1 negative 3 (3p) genes variations were showing defining character - and positive ccRCC patients benefited from pablizumab istic roles involving most importantly VHL gene known and axitinib combination therapy [22, 23]. by symbolic “double hit”. The aberrant change of VHL Meanwhile, as for the use of TMB in renal malignancy, including gene mutation and promoter methylation Robert M Samstein.et al. reported an pan-cancer MSK- causes the “first hit”, followed by "second hit", namely the IMPACT genome sequencing analysis based on 1600 3p chromosome deletion leading to tumor occurrence cases of cancer samples, of which 151 cases were RCC [3–5]. Besides VHL, other 3p gene variations were also samples, and the results showed that TMB was non sig- reported in ccRCC, for instance SETD2 [6], BAP1 [7] and nificant related with ccRCC overall survival [24]. And PBRM1 [8], which were reported to be survival related. since microsatellite instability (MSI-H and dMMR) is Regarding the clinical treatment, over the past two dec- very rare in RCC patients, MSI is not supported by evi- ades, molecular targeted therapies and immune therapies dence-based medicine yet to be an effective immune bio - have been showing great potential. Currently, at least 13 marker [14, 25]. drugs in 6 categories have been approved for metastatic Above all, ICIs has been an promising treatment in ccRCC, including VEGFR, mTORC1, c-Met, FGFR inhi- clinical ccRCC, but the effective biomarkers for immune bition, cytokines, and most recently anti PD-1/PD-L1 response prediction are still lacking, it is of great impor- immune checkpoint inhibitors (ICIs) which has been tance to keep exploring ccRCC genome and identifying a promising pillar of nowadays clinical treatment [4, 9]. new potential immune response biomarkers thus aiding Multi clinical trials most notably KEYNOTE-564 results more precise understanding of the disease and shed- supported ccRCC as immune sensitive and demonstrated ding promising light on further clinical immunotherapy the efficiency of ICIs in advanced patients clinical treat - application. ment using independent ICI therapies or ICI + TKI com- In the study, multi online ccRCC transcriptome pro- bination therapies [10]. files, our local hospital patients samples as well as vari - Although evidence-based medicine supported RCC ous bioinformatic analysis tools were combine used to as one of the malignancies that could benefit from neo- explore ccRCC genome data, identifying the survival adjuvant PD-1/PD-L1 blockade [11–13], the immune related meanwhile immune regulation associated genes response obviously vary among individuals indicating the and constructing a potential immune signature, fur- importance of usable and effective immune biomarkers ther necessary signaling mechanism was preliminary for selecting the potential patients that were most likely explored. The results shall provide meaningful insights to be able to benefit from the treatment [14]. Currently, to the unearth of potential new immune biomarkers PD-L1 expression, MSI and tumor mutation burden and shed promising light on further ccRCC immune (TMB) were three most widely used biomarkers in other researches. types of cancers, however, their use in ccRCC are still in dispute. Materials and methods As for PD-L1 expression, although it has been an effec - Data source: ccRCC transcriptome data from GEO database tive biomarker in other cancers for predicting immune From GEO online database, we widely screened ccRCC response [15–17], it’s function in ccRCC is inconclusive related profiles for exploring the changed genome Gui et al. BMC Cancer (2023) 23:649 Page 3 of 24 information in cancer comparing to normal renal sam- is short for Search Tool for the Retrieval of Interact- ples. The selection criteria of GEO profiles were set as: ing Genes was applied to construct the PPI network 1. the information of profiles were based on human tis - of above selected genes for observing the interaction sues (not animal models of any species); 2. the samples between individual genes. Further, based on PPI net- type was solid tissue (not tumor cell lines); 3. the contain- work, Molecular Complex Detection (MCODE) function ing data were mRNA/ cDNA/ transcriptome sequencing of Cytoscape3.6.0 software [33] was used to analyze the data; 4. covering both ccRCC cancer and normal renal promising function modules (gene clusters sharing simi- control samples; 5. each profile should contain at least 40 lar function) from the gene nest. or more samples. Further, Gene ontology analysis (GO) and Kyoto Ency- Based on above selection criteria, a total of five clopedia of Genes and Genomes (KEGG) [34] were ccRCC cDNA expression profiles were selected, includ - used to annotate basic biological attributes of the list of ing four profiles namely GSE53000 [26], GSE53757 [27], genes in each module including their main cellular loca- GSE68417 [28] and GSE71963 [29] containing a total of tion, involved biological processes, molecular functions 186 cases of ccRCC samples and 108 normal kidney sam- and the signaling pathways they mainly enriched in. The ples were selected for further potential genes selection. module that was predicted to be most related with tumor Meanwhile, another datasets GSE22541 which includes immune modulation and possess the highest module 68 cases of ccRCC samples was applied as validating data score would be highly focused and identified as a poten - source. (Table S1 for detailed information of the profiles tial gene cluster for further analysis. including samples amount, contributors and accessed online website). Univariate survival combine with Cox regression analysis of immune related gene cluster for hub genes Data processing: identify the differently expressed Following the identification of the immune related gene meanwhile immune related genes in ccRCC comparing cluster, each gene in the module would be in succession to normal control brought for firstly univariate survival analysis by UAL - The GEO transcriptome data were used to 1. explore CAN [35] and GEPIA [36], which have been two effec - the differently expressed genes in ccRCC comparing to tive online services for survival analysis. Then, the genes normal kidney samples; 2. combine with IMMPORT that were supported by both univariate analysis methods immune database [30] for collaborate identify the differ - to be statistical significantly associated with ccRCC sur - ently expressed meanwhile immune regulation related vival would be processed for multivariate COX regres- genes. sion analysis based on TCGA ccRCC data using SPSS19.0 To reveal the aberrant differently expressed genes in analysis. Any gene that was indicated by all three analysis ccRCC comparing to normal control samples, four GEO to be associating with patients survival would be identi- profiles GSE53000, GSE53757, GSE68417 and GSE71963 fied as credible prognostic indicating hub genes and pro - were in succession analyzed with GEO2R which was pro- cessed for next step interpretation. vided pared with each GEO profile. The analysis criteria was set as adjusted P value < 0.05 meanwhile |log2FC|< 1, Estimation of hub genes’ physicochemical properties 1 ≤|log2FC|< 2, 2 ≤|log2FC|< 3 and |log2FC|≥ 3 respec- For understanding the basic information of the selected tively, thus the genes expression change distribution hub genes, ProtParam [37], ProtScale [38], and Human namely the genes that were < twofold, 2 ~ fourfold, Protein Atlas [39] were combine used. As for the physico- 4 ~ eightfold and > eightfold different in cancer versus chemical properties of genes, ProtParam and ProtScale normal control would be preliminary understood. were applied to understand the basic information of the Then, Venn diagram [31] would be used to identify the genes encoding proteins including the aminoacid compo- immune related genes from all the high level differently sition, estimated molecular weight and protein half life, expressed genes based on IMMPORT immune genes list, computed protein instability index and theoretical isoe- therefore, the genes that were preliminary supported to lectric point, as well as hydrophobicity and hydrophilicity be both aberrant changed expressed and immune regula- of proteins. tion related were selected as candidate genes for further Besides ProtParam and ProtScale, Human Protein Atlas analysis. is also an effective and well used online service for inter - preting certain proteins information, in the study, it was Protein–protein interaction (PPI) network construction applied to predict the cellular location of the selected hub and function module analysis of the candidate genes genes for the convenience of further clinical test. After identifying the differently expressed meanwhile Additionally, UALCAN as well as GEPIA, which immune related candidate genes, STRING [32], which have been two resourceful web services constructed Gui et al. BMC Cancer (2023) 23:649 Page 4 of 24 Association between the selected hub genes expression based on TCGA and GTEx programs were in succes- and ccRCC clinical pathological features sion accessed to observe the expression difference of Ualcan has been a widely used integrated data-mining hub genes in broad-spectrum human cancers compar- platform for analyzing cancer transcriptome data, besides ing to corresponding normal control samples, espe- previous analysis of the expression difference of hub cially in ccRCC versus normal renal tissues. genes in broad-spectrum human cancers comparing to corresponding normal control samples, UALCAN was in addition applied to analyze the association between hub Quantitative real‑time PCR (QPCR) experiment genes’ expression and ccRCC clinical parameters includ- for evaluating the expression change of selected hub ing patients age, gender, tumor grade, stage and lymph genes in cancer vs normal tissues node metastasis, aiming to aiding the better understand- Besides above UALCAN as well as GEPIA online anal- ing of the potential biological roles of selected hub genes ysis, 30 pairs of ccRCC cancer tissues and adjacent in ccRCC. paracancerous normal renal tissues which were all col- lected from our local hospital were used for validating the expression changes of selected hub genes. All the patients tissues were collected from surgeries at local Other genetic alterations of the selected hub genes as well hospital General Surgery Department and sent for as their potential related signaling pathways analysis pathology examination then being long-term stored Besides the mRNA expression difference as well as at Pathology Department Biobank. The Informed con- association with clinical pathological parameters, sent from the patients as well as the approval by the the PPI networks which were centered on each of the Hospital Institutional Board were both obtained (Sec- four selected hub genes were constructed in succes- ond Hospital of ShanXi Medical University, China). sion, for the purpose of aiding better understanding The mRNA of 30 pairs of ccRCC cancer and adjacent of hub genes’ potential biological roles as well as their normal renal tissues were extracted using RNAiso-Plus related signaling pathways by KEGG analysis in ccRCC (TAKARA, DaLian, China). And then1 μg extracted development [40, 41]. mRNA was used for cDNA synthesis using cDNA Moreover, other types of variations of the selected key synthesis kit (TAKARA, DaLian, China) following genes including mutation ratio, copy number variation, operating instruction. Further, qPCR was performed amplification and deletion ratio were explored based on on Roche Light Cycler z 480 and the primers of the cBioPortal database which has been an effective cancer tested hub genes used during the process were listed genomics data website covering more than 2,8000 cancer as below: samples. After logging into the cBioPortal website, the MMP9: “cancer types summary” module of “quick search” section Former: AGA CCT GGG CAG ATT CCA AAC was used for exploring the genetic alteration character- Reverse: CGG CAA GTC TTC CGA GTA GT istics of previous selected hub genes in various cancer NFKB1: types, especially ccRCC. Former: AGC ACG ACA ACA TCT CAT T Reverse: CAG GCA CAA CTC CTT CAT IRF7: Construction and clinical features analysis of an prognosis Former: CCC ACG CTA TAC CAT CTA CCT related immune gene signature Reverse: GAT GTC GTC ATA GAG GCT GTTG To maximum the clinical utilization of hub gene indi- HMOX1: cators that were selected based on above processes, an Former: TGC CAG TGC CAC CAA GTT CAAG prognosis-related immune signature was constructed Reverse: TGT TGA GCA GGA ACG CAG TCTTG using LASSO algorithm performed with glmnet R pack- GAPDH: age based on TCGA ccRCC data, thus an unique regres- Former: AGA AGG CTG GGG CTC ATT TG sion coefficient was assigned to each gene indicator Reverse: AGG GGC CAT CCA CAG TCT TC which multiplies the gene expression. Based on the final The PCR cycling condition was set as: 95 °C 10 min score of each case calculated according to the gene signa- for 1 cycle; 95 °C 10 s, 58 °C 30 s, and 72 °C 34 s for ture, ccRCC patients were classified as high-risk and low- 35 cycles followed by the melting curve stage. And the risk groups (median score was set as cut off value). relative gene expression in each sample was recorded as Further, the clinical features of high-risk and low-risk the average 2^ − ΔΔCT calculation result of three rep- groups of patients were analyzed based on TCGA data licates. Further, T-test was used for detailed statistical which contains resourceful ccRCC samples, despite the analysis. P < 0.05 was considered statistically significant. censoring data, an effective pool of over 533 cases of Gui et al. BMC Cancer (2023) 23:649 Page 5 of 24 Association between the gene signature and ccRCC ccRCC patients information were applied for preliminary estimated environment immune score observing the clinical features association of the con- ESTIMATE, which is short for Estimation of Stromal structed signature. and Immune cells in Malignant Tumors using Expression data and has been a well accepted cancer immune evalu- Preliminary prognosis validation of the gene signature ation tool in multiple cancers was applied to estimate the To validate the survival relationship of last step con- immune score of ccRCC samples based on TCGA genes structed gene signature, a series of methods were com- expression data. The correlation between the signature bine used to analyze the TCGA ccRCC data including and ESTIMATE algorithm based ccRCC immune score, firstly Kaplan–Meier survival which was used to com - stromal score as well as tumor purity were evaluated pare the survival difference between high-risk and low- using R package for aiding the validation of immune rela- risk groups of patients, then AUC curve was performed tionship of the constructed gene signature. to observe the 1, 3 and 5 years survival prediction ability of the signature. Further, univariate as well as multivari- Correlation between gene signature and the expression ate Cox regression analysis were applied for testing the of immune checkpoint inhibitors independence survival prediction ability of the signa- Immune checkpoints have been showing inspiring drug ture together with other well accepted prognosis related targeting effects in multiple cancers by reversing the clinical parameters. Furthermore, a nomogram combin- tumor immuno suppressive microenvironment, and ing the signature and these validated clinical parameters expression of immune checkpoints especially PD-L1, was constructed for together evaluating clinical ccRCC CTLA4, TIGIT, TIM-3 and LAG-3 have been well patients prognosis, all based on the TCGA over 533 cases accepted as clinical biomarkers for selecting potential of ccRCC patients information. cancer patients that were most likely to benefit from Moreover, an independent GEO profile different from immunotherapy. Therefore, the association between the the four profiles that were used to identify the hub genes gene signature and expression level of these immune and construct the gene signature, namely GSE22541 was checkpoints in TCGA ccRCC samples were assessed in additionally applied to perform the Kaplan–Meier sur- the study, as well as the comparison of expression differ - vival as well as ROC curve analysis, for the purpose of ence of these immune checkpoints between high-risk and validating the prognosis correlation of the gene signature. low-risk groups of ccRCC patients. Gene set enrichment analysis (GSEA) of two ccRCC patients Evaluation of relationships between gene signature and 22 subgroups based on gene signature tumor infiltrating immune cells (TICs) After preliminary prognosis relationship analysis, the For characterizing the microenvironment immune land- gene signature was next step processed for further scape between high-risk and low-risk groups of ccRCC immune association evaluating. Based on the constructed patients, CIBERSORT algorithm [44] was performed gene signature, TCGA ccRCC patients were divided as to calculate the relative contents of 22 TICs based on high-risk and low-risk subgroups, for determining how TCGA profiles data, followed by analyzing the relation - the immunological pathways and corresponding immune ship between the 22 TICs and gene signature. Further, genes differ between the two ccRCC subgroups, GSEA the survival analysis of 22 TICs especially the ones that [42] was performed for signaling enrichment analysis, relates with gene signature were conducted for identi- and the threshold was set as P < 0.05. fying the specific immune cell infiltration that impacts patients prognosis. Difference of immunogenic cell death (ICD) between high‑risk and low‑risk groups of ccRCC patients Statistical analysis ICD has been gradually accepted as a form of regulated Statistical analysis was performed using SPSS 26.0. biological cell death meanwhile supported by evidence- For the enumeration data including QPCR experiment based medicine to be able to trigger cellular adaptive revealing the expression of different genes in ccRCC can - immune response through the emission of damage asso- cers vs normal control samples, the data were analyzed ciated molecular patterns (DAMPs), thus potentially using t-test. As for the measurement data for instance contributing to clinical immunotherapy. In the study, the association between constructed gene signature and the expression distribution of 32 ICD related genes [43] ccRCC clinical parameters, the data were analyzed by which were identified based on literature studies were χ test. And for the correlation analysis, for instance the explored in high-risk and low-risk groups of ccRCC correlation between gene signature and immune check- patients for preliminary evaluating the immune status points expression, the data were analyzed by Spearman difference between the two groups patients. Gui et al. BMC Cancer (2023) 23:649 Page 6 of 24 analysis. Meanwhile, for the survival analysis, Kaplan– promising gene clusters. The first gene cluster posses Meier were performed. p < 0.05 was considered statisti- the highest computed module score and contains 46 cally significant. genes with a big portion of them predicted to be related with immune system modulation (Fig. 2B). Meanwhile, Results the second and third gene modules contain 25 and 29 ccRCC transcriptome data identified 269 high level genes respectively, and genes were mostly related with differently expressed meanwhile immune related genes CXCR4, PI3K, EGF and mTOR related signaling path- in cancer versus normal renal samples ways (Fig. 2C, D). Four GEO profiles GSE53000, GSE53757, GSE68417 and Given the first module gene cluster possess the high - GSE71963 were combine applied to explore the aber- est score and a big percentage of containing genes were rant differently expressed genes in ccRCC comparing to immune system related which shows more potential for normal renal samples. And in GSE53000, a total of 5559 further clinical immune indicators selection, the 46 genes genes were identified to be differently expressed includ - in the gene cluster were mainly focused for next step ing 4270 genes with the expression change ≤ twofold, analysis. 1028 genes that were 2 ~ fourfold, 180 genes that were 4 ~ eightfold and 81 genes whose expression were > eight- Kaplan–Meier combine with Cox regression analysis fold in ccRCC comparing to normal renal samples of cluster genes identified 4 ccRCC prognosis related hub (Fig. 1A). And in GSE53757, a total of 28021 genes were genes identified, and the number was 21367, 4857, 1195 and Genes should be of more potential if they were both 602 in ≤ twofold, 2 ~ fourfold, 4 ~ eightfold and > eight- immune regulation and survival related, and promising fold groups respectively (Fig. 1B). In GSE68417, a total of immune biomarkers should also be prognosis related for 10150 genes were identified, and the number was 8276, potential clinical medical drug targeting use. To further 1425, 290 and 159 in each group (Fig. 1C). Meanwhile, in analyze survival relationship of the 46 selected immune GSE71963, a total of 10744 genes were identified, and the related candidate genes, univariate survival analysis number was 6266, 3120, 841 and 517 genes in each group including UALCAN and GEPIA, as well as multivariate respectively (Fig. 1D, Table S2). Cox regression analysis were in succession performed, Considering the feasibility of further clinical medical and the results supported four genes, namely MMP9 use, we mainly focused on the high differently expressed (Fig. 3A), NFKB1 (Fig. 3B), IRF7 (Fig. 3C) and HMOX1 genes (at least > fourfold in cancer vs. normal). As a (Fig. 3D) as independent prognostic indicators in ccRCC, result, besides the genes that were shared in multiple all four genes not only relate with patients overall sur- profiles, the analysis of 4 GEO profiles indicated a total vival but also progress free survival indicating their high of 3095 genes that were high level changed expressed in potential in clinical medical use (Table 1). ccRCC comparing to normal control samples (Fig. 1E). Further, the immune related genes list was obtained from Basic physicochemical properties of the 4 selected hub IMMPORT immune database, and Venn diagram analy- genes sis result identified 269 genes from the 3095 genes that Basic physiochemical properties of MMP9, NFKB1, IRF7 were both high level expression changed and immune and HMOX1 were preliminary interpreted before deeper regulation related for next step analysis (Fig. 1F, detailed scientific use of them (Table 2). As for MMP9, which is 269 genes information was listed in Table S3). a member of matrix metalloproteinase (MMP) family, locates in 20q13.12 and encodes a protein composed of PPI network of 269 genes highlighted a 46 707 amino acids with an estimated molecular weight of genes‑containing immune relating gene cluster 78.5KD. The theoretical isoelectric point of the protein The PPI network of 269 differently expressed mean - is estimated to be 5.69 and instability index to be 41.10, while potentially immune related genes was constructed meanwhile, the grand average of hydrophobic value of (Fig. 2A), and based on the network we identified three the protein is -0.394 indicating MMP9 works as a cellular (See figure on next page.) Fig. 1 Differential expression genes in ccRCC comparing to normal renal samples identified from GEO profiles. Four GEO profiles (A) GSE53000, (B) GSE53757, (C) GSE68417 and (D) GSE71963 were accessed to identify differently expressed genes in ccRCC vs normal renal samples, and based on these profiles, the up-regulated (right side) and down-regulated (left side) differential expression genes in ccRCC were identified. The genes were then divided into four groups based on the expression difference level as: < twofold genes (orange-colored spots), 2 ~ fourfold genes (red-colored spots), 4 ~ eightfold genes (green-colored spots) and > eightfold genes (black-colored spots). E The intersection of the genes in four GEO profiles for revealing the genes that were shared in different profiles. F The intersection of differential expressed genes revealed by GEO profiles and immune related genes from IMMPORT database, thus the genes that were both differential expressed and immune related were revealed Gui et al. BMC Cancer (2023) 23:649 Page 7 of 24 Fig. 1 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 8 of 24 unstable and hydrophilic protein which locates in cellular associated with the development of heme oxygenase1 cytoplasm or to be secreted in the extracellular region, deficiency and pulmonary disease, as well as chronic and the related signaling pathways include collagen chain obstructive (Fig. 3H). trimerization and apoptotic pathways in synovial fibro - blasts (Fig. 3E). Validation of the changed expression of selected hub NFKB1, which is short for Nuclear Factor Kappa B genes in ccRCC versus normal renal tissues Subunit 1 locates in 4q24 and encodes a protein com- Although the four selected hub genes were obtained from posed of 968 amino acids and estimated to be weighting the differently expressed gene clusters analyzed based on 105KD with computed theoretical isoelectric point as GEO data from the beginning, after preliminary inter- 5.20 and instability index as 38.15. Meanwhile, the grand pretation of the basic physicochemical information, it’s average of hydrophobic value of protein is -0.339 indicat- necessary to validate each of the gene’s aberrant changed ing NFKB1 to be cellular stable and hydrophilic. NFKB1 expression in ccRCC comparing to normal renal samples is predicted to locates in nucleoplasm and cytoplasm, it individually. In the study, both online analysis as well has been reported as a transcription regulator that could local hospital samples were used for detecting the genes be activated by various cellular stimuli such as cytokines, expression level. ultraviolet irradiation, and bacterial or viral products. Firstly, two analysis databases including UALCAN Activated NFKB1 translocates into cell nucleus and stim- and GEPIA were used, and the results revealed that as ulates the expression of genes involved in various biologi- for MMP9 and IRF7 genes, they gain of expression in cal functions (Fig. 3F). most of human cancers (Fig. 4A, D). And as for NFKB1 IRF7 is short for Interferon Regulatory Factor 7 and it’s and HMOX1, their expression vary in different cancers a member of the interferon regulatory factor (IRF) family, (Fig. 4G, J), although in ccRCC, all four genes were indi- locating in 11p15.5 and encoding a protein composed of cated to be statistical significantly up regulated in cancers 503 amino acids including 56 negatively charged amino comparing to normal renal tissues (Fig. 4B, E, H, K). acid residues (ASP + Glu) and 49 positively charged Then, the result of qRT-PCR experiment which was amino acid residues (Arg + Lys). The estimated protein conducted using 30 local hospital ccRCC and paired molecular weight is 54.2KD with theoretical isoelectric normal renal tissues also supported the aberrant gain point computed to be 5.89. Meanwhile, the estimated of expression of all four genes including MMP9, IRF7, instability index of the protein is 63.17 and grand aver- NFKB1 and HMOX1 in ccRCC (Fig. 4C, F, I, L). age of hydrophobic value is -0.367, the cellular location of the gene is predicted to be in nucleoplasm or cytoplasm Correlation analysis between the selected hub genes (Fig. 3G). and ccRCC clinical parameters Meanwhile, HMOX1 is short for Heme Oxygenase1, For analyzing the association between MMP9, IRF7, and locates in 22q12.3, encoding a protein composed of NFKB1 and HMOX1 expression and ccRCC clinical 288 amino acids including 35 negatively charged amino parameters, UALCAN online platform was used. And acid residues (ASP + Glu) and 36 positively charged the result revealed an inspiring fact that although all four amino acid residues (Arg + Lys). The estimated protein genes were indicated to express higher in cancer compar- molecular weight is 32.8KD with theoretical isoelec- ing to normal samples, the four genes tend to play oppo- tric point computed as 7.89. Moreover, the estimated site roles in cancer development. To be more specific, the instability index of the protein is 60.81 and grand aver- expression of MMP9 and IRF7 genes which were sup- age of hydrophobic value is -0.427 indicating HMOX1 ported by previous survival analysis to be related with works as a cellular unstable and hydrophilic protein worse patients prognosis tend to be higher as the cancer which is consistent with the ProtScale analysis result stage and grade advancing, more obviously, both of the of HMOX1 structure showing that the protein harbors genes express higher in patients with node metastasis more hydrophilic regions than hydrophobicity regions. (Fig. 5A-J). HMOX1 is predicted to locates in cellular Golgi appa- Meanwhile, as for NFKB1 and HMOX1 genes which ratus and plasma membrane, it has been reported to be were indicated to relate with better patients prognosis, (See figure on next page.) Fig. 2 PPI network construction of 269 differential expressed meanwhile immune related genes in ccRCC and function modules analysis. (A) Based on four GEO profiles as well as IMMPORT datasets, 269 differential expression meanwhile immune related genes in ccRCC were identified, and the PPI network of these 269 genes was constructed. And based on the PPI network, (B) the first, (C) second and (C) third genes function modules were analyzed, each module was shown with a diagrammatic sketch (left diagram) and the detailed module information (right table) including the computed module score, module description and detailed involving genes. (*The first module with the highest module score meanwhile immune regulation related was focused for further analysis) Gui et al. BMC Cancer (2023) 23:649 Page 9 of 24 Fig. 2 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 10 of 24 Fig. 3 Survival analysis and basic physicochemical properties exploration of four selected signature comprised genes. The overall survival (left) and disease free survival (right) analysis in ccRCC patients, including (A) MMP9 gene, (B) NFKB1 gene, (C) IRF7 gene and (D) HMOX1 gene. The predicted cellular location (left) and computed hydrophility/hydrophobicity property of (E) MMP9 protein, (F) NFKB1 protein, (G) IRF7 protein and (H) HMOX1 protein the expression tend to keep decreasing as the cancer the difference was not statistical significant presumably stage and grade advancing, and their expression were due to the limited patients cases number in N1 group lower in patients with lymph node metastasis, although (Fig. 5K-T). Gui et al. BMC Cancer (2023) 23:649 Page 11 of 24 Table 1 Univariate combine with multivariate Cox Regression analysis result of the 4 hub genes used for signature construction OSCC parameters P Value B value HR 95% CI Univariate analysis Multivariate analysis UALCAN GEPIA MMP9 0.001 0.016 0.002 0.256 1.292 1.009–1.519 NFKB1 0.013 2.9E-05 0.004 -0.216 0.805 0.696–0.932 IRF7 < 0.0001 9E-04 0.022 0.077 1.080 1.011–1.154 HMOX1 0.034 0.00062 < 0.001 -0.222 0.801 0.708–0.906 Table 2 Basic physicochemical properties of the 4 hub genes used for gene signature construction Gene Property MMP9 NFKB1 IRF7 HMOX1 Formula C H N O S C H N O S C H N O S C H N O S 3517 5298 958 1035 28 4643 7343 1271 1458 33 2418 3740 678 710 19 1475 2323 405 427 8 Molecular Weight 78.46KD 105.36KD 54.28KD 32.82KD Number of amino acids 707AA 968AA 503AA 288AA Theoretical pI 5.69 5.20 5.89 7.89 Aliphatic index 65.13 84.74 72.07 83.02 Hydrophobic value -0.394 -0.339 -0.367 -0.427 Estimated protein half life 30 h 30 h 30 h 30 h Instability index 42.10 38.15 63.17 60.81 Selected hub genes mainly enriched signaling pathways Besides mRNA expression and potential related signal- and other types of genetic alteration analysis ing pathways, other genetic alterations including mutation For preliminary exploring the mechanism behind the dif- ratio, protein structure variant and copy number varia- ferent roles of the four genes in ccRCC, the PPI network tion of the four genes were preliminary explored based on that were centered on each of the four genes including cBioPortal database. However, limited inspiration could MMP9 (Fig. 6A), IRF7 (Fig. 6C), NFKB1 (Fig. 6E) and be achieved based on this part of analysis considering the HMOX1 (Fig. 6G) were constructed, followed by GO/ fact that only few ccRCC samples were included in cBio- KEGG analyzing the biological functions as well as sign- Portal database, although various types of MMP9 (Figure aling pathways the fours genes and their surrounding S1A), IRF7 (Figure S1B), NFKB1 (Figure S1C) and HMOX1 partner genes most enriched. (Figure S1D) genes alteration were observed in different And the results revealed that the four genes potentially human cancers indicating the potential different functions evolved in different signaling in cancer development, as these genes play in human cancers. for MMP9, it mainly evolves in proteoglycan syndecan as well as integrin mediated signaling pathways (Fig. 6B), and Construction of a 4 genes containing ccRCC prognosis IRF gene was indicated to be relate with immune and inter- related immune gene signature and clinical features feron related biological functions (Fig. 6D). Meanwhile, analysis NFKB1 gene was related with the canonical NF-kB signal- To maximum the clinical prediction value of the four ing as well as Aurora A, CD40/40L and endogenous TLR selected genes, Cox-proportional hazards analysis signaling pathways (Fig. 6F). And HMOX1 was indicated to based on LASSO algorithm was applied to construct a be associated with HIF-1a gene related cancer hypoxia and MMP9, NFKB1, IRF7 and HMOX1 four genes contain- oxygen homeostasis regulation (Fig. 6H). ing signature which weights the normalized expression (See figure on next page.) Fig. 4 The differential expression of four selected signature comprised genes in ccRCC included human cancers. UALCAN prediction of (A) MMP9 gene, (D) IRF7 gene, (G) NFKB1 gene and (J) HMOX1 gene expression in broad spectrum human cancers. GEPIA analysis of (B) MMP9 gene, (E) IRF7 gene, (H) NFKB1 gene and (K) HMOX1 gene in ccRCC comparing to normal renal samples. QPCR experiment using local hospital ccRCC samples for validating the changed expression of (C) MMP9 gene, (F) IRF7 gene, (I) NFKB1 gene and (L) HMOX1 gene in ccRCC comparing to normal renal samples Gui et al. BMC Cancer (2023) 23:649 Page 12 of 24 Fig. 4 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 13 of 24 Fig. 5 The association between four selected hub genes expression and ccRCC clinical parameters. The association between MMP9 expression and ccRCC (A) patients gender, (B) age, (C) cancer stage, (D) cancer grade and (E) lymph node metastasis. The association between IRF7 expression and ccRCC (F) patients gender, (G) age, (H) cancer stage, (I) cancer grade and (J) lymph node metastasis. The association between NFKB1 expression and ccRCC (K) patients gender, (L) age, (M) cancer stage, (N) cancer grade and (O) lymph node metastasis. The association between HMOX1 expression and ccRCC (P) patients gender, (Q) age, (R) cancer stage, (S) cancer grade and (T) lymph node metastasis. (* p < 0.05, **p < 0.01, ***p < 0.001. The first layer * which is right above the error bar representing comparison to normal group, and the above layers * which were above a secondary line represent the comparison between corresponding groups that were covered by the line) Gui et al. BMC Cancer (2023) 23:649 Page 14 of 24 level of each gene to the regression coefficient of multi - a point scale was assigned for each variable, the sum of variate Cox regression analysis. And the result revealed all the variables points equal to the final score of each a formula: Risk Score = 0.256 * expression (MMP9)— patient, and the survival could be predicted by drawing 0.222 * expression (HMOX1)—0.216 * expression a vertical line from the total point axis downward to the (NFKB1) + 0.077 * expression (IRF7) as the best signature outcome axis (Fig. 7F). for combining the four differently expressed meanwhile Secondly, besides above TCGA data, an independent immune related hub genes (Fig. 7A, B). GEO ccRCC cDNA expression profile GSE22541 was Based on the signature, the risk score for each patient also included for validating the prognosis correlation of was calculated followed by the patients being categorized the gene signature. After observing the detailed expres- into high-risk or low-risk groups according to the median sion of all the four genes in the GSE22541 datasets (Fig- risk score which was set as the cut off point for the signa - ure S2E), the patients samples were divided as high and ture (Fig. 7C). Further, the association between the gene low risk groups based on the constructed gene signature signature and ccRCC clinical features was preliminary score (Figure S2A, S2B). Afterwards, Kaplan–Meier sur- analyzed, which result revealed that higher risk score vival (Figure S2C) as well as ROC curve (Figure S2D) was positively related with older age (> 45 years old) and were also conducted, and the results supported the high more advanced T, N and M stage, meanwhile, the low risk group of patients had a statistical significantly worse risk group of patients were tend to be younger (≤ 45 years prognosis than low risk group patients. old) female with lower TNM stage (Table 3). High‑risk group of ccRCC cases were more enriched High risk score based on the gene signature indicated in immune related phenotype worse ccRCC patients prognosis After preliminary demonstrating the association between All four genes used to construct the gene signature constructed gene signature and ccRCC prognosis, the were previously supported to be high level differently influence of gene signature on cancer immune profiles expressed in ccRCC comparing to normal renal samples, was to be investigated. And in the first step, GSEA was but changed gene expression doesn’t equal to survival utilized to analyze the immune-related biological pro- association. To validate the survival relationship of the cesses linked to the signature, and the analysis result gene signature, two independent analysis methods were showed that the high-risk group cases were significantly performed. enriched in multiple biological processes, of which Firstly, a series of analysis were applied using TCGA 4 immune-related processes were identified includ - ccRCC data including at first Kaplan–Meier survival ing HUMORAL_IMMUNE_RESPONSE (NES = 1,733, analysis, which result revealed that the high risk group of Nominal p value = 0.0), REGUL ATION_OF_T_CELL_ patients had a statistical significantly worse overall sur - MIGRATION (NES = 1.762, Nominal p value = 0.0), REG- vival than their low risk counterparts (Fig. 7D). Then the UL ATION_OF_LYMPHOCYTE_CELL_MIGR ATION ROC curve showed that the area under the ROC curve (NES = 1.743, Nominal p value = 0.003), POSITIVE_REG- (AUC) of gene signature for overall survival was 0.747 ULATION_OF_IMMUNOGLOBULIN_PRODUCTION at 1 year, 0.696 at 3 year and 0.705 at 5 years (Fig. 7E). (NES = 1.754,Nominal p value = 0.0). Meanwhile, the Meanwhile, univariate Kapkan-Meier survival as well low-risk group cases were not indicated to be enriched in as multivariate Cox regression analysis were applied for any immune-related biological processes (Fig. 7G). testing the survival prediction ability of the signature, and the results supported the risk score calculated based High‑risk and low‑risk groups of ccRCC patients revealed on the gene signature works as an independent prog- disparate ICD expression levels nosis indicator for ccRCC patients together with some Besides GSEA immune phenotype enrichment analysis, other well accepted prognosis related clinical param- given the significant roles of ICD in antitumor immuno - eters including patient T and M stage (Table 4). Further, logical responses, the connection between gene signature a nomogram was constructed and and in the nomogram, and ICD related genes were evaluated for additionally (See figure on next page.) Fig. 6 PPI network centered on four selected hub genes and GO/KEGG analysis of their enriched biological pathways. (A) The PPI network which is centered on MMP9 gene for analyzing (B) the main biological signaling pathways MMP9 and its connected genes mainly participated in. (C) The PPI network which is centered on IRF7 gene for analyzing (D) the main biological signaling pathwaysIRF7 and its connected genes mainly participated in. (E) The PPI network which is centered on NFKB1 gene for analyzing (F) the main biological signaling pathways NFKB1 and its connected genes mainly participated in. (G) The PPI network which is centered on MMP9 gene for analyzing (H) the main biological signaling pathways MMP9 and its connected genes mainly participated in. (KEGG software analysis permitted by Kanehisa laboratory) Gui et al. BMC Cancer (2023) 23:649 Page 15 of 24 Fig. 6 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 16 of 24 Fig. 7 Construction of a four genes containing meanwhile immune and prognosis related ccRCC gene signature. LASSO analysis to calculate (A) the coefficient and (B) the likelihood deviance for constructing a suitable immune meanwhile prognosis related signature which was comprised of strictly calculated four genes. (C) TCGA ccRCC patients were divided into high-risk and low-risk groups based on the calculated signature score (the cut off value was set as the median signature score in all samples). (D) Survival analysis of the high-risk and low-risk groups of ccRCC patients. (E) ROC curve of the gene signature to predict ccRCC patients survival of 1 year, 3 years and 5 years respectively. (F) ccRCC patients prognosis prediction nomogram constructed based on genes signature and clinical parameters which were supported by Cox Regression to be independently related with patients survival. (G) Significant enrichment of immune-related phenotype including immune response and immune cells migration in high-risk group of ccRCC patients compared with that in low-risk group patients Gui et al. BMC Cancer (2023) 23:649 Page 17 of 24 Table 3 Association between the gene signature and ccRCC exploring the immune status in high-risk and low-risk clinical features groups of ccRCC patients. And the results revealed that the expression of a large portion of the 32 ICD related Parameters Gene signature P Value genes were statistical significantly different between the Low‑risk group High‑risk group two groups of patients indicating the diverse immune status in the microenvironment of two groups of patients Gender (Fig. 8A). female 112 (59.6%) 76 (40.4%) 0.001 male 154 (44.6%) 191 (55.4%) Age Risk score calculated based on the gene signature ≤ 45 38 (64.4%) 21 (35.6%) 0.018 associated with ccRCC estimated environment immune > 45 228 (48.1%) 246 (51.9%) score Race For further validating the immune association of the gene White 234 (50.6%) 228 (49.4%) 0.697 signature, ESTIMATE was performed to evaluate the Yellow 4 (50.0%) 4 (50.0%) immune, stromal score and tumor purity of ccRCC sam- Black 25 (44.6%) 31 (55.4%) ples. And the result revealed that although no significant T stage correlation was found between the gene signature and T1 164 (60.01%) 109 (39.9%) < 0.001 ccRCC stromal score, a mediate correlation was revealed T2 31 (44.9%) 38 (55.1%) between the risk score which was calculated based on the T3 69 (38.3%) 111 (61.7%) gene signature and tumor immune score as well as tumor T4 2 (18.2%) 9 (81.8%) purity (Fig. 8B). Meanwhile, the high risk-group patients N stage were tend to posses higher immune score and lower N0 115 (47.9%) 125 (52.1%) 0.023 tumor purity, indicating the immune targeting potential N1 3 (18.8%) 13 (81.3%) of the group of patients (Fig. 8C). M stage M0 231 (54.7%) 191 (45.3%) < 0.001 M1 22 (27.8%) 57 (72.2%) Multi immune checkpoints expressed higher in high risk group of ccRCC patients based on the gene signature Besides estimation of immune score, the association Table 4 Survival prediction value of the gene signature included between the gene signature and clinical promising ccRCC clinical parameters immune checkpoints including PD-L1, CTLA4, TIGIT, Clinical parameters P Value Exp (B) TIM-3 and LAG-3 were evaluated (Fig. 8D). And a median association was revealed between the gene sig- Univariate Multivariate analysis analysis nature and CTLA4 expression. Moreover, mild correla- tion was indicated between gene signature and two of Age < 0.001 0.026 1.627 (1.058–2.500) the immune checkpoints including LAG3 and TIGIT, Gender 0.693 - - meanwhile, no significant relation was found between Race 0.719 - - the signature and PD-L1 or TIM-3 expression (Fig. 8E). T stage < 0.001 0.018 1.360 (1.054–1.755) An inspiring fact was that all CTLA4, LAG3 and TIGIT N stage < 0.001 0.480 1.290 (0.636–2.615) tend to express higher in high-risk group of patients M stage < 0.001 < 0.001 2.794 (1.728–4.517) which was categorized based on the gene signature Signature risk score < 0.001 0.002 1.871 (1.299–2.693) (Fig. 8F), and the distribution shall be an additional sup- port besides above ESMINATE immune score evaluation result for indicating the immune targeting potential for this group of ccRCC patients. (See figure on next page.) Fig. 8 Correlation between gene signature and ccRCC immune microenvironment landscape. (A) Relative expression of ICD related genes in high-risk and low-risk groups of ccRCC patients. (B) Correlation between gene signature and ccRCC computed immune score, stromal score and tumor purity calculated using ESTIMATE algorithm. (C) Estimated immune score, stromal score and tumor purity distribution in high-risk and low-risk ccRCC groups respectively. (D) Association between gene signature and immune checkpoints expression. (E) Correlation between gene signature and PD-L1, LAG-3, TIGIT and CALT-4 expression respectively. (F) Relative expression of five immune checkpoints including PD-L1, LAG-3, TIGIT, TIM3 and CALT-4 expression in high-risk and low-risk ccRCC groups respectively Gui et al. BMC Cancer (2023) 23:649 Page 18 of 24 Fig. 8 (See legend on previous page.) Gui et al. BMC Cancer (2023) 23:649 Page 19 of 24 Evaluation of relationships between gene signature and 22 and stratify patients into different groups for increasing tumor infiltrating immune cells (TICs) the potential effectiveness of ICIs therapy. Considering Previous analysis supported that the constructed gene most biological cellular functions were performed by signature was related to immunity, so we carried out different types of cell proteins, in the study, we mainly analyses on 22 TICs whose distribution profiles were focused on the protein encoding genes that were aber- draw based on CIBERSORT algorithm to further study rant differently expressed meanwhile prognosis as well the interaction between the gene signature and ccRCC as immune related for series of analysis. Thus, we mainly immune microenvironment. And the correlation analy- focused on GEO transcriptome profiles for identifing sis result found four types of TICs to be related with the potential immune regulation related gene candidates. gene signature including plasma cells, activated CD4( +) Based on four different ccRCC cDNA expression pro - T memory cells, activated dendritic cells and resting files which were all selected by strict criteria as see in the mast cells (Fig. 9A-C). Materials and Methods part, we identified the differen - Further, the prognostic abilities of the 22 TICs were tial expression genes in ccRCC cancer vs. normal renal tested and the results revealed that of the four signature tissues and then divided them into 4 groups according to related TICs, CD4( +) T cell and resting mast cell were the difference level as < twofold, 2 ~ fourfold, 4 ~ eightfold able to predict ccRCC patients prognosis. The resting and > eightfold genes considering the potential unique and activated CD4( +) T memory cells played opposite functions and clinical use of each group, for example, roles in patients survival, namely the activated CD4( +) an interesting phenomenon has been discovered that T memory cells were related with worse patients survival the more genes expression difference are, the more their (Fig. 9D), meanwhile, the resting CD4( +) T memory cells cellular location tend to be far away from cell nuclear predicting better patients survival (Fig. 9E). Also, the [51–53]. For the feasibility of further clinical medical use, resting mast cells were correlated with positive patients in the selection process of candidate genes, we mainly prognosis (Fig. 9F). focused on the high level differently expressed namely Combining the analysis results, one inspiring deduc- at least > fourfold genes that were more convenient to tion could be draw that CD4( +) T memory cells and be tested by IHC experiment which has been a common resting mast cells not only are significantly related to the method in clinical pathology diagnosis, considering that gene signature but also predict ccRCC patients progno- the genes shall harbor more chance to be translated into sis, indicating these immune cells may play important clinical use if they are suitable to be tested by IHC. Fur- roles in the immune regulation of the gene signature in ther, the intersection between GEO selected high level ccRCC microenvironment. aberrant differently expressed genes and immune related gene list from IMMPORT database indicated 269 genes Discussion that were both high level expression changed in ccRCC ICIs has been an increasing rising up clinical method and and immune related as candidate genes for next step holds great promise for treating ccRCC [9], but the effec - analysis. tive biomarkers for predicting immune response are still As for the construction of multi genes containing sig- lacking, the well accepted immune prediction biomark- nature, LASSO algorithm has been widely accepted as ers in other cancers for instance PD-L1 expression, MSI an effective tool that is suitable to construct gene mod - status and TMB haven’t been supported thoroughly by els basing on large numbers of correlated covariate. evidence-based medicine to be effective in ccRCC [45, But instead of constructing an gene signature directly 46]. For the clinical benefit from ICI therapy, it is of great from the 269 candidate genes, we further in succession importance to keep exploring ccRCC genome and iden- performed module analysis as well as multiple survival tifying new potential biomarkers thus benefiting further analysis to scale down the candidate genes and identify clinical application of immunotherapy in the cancer. the promising “unique key genes” during ccRCC devel- In recent years, multiple genes containing signatures opment, and only used LASSO for estimating the coef- representative of caner immune status have been iden- ficient of signature genes. This is for the considering tified in several cancers besides ccRCC, including gene that the signature and consist genes should be of more signatures that were comprised of LncRNAs, miRNAs clinical potential if they were not only immune regula- or immune regulation related genes, and they have been tion but also survival related for clinical medical drug showing inspiring clinical effects [47–50]. Based on targeting use. Therefore, the PPI network of the 269 these reports, it’s of clinical feasibility to explore ccRCC genes was constructed followed by genes module analy- genome information and develop meaningful immune sis which highlighted a 46 genes containing cluster, and prediction models which were also prognosis related to further survival analysis including GEPIA and UAL- evaluate the immune status of ccRCC microenvironment CAN univariate survival as well as multivariate Cox Gui et al. BMC Cancer (2023) 23:649 Page 20 of 24 Fig. 9 Correlation between gene signature and 22 immune cells infiltration in ccRCC. (A) Relative distribution of 22 immune cells in high-risk and low-risk groups of ccRCC patients. (B, C) Correlation between gene signature and various immune cells infiltration in ccRCC. Association between (D) T cells CD4 memory activated, (E) T cells CD4 memory resting and (F) Mast cells resting microenvironment infiltration and ccRCC patients survival Gui et al. BMC Cancer (2023) 23:649 Page 21 of 24 Regression analysis of each of the 46 genes supported Although survival relation was an important part, four genes: MMP9, NFKB1, IRF7 and HMOX1 to be the main aim of the signature was for potential associated with patients survival and worked as inde- immune prediction. Immune escape has been one of pendent prognostic indicators in ccRCC development. the major characteristics in malignant tumors involv- Interestingly, no direct relationship has yet been dis- ing multiple probable mechanisms [54, 55], for exam- covered among the four genes. MMP9 is s a member ple the increasing immune suppressive cells including of matrix metalloproteinase (MMP) family which is Treg cells and tumour-associated macrophages (TAM) well known to be involved in the breakdown of extra- in tumor microenvironment [47], and the up-regu- cellular matrix during multiple normal physiological lated expression of immunosuppressive molecules and diseases processes, and MMP9 has been reported for instance cytotoxic T lymphocyte associated anti- to be able to degrades type IV and V collagen which gen-4 (CTLA-4), also decreasing expression of can- are important microenvironment elements. NFKB1 is a cer antigens which results the inactivation of tumor transcription regulator that could be activated by cel- killing CD8 + T cells [56–59]. Therefore, we explored lular stimuli such as cytokines, ultraviolet irradiation, the probable relation between the gene signature and bacterial and viral products, and inappropriate acti- immune suppressive mechanisms. And the results vation of the gene has been known to associate with revealed that the difference expression of ICD related a number of inflammatory diseases, while persistent genes in high-risk and low-risk groups of patients inhibition of NFKB1 leads to inappropriate immune cell which were categorized based on the constructed gene development or delayed cell growth. IFR7 is predicted signature, as well as the statistical significant corre- to locates in nucleoplasm and cytoplasm and it has lation between the signature and ESTIATE immune been reported to play roles in innate immune response score supported the signature was immune modula- against DNA and RNA viruses. Meanwhile, HMOX1 tion associated. Further, we investigated the expres- is predicted to locates in cellular Golgi apparatus and sion of immune checkpoints including PD-L1, CTLA4, plasma membrane and it has been reported to be TIGIT, TIM-3 and LAG-3 between the high-risk and associated with the development of heme oxygenase1 low-risk groups of patients, and the results showed the deficiency and pulmonary disease, as well as chronic high-risk patients had higher expression of CTLA4, obstructive. Moreover, the PPI network centered on the LAG3 and TIGIT than the low-risk patients indicat- four genes and following KEGG analysis also supported ing the immune targeting potential for this group of the independent roles of these genes in cancer devel- ccRCC patients. opment. The together identification of the four genes As the results of the relation analysis between the and a gene signature combine all of them indicating the signature and 22 TICs indicated that the signature was elaborate collaboration network of various genes in cel- significantly related with CD4( +) T memory cells and lular activities, opening up further cancer researches of resting mast cells infiltration, and not only the two unlimited possibilities. immune cells were related with the signature, but also Based on the selected four genes and coefficient calcu - they were able to predict ccRCC patients prognosis. lated with LASSO algorithm for each gene, an immune CD4( +) T memory cells have already been reported meanwhile prognosis related gene signature was con- to confer vital functions on malignancy immune reg- structed. Survival relationship validation including ulation, including participating in the activation of Kaplan–Meier survival, Cox proportional-hazards model CD8 + T and NK killing cells, involving in the tumour and ROC curve based on both TCGA data and an inde- immunological reactions [60, 61]. And mast cells were pendent GEO profile all supported the signature worked reported to be able to not only influence tumor expan - as an prognostic factor after combining the four genes in sion via inducing angiogenesis and changing tumor one equation, proving the effectiveness of apply the gene extracellular matrix composition, but also could influ - signature in ccRCC prognosis prediction. Since many ence the infiltration and activity of dendritic cells, clinical parameters especially tumor TNM stage as been tumor-associated macrophages and lymphocytes, pro- well known as critical survival related aspects, we pro- moting pro-inflammatory reactions in tumor micro - posed a nomogram assessment that combines the signa- environment [47, 62, 63]. The association between the ture and other clinical features. Although current result signature and immune checkpoints expression as well has not supported the signature to be a better prognosis as different immune cells infiltration, suggesting the factor than TNM stage, the construction of the nomo- stronger immunosuppressive environment in high-risk gram shall work as a complementary perspective on indi- groups of patients comparing to low-risk group, high- vidual tumour and aiding the comprehensive evaluation lighting the potential of this group to benefit from fur - of clinical ccRCC patients prognosis. ther clinical immunotherapy. Gui et al. BMC Cancer (2023) 23:649 Page 22 of 24 manuscript. As the corresponding author, both WM and CW had full access to Conclusion all data of the manuscript, CW made the final decision to submit the article for The present study defined a four genes containing signa - publication. All authors read and approved the final manuscript. ture based on ccRCC genes expression information, the Funding signature was not only closely associated with patients The work was supported by the China central government funds for guiding survival, but also immune regulation related. Multiple local scientific and technological development (YDZJSX2021A042), the fund in vitro experiments data analysis supported the asso- of Shanxi Medical key scientific research project (2021XM34) and the fund of Natural Science Foundation of ShanXi Province in China (201901D211498). ciation between signature and ccRCC microenviron- ment immune aspects including immune checkpoints Availability of data and materials expression and various types of immune cells infiltration. Publicly available datasets were analyzed in this study. The data can be found here: GSE53000:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 Although the current result is not yet enough to sup- 000. GSE53757:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 port the application of the signature in clinical medical 757. GSE68417:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE68 immunotherapy, rigorous prospective studies performed 421. GSE71963:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE71 on animal models as well as clinical trials are still needed, GSE22541:https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE22 541. the results shall provide meaningful insight into better All data generated or analyzed based on the online datasets and other experi- understanding of the disease and shed lighting on further ments during this study are included in this published article. ccRCC immune regulation researches. Declarations Abbreviations Ethnic approval and consent to participate ccRCC Clear cell renal cell carcinoma All of the local hospital ccRCC patients samples used for qPCR experiment GEO G ene Expression Omnibus were stored at the hospital biobank (Second Hospital of ShanXi Medical Uni- PPI Protein–protein interaction network versity). All the sample donors have been informed of the possibility that their ICIs I mmune checkpoint inhibitors donated tissues might be used for certain scientific research. The Informed TMB Tumor mutation burden consent from the patients were all obtained, and the paper agreement signed ICD Immunogenic cell death with the donors’ signatures were kept by the biobank committee. And the TIC Tumor infiltrating immune cells biobank ccRCC samples being used in this research was approval by both TAM Tumour-associated macrophages the biobank committee and Hospital institutional ethics committee (Second Hospital of ShanXi Medical University, ShanXi Province, China). And all experi- ment methods were carried out in accordance with relevant guidelines and Supplementary Information regulations or declaration of Helsinki. The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s12885- 023- 11150-4. Consent for publication Not applicable. Additional file 1: Supplementary Figure 1. Genetic alterations of four Competing interests hub genes based on cBioPortal dataset. Different types of (A) MMP9, (B) All of the authors agreed the publication of the paper and declare no conflicts IRF7, (C) NFKB1 and (D) HMOX1 variations including gene amplification, of interests. deletion, mutation and structural variants in various human cancers revealed by cBioPortal dataset. Author details Additional file 2: Supplementary Figure 2. GEO profiles validating the 1 Department of Pathology, Second Clinical Medical College of ShanXi Medical prognosis correlation of the constructed gene signature. (A) GSE22541 2 University, Tai Yuan City, ShanXi Province, China. Department of Anesthesiol- patients were divided into high-risk and low-risk groups based on the ogy, Second Hospital of ShanXi Medical University, Tai Yuan, ShanXi Province, calculated signature score. (B) The survival status of all the GSE22541 3 China. Department of Pathology, Second Hospital of ShanXi Medical Univer- patients samples. (C) Survival analysis of the high-risk and low-risk groups sity, No.382 Wuyi Road, Tai Yuan, ShanXi Province, 030000, China. of GSE22541 patients. (D) ROC curve of the gene signature to predict GSE22541 patients survival. (E) Relative expression of MMP9, IRF7, NFKB1 Received: 15 December 2022 Accepted: 4 July 2023 and HMOX1 genes in GSE22541 samples displayed in a heatmap. Additional file 3: Supplementary Table 1. Detailed information of the GEO datasets used for identifying differently expressed genes in ccRCC vs normal renal tissues. Supplementary Table 2. Number of the four levels differently expressed genes in ccRCC analyzed based on four GEO profiles. References Supplementary Table 3. Detailed information of the 269 differently 1. Wu J, Xu WH, Wei Y, Qu YY, Zhang HL, Ye DW. An integrated score and expressed meanwhile immune modulation related genes in ccRCC. nomogram combining clinical and immunohistochemistry factors to pre- dict high isup grade clear cell renal cell carcinoma. Front Oncol. 2018;8:634. 2. Browning L, Colling R, Verrill C. WHO/ISUP grading of clear cell renal cell Acknowledgements carcinoma and papillary renal cell carcinoma; validation of grading on We sincerely appreciate the researchers for providing their GEO databases the digital pathology platform and perspectives on reproducibility of information online which were important data resources for the study, it is our grade. Diagn Pathol. 2021;16(1):75. pleasure to acknowledge their contributions. 3. Stone L. Kidney cancer: activation of oncogenes driven by VHL loss in ccRCC. Nat Rev Urol. 2017;14(11):637. Authors’ contributions 4. Expert Group of G, Urological Pathology CSoP. [Recent advances and ZG, JD and NW designed the study and together drafting the manuscript, expert consensus on molecular pathology of renal cell carcinomas (2020 contributed equally to the whole study. NS, ZY and XW performed the data version)]. Zhonghua Bing Li Xue Za Zhi. 2020;49(12):1232–41. collecting and analysis. HY, NZ and ZZ participated in the data interpretation 5. Williamson SR, Gill AJ, Argani P, Chen YB, Egevad L, Kristiansen G, and study design, RW were involved in the drafting and critical revision of the Grignon DJ, Hes O. Report from the International Society of Urological Gui et al. BMC Cancer (2023) 23:649 Page 23 of 24 Pathology (ISUP) consultation conference on molecular pathology 23. Bedke J, Stuhler V. KEYNOTE-426: axitinib plus pembrolizumab vs suni- of urogenital cancers: III: molecular pathology of kidney cancer. Am J tinib in the first-line treatment of metastatic renal cell carcinoma. Urologe Surg Pathol. 2020;44(7):e47–65. A. 2020;59(7):841–2. 6. Ibragimova I, Maradeo ME, Dulaimi E, Cairns P. Aberrant promoter 24. Samstein RM, Lee CH, Shoushtari AN, Hellmann MD, Shen R, Janjigian hypermethylation of PBRM1, BAP1, SETD2, KDM6A and other chro- YY, Barron DA, Zehir A, Jordan EJ, Omuro A, et al. Tumor mutational load matin-modifying genes is absent or rare in clear cell RCC. Epigenetics. predicts survival after immunotherapy across multiple cancer types. Nat 2013;8(5):486–93. Genet. 2019;51(2):202–6. 7. Wi YC, Moon A, Jung MJ, Kim Y, Bang SS, Jang K, Paik SS, Shin SJ. Loss 25. Raimondi A, Sepe P, Zattarin E, Mennitto A, Stellato M, Claps M, Gua- of nuclear BAP1 expression is associated with high WHO/ISUP grade in dalupi V, Verzoni E, de Braud F, Procopio G. Predictive biomarkers of clear cell renal cell carcinoma. J Pathol Transl Med. 2018;52(6):378–85. response to immunotherapy in metastatic renal cell cancer. Front Oncol. 8. Wang Y, Guo X, Bray MJ, Ding Z, Zhao Z. An integrative genomics 2020;10:1644. approach for identifying novel functional consequences of PBRM1 26. GSE53000: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 truncated mutations in clear cell renal cell carcinoma (ccRCC). BMC 000. Accessed 1 Jan 2022. Genomics. 2016;17(Suppl 7):515. 27. GSE53757: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE53 9. Srivastava A, Doppalapudi SK, Patel HV, Srinivasan R, Singer EA. The 757. Accessed 2 Jan 2022. roaring 2020s: a new decade of systemic therapy for renal cell carci- 28. GSE68417: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE68 noma. Curr Opin Oncol. 2022;34(3):234–42. 417. Accessed 2 Jan 2022. 10. Pandey A. KEYNOTE-564: Adjuvant immunotherapy for renal cell carci- 29. GSE71963: https:// www. ncbi. nlm. nih. gov/ geo/ query/ acc. cgi? acc= GSE71 noma. Indian J Urol. 2022;38(1):75–6. 963. Accessed 3 Jan 2022. 11. Doppalapudi SK, Leopold ZR, Thaper A, Kaldany A, Chua K, Patel HV, 30. IMMPORT database: https:// www. immpo rt. org/ home. Accessed 7 Feb Srivastava A, Singer EA. Clearing up clear cell: clarifying the immuno- 2022. oncology treatment landscape for metastatic clear cell RCC. Cancers 31. VENNY: https:// bioin fogp. cnb. csic. es/ tools/ venny/. . Accessed 9 Jan 2022. (Basel). 2021;13(16):4140. 32. STRING: https:// string- db. org/. Accessed 15 Feb 2022. 12. Rini BI, Pal SK, Escudier BJ, Atkins MB, Hutson TE, Porta C, Verzoni E, 33. Cytoscape:http:// www. softp edia. com/ get/ Scien ce- CAD/ Cytos cape. Needle MN, McDermott DF. Tivozanib versus sorafenib in patients shtml. Accessed 22 Feb 2022. with advanced renal cell carcinoma ( TIVO-3): a phase 3, multi- 34. FunRich3.1.3: http:// www. funri ch. org/. Accessed 27 Feb 2022. centre, randomised, controlled, open-label study. Lancet Oncol. 35. UALCAL: http:// ualcan. path. uab. edu/ analy sis. html. Accessed 7 Mar 2022. 2020;21(1):95–104. 36. GEPIA: http:// gepia. cancer- pku. cn/. Accessed 2 Mar 2022. 13. Schmidt AL, Tabakin AL, Singer EA, Choueiri TK, McKay RR. Next steps: 37. ProtParam: https:// web. expasy. org/ protp aram/. Accessed 13 Mar 2022. sequencing therapies in metastatic kidney cancer in the contemporary 38. ProtScale: https:// web. expasy. org/ prots cale/. Accessed 13 Mar 2022. era. Am Soc Clin Oncol Educ Book. 2021;41:1–11. 39. HumanProteinAtlas: https:// www. prote inatl as. org/. Accessed 14 Mar 14. Zhu J, Armstrong AJ, Friedlander T W, Kim W, Pal SK, George DJ, Zhang 2022. T. Biomarkers of immunotherapy in urothelial and renal cell carcinoma: 40. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. PD-L1, tumor mutational burden, and beyond. J Immunother Cancer. Nucleic Acids Res. 2000;28(1):27–30. 2018;6(1):4. 41. Kanehisa M, Furumichi M, Sato Y, Kawashima M, Ishiguro-Watanabe M. 15 Dumitru A, Dobrica EC, Croitoru A, Cretoiu SM, Gaspar BS. Focus on KEGG for taxonomy-based analysis of pathways and genomes. Nucleic PD-1/PD-L1 as a therapeutic target in ovarian cancer. Int J Mol Sci. Acids Res. 2023;51(D1):D587–92. 2022;23(20):12067. 42. GEPIA: https:// www. gsea- msigdb. org/ gsea/ index. jsp. Accessed 07 Jun 16. Nunez Abad M, Calabuig-Farinas S, de Lobo Mena M, Torres-Martinez 2022. S, Garcia Gonzalez C, Garcia Garcia JA, Iranzo Gonzalez-Cruz V, Camps 43. Wang X, Wu S, Liu F, Ke D, Wang X, Pan D, Xu W, Zhou L, He W. An immu- Herrero C. Programmed death-ligand 1 (PD-L1) as immunotherapy nogenic cell death-related classification predicts prognosis and response biomarker in breast cancer. Cancers (Basel). 2022;14(2):307. to immunotherapy in head and neck squamous cell carcinoma. Front 17. Deng H, Zhao Y, Cai X, Chen H, Cheng B, Zhong R, Li F, Xiong S, Li J, Liu Immunol. 2021;12:781466. J, et al. PD-L1 expression and Tumor mutation burden as Pathological 44. Cibersort: https:// ciber sortx. stanf ord. edu/. Accessed 16 May 2022. response biomarkers of Neoadjuvant immunotherapy for early-stage 45. Labriola MK, Zhu J, Gupta RT, McCall S, Jackson J, Kong EF, White JR, non-small cell lung cancer: a systematic review and meta-analysis. Crit Cerqueira G, Gerding K, Simmons JK, et al. Characterization of tumor Rev Oncol Hematol. 2022;170:103582. mutation burden, PD-L1 and DNA repair genes to assess relationship to 18. Motzer RJ, Escudier B, George S, Hammers HJ, Srinivas S, Tykodi SS, Sos- immune checkpoint inhibitors response in metastatic renal cell carci- man JA, Plimack ER, Procopio G, McDermott DF, et al. Nivolumab versus noma. J Immunother Cancer. 2020;8(1):e000319. everolimus in patients with advanced renal cell carcinoma: updated 46. Kotecha RR, Motzer RJ, Voss MH. Towards individualized therapy for meta- results with long-term follow-up of the randomized, open-label, phase static renal cell carcinoma. Nat Rev Clin Oncol. 2019;16(10):621–33. 3 checkmate 025 trial. Cancer. 2020;126(18):4156–67. 47. Long J, Wang A, Bai Y, Lin J, Yang X, Wang D, Yang X, Jiang Y, Zhao H. 19. Tomita Y, Fukasawa S, Shinohara N, Kitamura H, Oya M, Eto M, Tanabe Development and validation of a TP53-associated immune prognostic K, Saito M, Kimura G, Yonese J, et al. Nivolumab versus everolimus in model for hepatocellular carcinoma. EBioMedicine. 2019;42:363–74. advanced renal cell carcinoma: Japanese subgroup 3-year follow-up 48. Mei W, Jin L, Zhang B, Sun X, Yang G, Li S, Ye L. Computer classification analysis from the phase III checkmate 025 study. Jpn J Clin Oncol. and construction of a novel prognostic signature based on moonlighting 2019;49(6):506–14. genes in prostate cancer. Front Oncol. 2022;12:982267. 20. Motzer RJ, Choueiri TK, McDermott DF, Powles T, Vano YA, Gupta S, Yao 49 Yin Y, Li S, Liang X, Li K, Xie M, Hu B. Construction and validation of an J, Han C, Ammar R, Papillon-Cavanagh S, et al. Biomarker analysis from oxaliplatin-resistant gene signature in colorectal cancer patients who CheckMate 214: nivolumab plus ipilimumab versus sunitinib in renal underwent chemotherapy. Pharmaceuticals (Basel). 2022;15(9):1139. cell carcinoma. J Immunother Cancer. 2022;10(3):e004316. https:// doi. 50. Gao J, Huo S, Zhang Y, Zhao Z, Pan H, Liu X. Construction of ovarian org/ 10. 1136/ jitc- 2021- 004316. metastasis-related immune signature predicting prognosis of gastric 21. Jena R. CheckMate 214 trial: Immune checkpoint regulators for cancer patients. Cancer Med. 2022;12(1):913–29. advanced renal cell carcinoma. Indian J Urol. 2020;36(3):231–3. 51. Zhang X, Wang Z, Zeng Z, Shen N, Wang B, Zhang Y, Shen H, Lu W, Wei 22. Powles T, Plimack ER, Soulieres D, Waddell T, Stus V, Gafanov R, Nosov R, Ma W, et al. Bioinformatic analysis identifying FGF1 gene as a new D, Pouliot F, Melichar B, Vynnychenko I, et al. Pembrolizumab plus prognostic indicator in clear cell renal cell carcinoma. Cancer Cell Int. axitinib versus sunitinib monotherapy as first-line treatment of 2021;21(1):222. advanced renal cell carcinoma (KEYNOTE-426): extended follow- 52. Wei R, Qi G, Zeng Z, Shen N, Wang Z, Shen H, Gao L, Song C, Ma W, Wang up from a randomised, open-label, phase 3 trial. Lancet Oncol. C. IMUP and GPRC5A: two newly identified risk score indicators in pancre - 2020;21(12):1563–73. atic ductal adenocarcinoma. Cancer Cell Int. 2021;21(1):620. Gui et al. BMC Cancer (2023) 23:649 Page 24 of 24 53. Ma W, Wang B, Zhang Y, Wang Z, Niu D, Chen S, Zhang Z, Shen N, Han W, Zhang X, et al. Prognostic significance of TOP2A in non-small cell lung cancer revealed by bioinformatic analysis. Cancer Cell Int. 2019;19:239. 54. Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer immunoed- iting: from immunosurveillance to tumor escape. Nat Immunol. 2002;3(11):991–8. 55. Schreiber RD, Old LJ, Smyth MJ. Cancer immunoediting: integrat- ing immunity’s roles in cancer suppression and promotion. Science. 2011;331(6024):1565–70. 56. Li X, Shao C, Shi Y, Han W. Lessons learned from the blockade of immune checkpoints in cancer immunotherapy. J Hematol Oncol. 2018;11(1):31. 57 Li Y, Li F, Jiang F, Lv X, Zhang R, Lu A, Zhang G. A mini-review for cancer immunotherapy: molecular understanding of PD-1/PD-L1 path- way & translational blockade of immune checkpoints. Int J Mol Sci. 2016;17(7):1151. 58. Pardoll DM. The blockade of immune checkpoints in cancer immuno- therapy. Nat Rev Cancer. 2012;12(4):252–64. 59. Zou W, Chen L. Inhibitory B7-family molecules in the tumour microenvi- ronment. Nat Rev Immunol. 2008;8(6):467–77. 60. Khan M, Arooj S, Wang H. Soluble B7-CD28 family inhibitory immune checkpoint proteins and anti-cancer immunotherapy. Front Immunol. 2021;12:651634. 61. Schildberg FA, Klein SR, Freeman GJ, Sharpe AH. Coinhibitory pathways in the B7-CD28 ligand-receptor family. Immunity. 2016;44(5):955–72. 62. Kalkusova K, Smite S, Darras E, Taborska P, Stakheev D, Vannucci L, Bartunkova J, Smrz D. Mast cells and dendritic cells as cellular immune checkpoints in immunotherapy of solid tumors. Int J Mol Sci. 2022;23(19):11080. 63. Honda T, Keith YH. Novel insights into the immune-regulatory func- tions of mast cells in the cutaneous immune response. Front Immunol. 2022;13:898419. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : fast, convenient online submission thorough peer review by experienced researchers in your ﬁeld rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions

Journal

BMC Cancer – Springer Journals

Published: Jul 12, 2023

Keywords: Clear cell renal cell carcinoma (ccRCC); Immune response; LASSO analysis; Gene signature; Prediction biomarker

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Immune regulation and prognosis indicating ability of a newly constructed multi-genes containing signature in clear cell renal cell carcinoma

Immune regulation and prognosis indicating ability of a newly constructed multi-genes containing signature in clear cell renal cell carcinoma

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Immune regulation and prognosis indicating ability of a newly constructed multi-genes containing signature in clear cell renal cell carcinoma

Immune regulation and prognosis indicating ability of a newly constructed multi-genes containing signature in clear cell renal cell carcinoma

References (78)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies