Development and validation of immunogenic cell death-related signature for predicting the prognosis and immune landscape of uveal melanoma

Yuanyuan Hu, Jiayang Cai, Meng Ye, Qianxue Mou, Bowen Zhao, Qian Sun, Xiaotong Lou, Hong Zhang, Yin ZhaoFrontiers in Immunology 2022 · DOI

TL;DR

Uveal melanoma (UM) is a rare but aggressive eye cancer with poor prognosis once it metastasizes. This study identifies a novel 5-gene signature related to immunogenic cell death (ICD) that accurately predicts UM patient prognosis and reveals distinct immune microenvironments associated with different risk levels. This signature could pave the way for more effective immunotherapies for UM.

What Problem Does This Paper Address?

Uveal melanoma (UM) is the most common primary intraocular malignancy, but it has a high propensity for metastasis, leading to a poor prognosis for affected patients. Current treatments are often limited in their effectiveness, especially once the cancer has spread. Identifying reliable biomarkers is crucial for predicting patient outcomes and guiding therapeutic strategies. Immunogenic cell death (ICD) is a form of programmed cell death that can trigger adaptive immune responses and long-term immune memory, making it a promising avenue for cancer therapy. However, the role of ICD and its associated biomarkers in UM, particularly in predicting prognosis and understanding the tumor's immune landscape, remains underexplored. This paper aims to address this gap by developing and validating an ICD-related gene signature for UM.

Methods

The researchers employed a multi-faceted approach combining bioinformatics analysis with experimental validation. They utilized data from The Cancer Genome Atlas (TCGA) as a training set and the Gene Expression Omnibus (GEO) database as a validation set.

Key methods included:

  • Differential Gene Expression Analysis: Comparing ICD-related gene expression between normal and UM tissues using Wilcoxon tests.
  • Consensus Clustering: Identifying distinct molecular subtypes of UM based on ICD-related gene expression.
  • LASSO Cox Regression: Constructing a prognostic risk signature by selecting the most relevant ICD-related genes.
  • Immune Microenvironment Analysis: Assessing stromal score, immune score, and tumor purity, as well as the expression of immune checkpoints (ICPs) and human leukocyte antigen (HLA) genes.
  • Somatic Mutation Analysis: Investigating the mutational landscape of UM.
  • Prognostic Analysis: Using Kaplan-Meier (KM) curves and Cox regression to evaluate the predictive power of the risk signature.
  • Cellular Experiments: Validating the role of a key gene (FOXP3) in UM cell proliferation using siRNA knockdown, CCK-8 assays, and EdU-DNA synthesis assays.

Datasets

  • TCGA (The Cancer Genome Atlas): Used as the primary training set, comprising RNA-seq expression data and clinical data from 80 uveal melanoma patients.
  • GEO (Gene Expression Omnibus): Used as the validation set, including RNA-seq transcriptome information from 57 uveal melanoma patients (accession number: GSE44299).
  • GEO (Gene Expression Omnibus): Normal retinal pigmented epithelium (RPE)-choroid complex samples from 175 individuals were used as a normal control group (accession number: GSE29801).
  • GDSC (Genomics of Drug Sensitivity in Cancer) database: Used for forecasting drug sensitivity.
  • CellMiner and GSCA databases: Also used for drug sensitivity analysis.

Key Insights and Ideas

  • ICD as a Prognostic Indicator in UM: The study establishes that ICD-related genes are not just involved in immune responses but also hold significant prognostic value in UM.
  • Distinct Immune Phenotypes: The research clearly delineates two UM subtypes: an "ICD-high" group associated with an "immune-hot" phenotype (higher immune infiltration, more immune checkpoints) and a "cold" phenotype in the "ICD-low" group. This distinction is crucial for understanding treatment responses.
  • FOXP3's Dual Role: While FOXP3 is often associated with immune suppression, this study highlights its role in promoting UM tumor growth, suggesting it could be a therapeutic target.
  • Risk Score as a Comprehensive Predictor: The developed 5-gene risk signature effectively predicts not only overall survival but also correlates with tumor stage and immune infiltration, making it a potentially valuable clinical tool.

Key Results

  • Differential Gene Expression: Approximately half of the 34 ICD-related genes showed significantly different expression levels between normal and UM tissues, with many being highly expressed in UM.
  • ICD Subtypes and Prognosis: Consensus clustering identified two subtypes: ICD-high and ICD-low. The ICD-low subtype demonstrated a better prognostic potential.
  • 5-Gene Risk Signature: A LASSO Cox regression identified a 5-gene signature that was validated in both TCGA and GEO datasets as a significant predictor of overall survival (OS).
  • Immune Landscape Differences: The high-risk group exhibited higher ESTIMATE, stromal, and immune scores, along with lower tumor purity, indicating increased immune cell infiltration. They also showed significantly upregulated expression of immune checkpoints (ICPs) and HLA genes.
  • FOXP3 Validation: In cellular experiments, FOXP3 mRNA and protein levels were higher in UM cell lines compared to normal cells. Knockdown of FOXP3 significantly inhibited UM cell proliferation and DNA synthesis.
  • Drug Sensitivity: The high-risk group showed increased sensitivity to several drugs, suggesting potential therapeutic implications.

Limitations and Future Work

  • Limited Experimental Validation: While cellular experiments validated FOXP3's role, further in vivo studies would strengthen these findings.
  • Generalizability of FOXP3: The study focused on FOXP3's role in proliferation. Its broader impact on immune evasion and metastasis in UM warrants further investigation.
  • Clinical Translation: The developed risk signature needs prospective validation in larger, diverse patient cohorts to confirm its clinical utility.
  • Mechanism of ICD in UM: While the study links ICD to immune microenvironment and prognosis, the precise mechanisms by which ICD influences UM progression and immune evasion require deeper exploration.
  • Therapeutic Strategies: The findings suggest that targeting ICD pathways or leveraging the immune-hot phenotype of high-risk patients could be promising therapeutic strategies, which should be explored in future research.

Summary updated

Abstract

ICD-related genes play a critical role in the tumor immune microenvironment. Our results may contribute to the development of effective immunotherapies.
BibTeX
@inproceedings{hu2022development,
  author = {Yuanyuan Hu and Jiayang Cai and Meng Ye and Qianxue Mou and Bowen Zhao and Qian Sun and Xiaotong Lou and Hong Zhang and Yin Zhao},
  title = {{Development and validation of immunogenic cell death-related signature for predicting the prognosis and immune landscape of uveal melanoma}},
  booktitle = {Frontiers in Immunology},
  year = {2022},
  doi = {10.3389/fimmu.2022.1037128},
}
Figure