Identification of tumor antigens and immune subtypes in breast cancer for mRNA vaccine development

Ruo Qi Li, Wei Wang, Lei Yan, Song Li, Xin‐Yuan Guan, Wei Zhang, Jing LianFrontiers in Oncology 2022 · DOI

TL;DR

This study identifies three promising tumor antigens (CD74, IRF1, and PSME2) for breast cancer mRNA vaccine development and defines a distinct "immune-hot" subtype (Cluster B) that shows higher immune cell infiltration and better prognosis, making them ideal candidates for immunotherapy. The research also proposes biomarkers for assessing vaccine suitability and suggests personalized treatment strategies based on immune subtypes.

What Problem Does This Paper Address?

Breast cancer (BC) remains a significant health challenge, characterized by poor prognosis, resistance to conventional treatments, and severe side effects. While immunotherapy, particularly mRNA vaccines, holds immense promise, a major hurdle is the identification of effective tumor antigens and the selection of patient populations most likely to benefit. Current challenges include predicting which patients will respond to immunotherapy and tailoring treatments to individual needs. This paper aims to address these issues by identifying key tumor antigens and characterizing distinct immune subtypes within breast cancer to guide the development of more effective mRNA vaccines and personalized treatment strategies.

Methods

The researchers employed a multi-pronged approach integrating various bioinformatics and statistical analyses:

  • Data Acquisition: Gene expression, clinical data, and mutation profiles were obtained from large-scale public datasets: The Cancer Genome Atlas (TCGA) and the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC). Single-cell sequencing data was also utilized.
  • Tumor Antigen Identification: They analyzed gene expression, copy number variations, and mutation data from TCGA and cBioPortal to identify differentially expressed and frequently mutated genes that could serve as tumor-associated antigens (TAAs).
  • Immune Cell Infiltration Analysis: The Tumor Immune Estimation Resource (TIMER) was used to assess the correlation between TAA expression and immune cell infiltration.
  • Immune Subtyping: Consensus clustering and t-distributed stochastic neighbor embedding (t-SNE) were applied to gene expression data of immune-related genes to stratify BC patients into distinct immune subtypes.
  • Prognostic and Biomarker Analysis: Kaplan-Meier survival analysis (Overall Survival - OS, Disease-Free Survival - DFS) was performed to link TAAs and immune subtypes with patient outcomes. Receiver Operating Characteristic (ROC) curves were used to identify potential biomarkers for vaccine response.
  • Functional Enrichment Analysis: Gene Set Variation Analysis (GSVA) was employed to understand the biological pathways associated with different immune subtypes.
  • Drug Sensitivity Analysis: The pRRophetic R package was used to predict the sensitivity of different immune clusters to various anti-cancer drugs.
  • Weighted Gene Co-expression Network Analysis (WGCNA): This method was used to identify modules of co-expressed immune genes and their association with clinical phenotypes.

Datasets

  • The Cancer Genome Atlas (TCGA): Used for gene expression, patient prognosis, and clinicopathology data. Specifically, the TCGA-BRCA cohort was utilized.
  • Molecular Taxonomy of Breast Cancer International Consortium (METABRIC): Provided additional gene expression and clinical data for validation.
  • Single Cell Portal: Used for single-cell sequencing data from primary BC samples.
  • cBioPortal: Utilized for exploring and visualizing genetic alterations in BC.
  • IMMPORT database: Provided data on immune-associated genes.
  • NCG database: Used to extract tumor driver genes for BC.

Key Insights and Ideas

  • Identification of Key TAAs: The study pinpointed CD74, IRF1, and PSME2 as promising TAAs for mRNA vaccines. Their overexpression was linked to poorer prognosis, but importantly, also to increased infiltration of antigen-presenting cells and B cells, suggesting they can effectively prime an anti-tumor immune response.
  • Defining an "Immune-Hot" Subtype: The researchers successfully stratified BC patients into three distinct immune subtypes (Clusters A, B, and C). Cluster B emerged as a particularly promising group, characterized by a significantly higher abundance of immune cells, elevated expression of immune checkpoints and immunogenic cell death modulators, and a higher tumor mutation burden (TMB). This "immune-hot" profile suggests a favorable microenvironment for immunotherapy.
  • Biomarkers for Vaccine Suitability: The study identified specific genes (ABCG4, MYO1E, REXO2, USP41, and VTA1P1) that could serve as biomarkers to assess a patient's potential response to mRNA vaccination, offering a path towards personalized treatment selection.
  • Subtype-Specific Drug Sensitivities: The analysis revealed that different immune subtypes exhibit varying sensitivities to chemotherapy drugs. This highlights the critical need for individualized treatment approaches, suggesting that patients in Cluster B might benefit more from immunotherapy, while others might respond better to specific chemotherapies.
  • Validation Across Cohorts: The immune subtype classification and its association with prognosis and immune infiltration were validated in an independent cohort (METABRIC), strengthening the robustness of the findings.

Key Results

  • Three TAAs Identified: CD74, IRF1, and PSME2 were identified as TAAs associated with prognosis and immune cell infiltration.
  • Three Immune Subtypes Defined: Patients were classified into three clusters (A, B, C), with Cluster B exhibiting a more favorable immune microenvironment and prognosis.
  • Cluster B as Ideal for Vaccination: Cluster B patients showed significantly higher immune cell infiltration, elevated expression of immune checkpoints and immunogenic cell death modulators, and higher TMB, indicating suitability for mRNA vaccination.
  • Biomarker Genes: Five genes were identified as potential biomarkers for assessing mRNA vaccination response.
  • Subtype-Specific Drug Responses: Significant differences in IC50 values were observed for various anti-cancer drugs across the identified immune clusters, underscoring the potential for personalized therapy.

Limitations and Future Work

The authors acknowledge several limitations:

  • Retrospective Data: The study relied on retrospective data from public databases, which may be subject to selection bias.
  • Impact of Immune Cell Infiltration: The direct impact of the identified potential antigens on the efficacy of vaccines in tumor-infiltrating immune cells requires further verification through subsequent experimental studies.

Future work should focus on:

  • Experimental Validation: Conducting in vitro and in vivo experiments to validate the identified TAAs (CD74, IRF1, PSME2) and their immunogenicity.
  • Clinical Trials: Designing and conducting clinical trials to assess the efficacy of mRNA vaccines targeting these TAAs in the identified "immune-hot" patient subtype (Cluster B).
  • Personalized Treatment Strategies: Further exploring and validating the proposed biomarkers for vaccine suitability and refining personalized treatment plans based on immune subtype and drug sensitivity profiles.
  • Investigating Other Immune Cells: Further research into the role of specific immune cells and their interactions with TAAs in the context of mRNA vaccination.

Summary updated

Abstract

Our findings suggest new directions for the development of mRNA vaccines against breast cancer.
BibTeX
@inproceedings{li2022identification,
  author = {Ruo Qi Li and Wei Wang and Lei Yan and Song Li and Xin‐Yuan Guan and Wei Zhang and Jing Lian},
  title = {{Identification of tumor antigens and immune subtypes in breast cancer for mRNA vaccine development}},
  booktitle = {Frontiers in Oncology},
  year = {2022},
  doi = {10.3389/fonc.2022.973712},
}
Figure