Scholarlink

Background <p>Non-small cell lung cancer (NSCLC) is one of the most prevalent malignancies and currently shows a poor clinical prognosis. Type II pneumocyte, as one of the main sources of cancer cells in NSCLC, is important to explore the molecular functions of its related markers for a deeper understanding of NSCLC.</p> Methods <p>In this study, NSCLC single-cell profiles and transcriptomic data were derived from Gene Expression Omnibus (GEO). Single-cell clustering was used to characterize the infiltration of type II pneumocyte and to reveal differences in the regulatory function of this cell compared to other cell subpopulations. Interactions between cellular subpopulations were clarified by CellChat. High-dimensional weighted gene co-expression network analysis (hdWGCNA) was utilized to mine modular signature genes closely relevant to type II pneumocyte in NSCLC. Differential expression analysis combined with least absolute shrinkage and selection operator (LASSO) and support vector machine recursive feature elimination (SVM–RFE), two machine learning algorithms, was used to screen key genes in NSCLC that are associated with disease progression and type II pneumocyte. The CIBERSORT and ESTIMATE algorithms were mainly used to assess the correlation between biomarkers and immune cell infiltration in NSCLC. Gene set enrichment analysis (GSEA) was used to evaluate the regulatory functions of biomarkers. CCK-8, Transwell assay, and wound healing assay were used to illustrate the regulation of biomarkers to NSCLC cell lines.</p> Results <p>Seven types of cell subpopulations were found in NSCLC samples, including macrophages, type II pneumocyte, T cell, ciliated cell, B cell, endothelial cell, and plasma cell. Among them, there is extensive cellular communication between type II pneumocyte and macrophages, mainly through ITGB2–ICAM1, HLA–DPB1–CD4, and HLA–DQB1–CD4 ligand–receptor pair binding. hdWGCNA combined with machine learning algorithms screened three key genes, <i>HN1</i>, <i>OCIAD2</i>, and <i>SFTA2,</i> associated with type II pneumocyte and NSCLC progression. Among them, <i>HN1</i> and <i>OCIAD2</i> were significantly negatively correlated with the stromal infiltration score, immunization score, and ESTIMATE score, respectively, whereas <i>SFTA2</i> was significantly positively correlated with these three indicators. In vitro cellular assays indicate that <i>HN1</i> and <i>OCIAD2</i> are significantly upregulated in the NCI-H838 and A549&#xa0;cell lines, whereas <i>SFTA2</i> expression is downregulated. Furthermore, silencing <i>HN1</i> significantly inhibits the proliferation, migration, and invasive capacity of NSCLC cell lines.</p> Conclusion <p>This study elucidates genes in NSCLC that are highly associated with both disease progression and type II pneumocyte features as potential biomarkers of NSCLC, and predicts their potential regulatory functions as well as their impact on the immune infiltration profile of NSCLC, providing guidance for the unraveling of the immune mechanisms of NSCLC.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Single cell and machine learning identify type II pneumocyte-derived biomarkers HN1/OCIAD2/SFTA2 for non-small cell lung cancer prognosis and immune regulation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月26日

查看全文
  • Zhaofan Lin,
  • Hanguang Yu,
  • Kai Yang,
  • Huijuan Xu

摘要

Background

Non-small cell lung cancer (NSCLC) is one of the most prevalent malignancies and currently shows a poor clinical prognosis. Type II pneumocyte, as one of the main sources of cancer cells in NSCLC, is important to explore the molecular functions of its related markers for a deeper understanding of NSCLC.

Methods

In this study, NSCLC single-cell profiles and transcriptomic data were derived from Gene Expression Omnibus (GEO). Single-cell clustering was used to characterize the infiltration of type II pneumocyte and to reveal differences in the regulatory function of this cell compared to other cell subpopulations. Interactions between cellular subpopulations were clarified by CellChat. High-dimensional weighted gene co-expression network analysis (hdWGCNA) was utilized to mine modular signature genes closely relevant to type II pneumocyte in NSCLC. Differential expression analysis combined with least absolute shrinkage and selection operator (LASSO) and support vector machine recursive feature elimination (SVM–RFE), two machine learning algorithms, was used to screen key genes in NSCLC that are associated with disease progression and type II pneumocyte. The CIBERSORT and ESTIMATE algorithms were mainly used to assess the correlation between biomarkers and immune cell infiltration in NSCLC. Gene set enrichment analysis (GSEA) was used to evaluate the regulatory functions of biomarkers. CCK-8, Transwell assay, and wound healing assay were used to illustrate the regulation of biomarkers to NSCLC cell lines.

Results

Seven types of cell subpopulations were found in NSCLC samples, including macrophages, type II pneumocyte, T cell, ciliated cell, B cell, endothelial cell, and plasma cell. Among them, there is extensive cellular communication between type II pneumocyte and macrophages, mainly through ITGB2–ICAM1, HLA–DPB1–CD4, and HLA–DQB1–CD4 ligand–receptor pair binding. hdWGCNA combined with machine learning algorithms screened three key genes, HN1, OCIAD2, and SFTA2, associated with type II pneumocyte and NSCLC progression. Among them, HN1 and OCIAD2 were significantly negatively correlated with the stromal infiltration score, immunization score, and ESTIMATE score, respectively, whereas SFTA2 was significantly positively correlated with these three indicators. In vitro cellular assays indicate that HN1 and OCIAD2 are significantly upregulated in the NCI-H838 and A549 cell lines, whereas SFTA2 expression is downregulated. Furthermore, silencing HN1 significantly inhibits the proliferation, migration, and invasive capacity of NSCLC cell lines.

Conclusion

This study elucidates genes in NSCLC that are highly associated with both disease progression and type II pneumocyte features as potential biomarkers of NSCLC, and predicts their potential regulatory functions as well as their impact on the immune infiltration profile of NSCLC, providing guidance for the unraveling of the immune mechanisms of NSCLC.