Copper-binding protein modelling by single-cell transcriptome and Bulk transcriptome to predict overall survival in lung adenocarcinoma patients

Abstract

Background: Copper and copper-binding proteins are key components of tumour progression as they play an important role in tumour invasion and migration, and abnormal accumulation of copper (Cu) may be intimately linked to with lung adenocarcinoma (LUAD). Methods: Data on lung adenocarcinoma were sourced from the Cancer Genome Atlas (TCGA) database and the National Centre for Biotechnology Information (GEO). 10x scRNA sequencing, which is from Bischoff P et al, was used for down-sequencing clustering and subgroup identification using TSNE. The genes for Copper-binding proteins (CBP) were acquired from the MSigDB database. LASSO-Cox analysis was subsequently used to construct a model for copper-binding proteins (CBPRS), which was then compared to lung adenocarcinoma models developed by others. External validation was carried out in the GSE31210 and GSE50081 cohorts. The effectiveness of immunotherapy was evaluated using the TIDE algorithm and the IMvigor210, GSE78220, and TCIA cohorts. Furthermore, differences in mutational profiles and the immune microenvironment between different risk groups were investigated. The CBPRS's key regulatory genes were screened using ROC diagnostic and KM survival curves. The differential expression of these genes was then verified by RT-qPCR. Results: The six CBP genes were identified as highly predictive of LUAD prognosis and significantly correlated with it. Multivariate analysis showed that patients in the low-risk group had a higher overall survival rate than those in the high-risk group, indicating that the model was an independent predictor of LUAD. The CBPRS demonstrated superior predictive ability compared to 11 previously published models. We constructed a column-line graph that includes CBPRS and clinical characteristics, which exhibits high predictive performance. Additionally, we observed significant differences in biological functions, mutational landscapes, and immune cell infiltration in the tumour microenvironment between the high-risk and low-risk groups. It is noteworthy that immunotherapy was also significant in both the high- and low-risk groups. These results suggest that the model has good predictive efficacy. Conclusions: The CBP model demonstrated good predictive performance, revealing characteristics of the tumour microenvironment. This provides a new method for assessing the efficacy of pre-immunisation and offers a potential strategy for future treatment of lung adenocarcinoma.