Recognition Of Novel Prognostic Biomarkers For Tumor Immune Microenvironment Refactoring By Tertiary Lymphoid Structures In Skin Cutaneous Melanoma

Background Cutaneous melanoma is one of the most lethal skin cancers,and its incidence continues to rise worldwide.The tertiary lymphoid structure represents a lymphoid regeneration site that may promote adaptive immune response.Recent studies have shown that the tertiary lymphoid structure plays a role in the immune checkpoint inhibitor treatment response,and the tertiary lymphoid structure in tumors is associated with better prognosis after immunotherapy.However,the driving factors of TLS formation and its effect on immune response are not fully understood,and how to define and describe the markers of TLSs is not clear,which makes the prediction and evaluation of the content of tertiary lymphoid structure in the immune microenvironment of cutaneous melanoma complicated.Objective The genes related to tertiary lymphoid structure with prognostic value were identified,and the molecular subtypes related to tertiary lymphoid structure were identified.The number of tertiary lymphoid structures in cutaneous melanoma was evaluated based on transcriptome construction algorithm.Locate the key genes for maintaining the homeostasis of the tertiary lymphoid structure,and clarify the pan-cancer characteristics of the tertiary lymphoid structure.Methods(1)Univariate cox analysis was used to screen the prognostic tertiary lymphoid structure genes related to the specific survival of cutaneous melanoma.Based on the non-negative matrix factorization algorithm,unsupervised consensus clustering of cutaneous melanoma samples in TCGA and GEO databases was performed to identify the molecular subtypes of tertiary lymphoid structure.(2)Lasso-cox was used to further remove the collinearity of prognostic tertiary lymphoid structure genes,calculate the risk characteristics of tertiary lymphoid structure,and verify the repeatability of the characteristics.Then,the average expression of the genes in the features was defined as the tertiary lymphoid structure possibility score(TPS).Combined with WGCNA,machine learning and other algorithms,the robustness of TPS in predicting TLS content was verified from biological function,immune cell content and immune cell composition.(3)The predictive value of TPS for immunotherapy was evaluated by using a variety of classical markers of immunotherapy,and the AUC value of TPS for immunotherapy response prediction was verified by 8 data sets.In addition,the difference of chemotherapy sensitivity between high and low TPS groups was evaluated based on IC50.We also clarified the relationship between TPS and skin melanoma mutations.(4)Random forest was used to further screen and identify the most critical genes of TPS gene,and immunohistochemistry was used to verify the difference in the expression of melanoma compared with melanoma hemorrhoids.Six melanoma single cell sequencing datasets were used to locate the expression sources of key genes,and the results were verified by immunofluorescence co-localization.Finally,we also clarified the heterogeneity and unity of TPS in pan-cancer.Results(1)We identified 34 prognostic tertiary lymphoid structural genes associated with disease-specific survival of cutaneous melanoma,and identified two molecular subtypes C1 and C2 with significant biological uniqueness.C2 has statistically better survival,which may be related to more immune cell infiltration.(2)Lasso-cox included 12 genes to construct risk characteristics,which had good predictive accuracy for the survival of patients with cutaneous melanoma and could be repeated in external data sets.The average expression value of these 12 genes was defined as TPS to standardize the measurement method of TLS number.The biological function,immune cell composition and immune cell content of high TPS group were highly consistent with TLS,which proved that TPS was likely to be a marker for high-precision prediction of TLS.(3)The prediction results of a variety of classical immunotherapy markers and 8immunotherapy data sets consistently showed that TPS was a superior immunotherapy predictive marker,and patients with high TPS were more likely to respond to immunotherapy.Interestingly,melanoma mutations had no statistical correlation with TPS.(4)TRAITS(EAF2)is the most important TPS gene.Immunohistochemistry suggests that it is highly expressed in melanoma.The consistency of multiple single cell data sets indicates that TRAITS is derived from B cell lines.TRAITS was significantly associated with CASP7 at both transcriptional and protein levels.Chip-seq suggested that TRAITS binds to the promoter region of CASP7.Pan-cancer analysis suggests that TPS is different in a large number of tumors and is related to survival.Conclusion TLS risk characteristics are a robust prognostic feature for predicting the prognosis of SKCM,and TPS as a direct indicator can predict the therapeutic response of immune checkpoint inhibitors.In addition,TRAITS is a gene that plays a key role in TLS regulation,which is derived from B cell line and can bind to the CASP7 promoter region.Targeting the TRAITS-CASP7-TLS axis may reconstruct the TME of cutaneous melanoma and induce a broader and stronger immune response to enhance the immunotherapy effect of SKCM.