Anoikis-related gene signature is associated with immune infiltration and predicts the prognosis of non-small cell lung cancer

Yixuan Wu, Zhou Zhou, Qianyi Qi, Shirong Xu, Lin Chen, Feng WangAging 2024 · DOI

Abstract

Non-small cell lung cancer (NSCLC) is the most common histological type of lung cancer. With the in-depth exploration of cell death manners, numerous studies found that anoikis is an important mechanism that associated with treatment. Therefore, we aimed to explore the prognostic value and treatment guidance of anoikis in NSCLC patients. In the current study, we first constructed a prognostic model based on the anoikis-related genes based on bulk RNA-sequencing and single-cell RNA-sequencing (scRNA-seq) dataset. Then, immuno-correlations of anoikis-related risk scores (ARGRS) were analyzed. In addition, HMGA1, a risky gene in ARGRS, was further explored to define its expression and immuno-correlation. Results showed that patients with higher ARGRS had worse clinical outcomes. Moreover, the five genes in the prognostic model were all highly expressed on tumor cells. Moreover, further analysis found that the ARGRS was negatively correlated with ImmuneScore, but positively with tumor purity. Besides, patients in the ARGRS-high group had lower levels of immunological characteristics, such as the immune-related signaling pathways and subpopulations. Additionally, in the immunotherapy cohorts, patients with the ARGRS-high phenotype were more resistant to immunotherapy and tended to not achieve remission after treatment. Last, HMGA1 was chosen as the representative biomarker, and analysis of the in-house cohort showed that HMGA1 was highly expressed in tumor tissues and correlated with decreased T cell infiltration. To sum up, ARGRS was correlated with a desert tumor microenvironment and identified immune-cold tumors, which can be a novel biomarker for the recognition of immunological characteristics and an immunotherapeutic response in NSCLC.

Introduction

Non-small cell lung cancer (NSCLC) remains a leading cause of global cancer-related mortality. While the advent of immunotherapy, particularly immune checkpoint inhibitors, has significantly improved outcomes, these treatments are only effective in a subset of patients. A major clinical challenge is distinguishing between "hot" tumors, which are inflamed and responsive, and "cold" tumors, which act as immune deserts and resist treatment.

This study investigates anoikis resistance as a potential driver of these "cold" tumor phenotypes. Anoikis is a form of programmed cell death that occurs when a cell detaches from the extracellular matrix. Cancer cells that develop resistance to anoikis can survive in the bloodstream, facilitating metastasis. The researchers hypothesized that the genetic signature of anoikis resistance could serve as a reliable biomarker for predicting both patient prognosis and the immune landscape of the tumor microenvironment (TME).

Methods

The researchers employed a multi-omic approach to develop and validate a prognostic model:

  1. Signature Construction: Using bulk RNA-sequencing data from The Cancer Genome Atlas (TCGA), 33 anoikis-related genes (ARGs) were identified as significantly upregulated in NSCLC compared to normal tissue.
  2. Model Refinement: Through univariable COX and LASSO regression analysis, the initial set was narrowed to a five-gene "Anoikis-Related Gene Risk Score" (ARGRS), comprising HMGA1, PLK1, ETV4, PHLDA2, and ITGB4.
  3. Validation: The model was validated using multiple independent datasets (GEO cohorts) and single-cell RNA-sequencing (scRNA-seq) to confirm tumor-specific expression. Protein-level validation was conducted via immunohistochemistry (IHC) on an in-house tissue microarray.
Initial identification and characterization of anoikis-related genes in NSCLC, including PCA, differential expression analysis, and genomic localization.

Key Insights and Results

Prognostic Power and Immunotherapy Resistance

Across all tested cohorts, patients categorized into the high-ARGRS group exhibited significantly shorter overall survival (OS). Furthermore, the ARGRS proved to be a potent predictor of immunotherapy failure. In one clinical cohort (GSE126044), 87.5% of high-risk patients failed to achieve remission following immunotherapy, highlighting the model's potential for guiding treatment selection.

The "Immune Desert" Phenotype

A critical finding was the strong correlation between high ARGRS scores and an "immune-cold" tumor microenvironment. High-risk tumors were characterized by high tumor purity but low "ImmuneScore," indicating a dense mass of cancer cells that effectively excludes protective immune cells. Cell-cell communication analysis suggested that these tumor cells interact with T-cells through inhibitory pathways, such as TIGIT-NECTIN2, potentially inducing T-cell exhaustion.

Genomic Complexity

High-ARGRS tumors were associated with increased genomic instability. Analysis of the mutation landscape revealed that the candidate genes themselves exhibit varying degrees of mutation, contributing to the overall complexity of the high-risk phenotype.

Mutation landscape of the 33 candidate anoikis-related genes across 1001 NSCLC samples, highlighting the genomic complexity associated with these markers.

HMGA1 as a Key Biomarker

Among the five signature genes, HMGA1 emerged as a primary driver of poor outcomes. It was highly expressed in tumor cells and showed a strong negative correlation with CD8+ T-cell infiltration. IHC staining confirmed that HMGA1 protein levels are significantly elevated in tumor tissues and inversely related to T-cell density, suggesting it may actively shape the immunosuppressive TME.

Discussion and Limitations

The ARGRS provides a novel framework for identifying "immune-cold" tumors in NSCLC, offering a potential roadmap for personalizing lung cancer therapy. By identifying patients likely to resist standard immunotherapy, clinicians can explore alternative or combination strategies earlier in the treatment process.

However, several limitations remain. The study is primarily retrospective, relying on existing databases; prospective clinical trials are necessary to validate the ARGRS in real-time clinical settings. Additionally, while the correlations are strong, the exact molecular mechanisms by which genes like HMGA1 suppress the immune system require further functional "wet-lab" studies to be fully understood.

Summary updated

Abstract

Non-small cell lung cancer (NSCLC) is the most common histological type of lung cancer. With the in-depth exploration of cell death manners, numerous studies found that anoikis is an important mechanism that associated with treatment. Therefore, we aimed to explore the prognostic value and treatment guidance of anoikis in NSCLC patients. In the current study, we first constructed a prognostic model based on the anoikis-related genes based on bulk RNA-sequencing and single-cell RNA-sequencing (scRNA-seq) dataset. Then, immuno-correlations of anoikis-related risk scores (ARGRS) were analyzed. In addition, HMGA1, a risky gene in ARGRS, was further explored to define its expression and immuno-correlation. Results showed that patients with higher ARGRS had worse clinical outcomes. Moreover, the five genes in the prognostic model were all highly expressed on tumor cells. Moreover, further analysis found that the ARGRS was negatively correlated with ImmuneScore, but positively with tumor purity. Besides, patients in the ARGRS-high group had lower levels of immunological characteristics, such as the immune-related signaling pathways and subpopulations. Additionally, in the immunotherapy cohorts, patients with the ARGRS-high phenotype were more resistant to immunotherapy and tended to not achieve remission after treatment. Last, HMGA1 was chosen as the representative biomarker, and analysis of the in-house cohort showed that HMGA1 was highly expressed in tumor tissues and correlated with decreased T cell infiltration. To sum up, ARGRS was correlated with a desert tumor microenvironment and identified immune-cold tumors, which can be a novel biomarker for the recognition of immunological characteristics and an immunotherapeutic response in NSCLC.
BibTeX
@inproceedings{wu2024anoikis,
  author = {Yixuan Wu and Zhou Zhou and Qianyi Qi and Shirong Xu and Lin Chen and Feng Wang},
  title = {{Anoikis-related gene signature is associated with immune infiltration and predicts the prognosis of non-small cell lung cancer}},
  booktitle = {Aging},
  year = {2024},
  doi = {10.18632/aging.205522},
}
Figure