Abstract
Background:
Lung adenocarcinoma (LUAD), as a major subtype of lung cancer, continues to have high incidence and mortality rates worldwide. Macrophages play a complex role in the LUAD tumor microenvironment (TME), promoting tumor growth and metastasis while also participating in tumor resistance mechanisms. Given this, our study aims to delve into the heterogeneity of macrophages within LUAD and their impact on disease prognosis, further explore and validate the expression and function of macrophage marker genes in LUAD, and construct a prognostic model based on these marker genes. This will provide new biomarkers and strategies for the early diagnosis, treatment selection, and prognosis assessment of LUAD.
Method:
This study meticulously dissected the LUAD TME using single-cell RNA sequencing (scRNA-seq) technology, visualized cell clusters through UMAP technology, and accurately identified various cell subtypes using manual annotation and the SingleR automatic annotation method. We analyzed intercellular communication networks using the CellChat software package to explore the subgroup distribution and function of macrophages in the TME and further revealed the developmental trajectory and differentiation pathways between macrophage subgroups through pseudotime analysis. The ssGSEA algorithm was applied to calculate the macrophage subgroup enrichment scores for each patient in the TCGA database, further evaluating the impact of each macrophage subgroup on LUAD prognosis based on these enrichment scores. Subsequently, based on the marker genes of macrophages with significant effects on LUAD prognosis, we constructed a prognostic model in the TCGA-LUAD cohort and validated it with independent cohorts from the GEO database. Additionally, expression validation was conducted in LUAD patient samples from the Chinese population using RT-PCR technology, and a series of in vitro and in vivo cellular experiments were conducted to explore the functional role of the COL5A1 gene in LUAD. Finally, through a co-culture system, we confirmed that COL5A1 can promote the polarization of anti-inflammatory macrophages.
Results:
Among the numerous pathways emanating from macrophages, we discovered that signals such as SPP1 and MIF were more active in tumor tissues, suggesting potential oncogenic mechanisms in macrophages. Utilizing macrophage marker genes, we developed a LUAD prognostic model using Lasso regression combined with multivariate COX regression. This model robustly predicts the prognosis and immunotherapy efficacy in LUAD patients. Based on the model's risk score and other clinical features, we constructed a nomogram capable of predicting LUAD prognosis. Additionally, we systematically analyzed the differences between high and low-risk groups in terms of TME, enrichment analysis, mutational landscape, and predicted immunotherapy efficacy. RT-PCR validated the expression of genes used in the model construction, partially corroborating our bioinformatics analysis and underscoring the robustness of our approach. The final series of experiments demonstrated that COL5A1 might promote the progression of LUAD by facilitating the polarization of anti-inflammatory macrophages.
Conclusion:
Our study unveiled potential oncogenic mechanisms of macrophages and identified the influence of various macrophage subtypes on the prognosis of LUAD patients. We developed a robust prognostic model based on macrophage marker genes, which demonstrated exceptional performance in predicting prognosis and the efficacy of immunotherapy. Ultimately, a series of cellular experiments established COL5A1 as a potential therapeutic target for LUAD