| Purpose:Renal cell carcinoma(RCC)is one of the most common malignant tumors in the urinary system,and its pathological subtypes include clear-cell renal carcinoma(cc RCC)and non-clear-cell renal carcinoma(ncc RCC).Tumor metastasis and treatment resistance are important malignant progression of RCC.The treatment strategy for patients with advanced cc RCC is mainly based on angiogenesis inhibitors combined with immune checkpoint inhibitors,but there are still some patients who do not respond to therapies or develop secondary resistance.Therefore,it is urgent to further explore the key prognostic factors and potential therapeutic targets of RCC.One of the characteristics of tumors is heterogeneity,which can lead to different responses for patients to guidance-recommended therapies.Studies have shown that the characterization of tumor heterogeneity should not be limited to tumor cells,but should deepen the understanding of tumor microenvironment heterogeneity.The heterogeneity of tumor and its microenvironment affects the biological behavior of RCC and is thought to be associated with the survival prognosis of RCC,but the mechanism is not fully understood.Different from traditional sequencing,single-cell transcriptome sequencing has achieved a breakthrough in transcriptome analysis at the single cell level,making it possible to fully reveal the heterogeneity of tumor cells and microenvironment.We expected to further identify the molecular targets of tumor cells and microenvironment affecting the prognosis of RCC through single-cell transcriptome data analysis of patients’ tissues with metastatic RCC and late-stage RCC,as well as in vivo and in vitro experiments,and conduct cohort verification in multi-center clinicopathological specimens.We aimed to investigate the role and possible mechanism of tumor cell and microenvironment heterogeneity in RCC metastasis and resistance to targeted therapy.Methods:First,fresh tissue samples from patients with clinical metastatic RCC and advanced RCC were collected for single-cell transcriptome sequencing,including tumor primary sites,metastases and para-cancer tissues from patients;and single-cell sequencing data from tissues of patients with advanced RCC from public databases were collected for comprehensive integrated bioinformatics analysis.The heterogeneity of RCC tumor cells and microenvironment was analyzed.After that,the sequencing data from TCGA,ICGC and multi-center RCC public clinical drug trial cohort of IMmotion151 and Checkmate-025 were combined to select key molecules that pointed to RCC heterogeneity and metastasis,resistance and survival,and their prognostic value was verified in clinical patient pathological sample from our center.Finally,in vitro and in vivo intervention experiments were used to investigate the mechanism of key molecules involved in the regulation of RCC tumor growth,metastasis and drug resistance.Results:Part I Identification and validation of Serine protease inhibitor E2(SERPINE2),a key molecular marker of renal cell carcinoma metastasis and progression.1.Construction of single-cell transcriptome landscape of primary and metastatic RCC.We analyzed single-cell data from 3 RCC primary sites and 5 RCC lymph node metastases from 6 patients.30,514 malignant cells and 14,762 non-malignant cells were distinguished by copy number variation and malignant cell epithelial score.T cells,endothelial cells,macrophages,B/plasma cells,NK cells and fibroblasts were identified by specific markers.2.Identification of the heterogeneity of expression patterns in malignant cells and SERPINE2,a molecule that is important for the biological behavior of metastasis.Unsupervised non-negative matrix factorization(NNMF)reveals a consistent set of genes that are preferentially co-expressed by malignant cell populations and identifies the tumor metastasis program of RCC.All genes in the cluster were screened by survival analysis,tissue expression differences,and single-cell expression differences,and SERPINE2 was found to be the most relevant molecule for metastasis and prognosis.3.SERPINE2,as a metastasis-associated oncogene,plays a predictive role in RCC and its therapy response.Gene set enrichment analysis(GSEA)showed that patients with high expression of SERPINE2 in the TCGA database were significantly enriched in EMT-related pathways,and the published data of TCGA,GSE105261 and GSE53757 showed high expression of SERPINE2 in patients with tumors or metastases compared with para-cancer tissues.Meanwhile,cancer therapy sensitivity genomics,single-cell expression level,and immunotherapy cohort suggest that high SERPINE2 expression corresponds to targeted therapy resistance and is associated with immune infiltration and immunosuppressive therapy.4.SERPINE2 promotes RCC invasion in vivo and in vitro,and is associated with epithelial-mesenchymal transition(EMT).SERPINE2 was highly expressed in Caki-1 cells and lowly expressed in 786-O cells.After Caki-1 knockdown of SERPINE2,migration,invasion and colony formation ability were weakened,while after 786-O overexpression of SERPINE2,migration,invasion and colony formation ability were enhanced.WB and q PCR results indicated that EMT activation signal was enhanced after overexpression and weakened after knockdown.Stable luciferase labeled SERPINE2-overexpressed 786-O cell lines showed more severe lung metastasis in a nude mouse tail-vein metastasis model.5.The clinicopathological samples confirmed that SERPINE2 predicted RCC metastasis,and its higher expression corresponded to poorer overall survival and progression-free survival.SERPINE2 expression was an independent risk factor for overall survival.Part II1.Single cell transcriptome analysis of cc RCC patients before and after neoadjuvant TKI treatment.Biopsy tumor samples from 4 patients before targeted therapy and surgically paired tumor and paracancer tissue samples after treatment were collected(Sensitive: Resistant =2:2),a total of 81135 cells were retained after quality control by 10 X Genomic single cell sequencing,among which the proportion of tumor cells,endothelial cells and T cells changed significantly before and after treatment,while only endothelial cells showed differences between the response group and the non-response group.2.Tumor-infiltrating T cells exhibit transcriptional heterogeneity and resistance-related exhaustion and senescence.After unsupervised clustering,18382 T cells were divided into 8 cell subsets,including cytotoxic CD8+T cells,exhausted CD8+T cells and central memory CD4+T cells,according to the expression of key markers.The functional exhaustion scores of treatment-sensitive patients were significantly reduced after treatment,and exhausted CD8+T cells showed lineage heterogeneity changes in 4subsets.There was no significant difference in the proportion of T subsets between sensitive and drug-resistant patients before and after treatment.After targeted treatment,only exhausted CD8+T cells in sensitive patients showed a change in treatment senescence level,while more cell subsets in resistant patients showed a decrease in senescence score.3.The dynamic changes of endothelial cell transcription level and risk score in neoadjuvant therapy are correlated with the prognosis of RCC.The 9672 endothelial cells were classified into 4 subgroups(Endo＿1,Endo＿2,Endo＿C and Endo＿p)by unsupervised clustering.Analysis of gene set variation(GSVA)showed that the four subgroups had different functions.The top 10 markers of Endo＿p subgroup were used to establish a risk score model.The Cox regression analysis based on the Lasso model retained 6 genes for modeling,which showed that high-risk score was associated with poor prognosis,and drug sensitivity analysis also showed that patients with high-risk score had a higher IC50 for Pazopanib.When using the IMmotion151 cohort receiving targeted therapy as an external validation,the risk score model can well distinguish progression-free survival.Finally,multivariate Cox analysis showed that Endo＿p score was an independent risk factor for the prognosis of renal cell carcinoma.Part III1.Tumor cells showed different biological functional characteristics after targeted therapy.In the single-cell analysis of neoadjuvant therapy,37,118 epithelial cells could be divided into proximal tubules,distal tubules,collecting tubules,and three tumor cell groups: S＿RCC1,S＿RCC2,and I＿RCC.All I＿RCC cells were from drug-resistant patients.Copy number variation results showed that chromosome 3p deletion was more obvious in I＿RCC cells and tumor cells of resistant patients.Cell cycle analysis showed that I＿RCC was more proliferative,and S＿RCC2 and I＿RCC were enriched in more tumor-related pathways such as oxidative phosphorylation,hypoxia,and EMT.2.To identify ferroptosis as a core transcriptional program associated with resistance in RCC.The NNMF algorithm was used to find six gene sets that were co-expressed in all tumor samples,and the GSEA analysis was used to annotate the cell cycle,hypoxia,inflammation,stemness and metastasis,and ferroptosis program.Differential gene analysis found that VEGF and ferroptosis were positively enriched in tumor cells of sensitive patients,and the scores of ferroptosis gene sets of resistant tumor cells were significantly reduced.3.Bioinformatics analysis revealed that IL-6 promoted the expression of SLC7A11 through JAK2-STAT3.The cross-validation of ferroptosis biomarkers with the core genes of the NNMF transcription program only IL-6 showed higher transcriptional levels in resistant tumor cells I＿RCC.Public data analysis showed that high IL-6 expression was associated with poorer survival,and IL-6 levels were positively correlated with SLC7A11 expression.Bulk-seq and single-cell transcriptome data from IMmotion151 cohorts receiving targeted therapy showed a significant high correlation between IL-6 and SLC7A11.Further GSEA analysis showed that IL-6-JAK2-STAT3 signaling pathway is an important pathway associated with drug resistance.4.Clinicopathological specimens showed that IL-6 was correlated with SLC7A11 expression and lipid peroxidation.Immunohistochemical(IHC)analysis of tissue samples showed that IL-6 and SLC7A11 levels were higher in advanced cc RCC than in normal epithelial cells,while lipid peroxidation marker 4-hydroxynonenal(4-HNE)levels were lower in normal kidney tissue.In the progression of cc RCC,the up-regulation of IL-6expression was positively correlated with the level of SLC7A11,and the expression of SLC7A11 was negatively correlated with 4-HNE.5.Establishment of a targeted therapy resistant 786-O cell line and verification in vivo and in vitro that IL-6 inhibits ferroptosis of tumor cells by upregulating SLC7A11.Sunitinib inhibited the proliferation rate of normal renal epithelial cell lines HK2 and cc RCC cell lines A498,ACHN,Caki1,769-P and 786-O,but did not inhibit the drug-resistant 786O-R cell line.The relative m RNA expression and protein level of SLC7A11 in 786O-R were higher than those of wild type.Bulk-seq analysis showed that the IL-6-JAK2-STAT3 signaling pathway was significantly enriched in 786O-R resistant cell lines.After SLC7A11 knockdown with 786O-R,the ability of colony formation and proliferation were decreased,while the ferroptosis and malondialdehyde level were increased.Subcutaneous tumor formation in nude mice showed that SLC7A11 knockdown effectively inhibited tumor growth,and IHC analysis of xenograft showed that 4-HNE was up-regulated after SLC7A11 knockdown.6.The process of IL-6 regulation of SLC7A11 is mediated by JAK2-STAT3 signaling,and ferroptosis agonists can reverse the resistance of RCC targeted therapy.IL-6 promotes the expression of SLC7A11 in a dose-dependent manner,while up-regulating p-JAK2 and p-STAT3 levels,and blocking the IL-6-induced up-regulation of SLC7A11 in the JAK2/STAT3 signaling pathway is inhibited.IL-6 significantly reversed the SLC7A11 knockdown or iron-death agonist Erastin induced decline in cell colony formation.IL-6significantly reversed the increase of intracellular ferrous ions and MDA levels induced by SLC7A11 knockdown.In vivo,with subcutaneous injection of 786-O cells to establish xenograft models,Erastin significantly reversed the increase in tumor weight and volume caused by IL-6,and IHC analysis of xenografts showed that IL-6 reduced 4-HNE levels,while Erastin increased 4-HNE levels.Conclusion:In this study,single-cell transcriptional sequencing data from primary and metastatic sites of RCC,as well as from cancer and paired para-cancer tissues before and after targeted therapy were integrated,and transcriptome and genome sequencing were combined to analyze the heterogeneity of RCC tumor microenvironment through multi-omics data,and to identify key modules related to metastasis and drug resistance.SERPINE2 was identified as an important target for metastasis of advanced RCC and promoted the malignant progression of RCC in vivo and in vitro.Meanwhile,IL-6,as an upstream factor,stimulated the upregulation of SLC7A11 through the JAK2-STAT3 signaling pathway,resulting in ferroptosis inhibition and resistance to targeted therapy.Erastin may play a role in reversing targeted drug resistance.Finally,translational medicine verification was conducted by using multiple cohorts from TCGA,ICGC,Checkate-025 and IMmotion151 clinical trials,as well as patients with different stages of RCC or targeted therapy in our center,providing new ideas and therapeutic targets for advanced RCC progression and drug resistance dilemma. |
PDF Full Text Request