A novel genomic instability-derived lncRNA signature to predict prognosis and immune characteristics of pancreatic ductal adenocarcinoma

Huijie Yang, Weiwen Zhang, Jin Ding, Jingyi Hu, Yi Eve Sun, Weijun Peng, Yi Chu, Lingxiang Xie, Zubing Mei, Zhuo Shao, Yang XiaoFrontiers in Immunology 2022 · DOI

TL;DR

Researchers have developed a new way to predict the survival of pancreatic cancer patients using a signature of five long non-coding RNAs (lncRNAs) linked to genomic instability. This "GILncSig" not only predicts prognosis but also offers insights into the tumor's microenvironment, suggesting potential avenues for more personalized immunotherapies.

What Problem Does This Paper Address?

Pancreatic ductal adenocarcinoma (PDAC) is a notoriously aggressive cancer with a grim prognosis and limited treatment options. Early diagnosis is challenging, and many patients don't respond well to existing therapies. While genomic instability is a known hallmark of cancer and a prognostic factor, its specific role and how it's regulated in PDAC, particularly by long non-coding RNAs (lncRNAs), remain underexplored. This paper aims to bridge this gap by identifying lncRNAs associated with genomic instability in PDAC and developing a predictive model that can guide treatment strategies.

Methods

The study employed a multi-pronged approach:

  • Computational Framework for Identifying Genomic Instability-Associated lncRNAs: They developed a method to identify lncRNAs linked to genomic instability by comparing gene expression profiles of "genomic instability (GU)" and "genomic stability (GS)" patient groups, defined by their somatic mutation counts.
  • Prognostic Signature Development: Using a multivariate Cox proportional hazards regression analysis on a training dataset, they screened for lncRNAs that independently predicted overall survival in PDAC patients. This led to the construction of the GILncSig model.
  • Validation: The GILncSig model was rigorously validated using independent testing and combined datasets from The Cancer Genome Atlas (TCGA).
  • Functional Enrichment Analysis: To understand the biological implications of the identified lncRNAs and the GILncSig, they used graph-based clustering and gene set enrichment analysis to explore associations with epithelial-mesenchymal transition (EMT) and tumor microenvironment (TME) characteristics.
  • Tumor Microenvironment (TME) Estimation: The CIBERSORT algorithm was used to estimate the abundance of different immune cell subsets within the TME, and the ESTIMATE algorithm was used to calculate the Immune Score.
  • Real-time Quantitative PCR (qRT-PCR): This technique was used for experimental validation of lncRNA expression in a small set of patient tissue samples.

Datasets

  • The Cancer Genome Atlas (TCGA): This was the primary source for RNA-seq expression data, clinical features, clinicopathological characteristics, survival information, and somatic mutation information for pancreatic ductal adenocarcinoma patients.
  • TANRIC Database: Used for lncRNA expression data.
  • Five PDAC Tissue Specimens: A small set of formalin-fixed paraffin-embedded pancreatic ductal adenocarcinoma tumor specimens and their adjacent normal tissues were used for qRT-PCR validation.

Key Insights and Ideas

  • Genomic Instability as a lncRNA Regulator: The study highlights that lncRNAs are not just bystanders but actively participate in maintaining or reflecting genomic instability in PDAC.
  • GILncSig: More Than Just a Prognostic Score: The developed signature isn't just a number; it's linked to fundamental cancer processes like EMT and the immune landscape of the tumor. This suggests it could be a window into the tumor's behavior.
  • EMT and Immune Evasion Link: The research strongly suggests that high GILncSig scores are associated with increased EMT (making cancer cells more mobile and invasive) and a suppressed adaptive immune response within the tumor. This dual threat likely contributes to the poor prognosis.
  • Potential for Targeted Immunotherapy: By understanding the immune profile associated with high GILncSig, clinicians might be able to better select patients who would benefit from specific immunotherapies or develop strategies to overcome immune evasion.

Key Results

  • Identification of 206 Genomic Instability-Associated lncRNAs: The computational framework successfully identified a significant number of lncRNAs linked to genomic instability in PDAC.
  • Development of a 5-lncRNA Prognostic Signature (GILncSig): Five specific lncRNAs were identified as independent prognostic factors, forming the GILncSig.
  • Robust Prognostic Performance: GILncSig demonstrated significant prognostic value across training, testing, and combined TCGA cohorts, with AUC values around 0.72.
  • Independent Prognostic Factor: GILncSig was validated as an independent prognostic factor, even when considering other clinical variables like age and tumor grade.
  • Association with EMT: Patients with high GILncSig scores showed significantly higher EMT signature scores, indicating a more invasive tumor phenotype.
  • Impaired Adaptive Immunity: The high-risk GILncSig group exhibited a significantly diminished presence of key adaptive immune cells (e.g., B cells, CD8+ T cells) in the tumor microenvironment, suggesting immune evasion.
  • KRAS Mutation Correlation: GILncSig was able to identify a subpopulation of patients with KRAS mutations who had a worse prognosis, suggesting its utility in refining risk stratification.

Limitations and Future Work

  • Dataset Size: The TCGA cohort used (171 patients) is considered relatively small compared to datasets for other cancer types.
  • Independent Validation: GILncSig was primarily validated within the TCGA dataset and a small set of patient specimens. The authors acknowledge the need for larger, independent datasets for further validation.
  • Functional Studies: While bioinformatics analyses suggest links to EMT and TME, further in-depth functional studies are needed to elucidate the exact regulatory mechanisms of GILncSig in maintaining genomic instability and its downstream effects.
  • Immunotherapy Data: The TCGA dataset lacked immunotherapy data, meaning the predictive power of GILncSig for immunotherapy response was indirectly assessed. Future studies should incorporate this to guide treatment decisions.
  • Experimental Validation: While qRT-PCR was performed on a small sample, more extensive experimental validation of the identified lncRNAs and their target genes is warranted.

Summary updated

Abstract

Our study established a genomic instability-associated lncRNAs-derived model (GILncSig) for prognosis prediction in patients with PDAC, and revealed the potential functional regulatory role of GILncSig.
BibTeX
@inproceedings{yang2022novel,
  author = {Huijie Yang and Weiwen Zhang and Jin Ding and Jingyi Hu and Yi Eve Sun and Weijun Peng and Yi Chu and Lingxiang Xie and Zubing Mei and Zhuo Shao and Yang Xiao},
  title = {{A novel genomic instability-derived lncRNA signature to predict prognosis and immune characteristics of pancreatic ductal adenocarcinoma}},
  booktitle = {Frontiers in Immunology},
  year = {2022},
  doi = {10.3389/fimmu.2022.970588},
}
Figure