Combined analysis of entosis and disulfidptosis to predict the prognosis and tumor microenvironment (TME) landscape of hepatocellular carcinoma

Shuai Hu, Kai Huang, Guangquan Zhang, Yongfei He, Junming Xu, Hang Zhai, Jicai Wang, Shengjie Hong, Fenfang Wu, Xianjie ShiResearch Square 2024 · DOI

TL;DR

This study develops a novel prognostic model for hepatocellular carcinoma (HCC) by integrating two distinct cell death mechanisms: entosis and disulfidptosis. By identifying a two-gene signature (KIF2C and SLC7A11), the researchers successfully stratified patients into high- and low-risk groups, providing a roadmap for personalized immunotherapy and targeted drug selection.

What Problem Does This Paper Address?

Hepatocellular carcinoma (HCC) remains a global health challenge with high mortality rates and complex molecular drivers. While traditional cell death pathways like apoptosis are well-studied, newer mechanisms—specifically entosis (where one cell "eats" another) and disulfidptosis (cell death triggered by disulfide stress under glucose deficiency)—have not been combined to predict patient outcomes.

The authors sought to bridge this gap, investigating whether the interplay between these two pathways could better predict a patient's prognosis and reveal the "landscape" of the tumor microenvironment (TME). Understanding this landscape is crucial for determining why some patients respond to immunotherapy while others do not.

Methods

The researchers employed a multi-step bioinformatic pipeline:

  1. Gene Screening: They started with 22 disulfidptosis-related genes (DRGs) and 25 entosis-related genes (ERGs). Using the Deseq2 R package, they identified 15 genes that were significantly differentially expressed in tumor vs. normal tissue.
  2. Model Construction: Through single-factor Cox regression and LASSO (Least Absolute Shrinkage and Selection Operator) regression, they narrowed the field to two key prognostic genes: KIF2C and SLC7A11.
  3. Validation: The resulting "DRG-ERG score" was used to create a nomogram (a predictive tool) and was validated using an independent external dataset.
  4. TME & Drug Analysis: They used tools like Cibersort and estimate to analyze immune cell infiltration and OncoPredict to simulate how different risk groups might respond to specific chemotherapy and targeted drugs.

Datasets

  • Training Set: RNA-seq profiles and clinical data from 369 HCC patients and 50 normal controls from The Cancer Genome Atlas (TCGA-LIHC).
  • Validation Set: Clinical information from 61 liver cancer patients from the International Cancer Genome Consortium (ICGC).
  • Gene Databases: Disulfidptosis genes were sourced from FerrDb V2, and entosis genes were pulled from GeneCards.

Key Insights and Ideas

  • The Glucose Connection: Both entosis and disulfidptosis are heavily influenced by glucose scarcity. In the nutrient-poor environment of a tumor, cells either cannibalize neighbors (entosis) or succumb to internal chemical stress (disulfidptosis). This paper is the first to link them for HCC prognosis.
  • The Power of Two: Despite starting with 47 genes, the model's heavy lifting is done by just two:
  • SLC7A11: A transporter involved in glutathione synthesis; its overexpression is a hallmark of disulfidptosis vulnerability.
  • KIF2C: A protein involved in chromosome separation during mitosis, often linked to poor prognosis in various cancers.
  • Immune "Purity": The study found that low-risk tumors were "less pure," meaning they contained more stromal and immune cells. Conversely, high-risk patients showed a higher sensitivity to immune checkpoint therapies, suggesting that the DRG-ERG score could be a biomarker for immunotherapy eligibility.

Key Results

  • Strong Predictive Accuracy: The model achieved Area Under the Curve (AUC) values for 1-year survival of 0.778 in the training set and 0.800 in the validation set, indicating high reliability.
  • Survival Disparity: Patients in the low-risk group had significantly higher overall survival rates compared to the high-risk group (p < 0.001).
  • Drug Sensitivity Profiles:
    • High-Risk Group: Showed higher sensitivity to fluorouracil, imatinib, and daraparib.
  • Low-Risk Group: Showed higher sensitivity to gemcitabine, olaparib, and darafenib.
  • Mutation Landscape: The high-risk group featured a higher frequency of mutations in genes like CTNNB1, TTN, and TP53.

Limitations and Future Work

The authors acknowledge that their findings are primarily based on publicly available bioinformatic databases. While the statistical correlations are robust, the study lacks "wet lab" validation.

Future directions include: * Performing in vitro (cell culture) and in vivo (animal model) experiments to observe the physical interaction between entosis and disulfidptosis. * Validating the six identified sensitive drugs in clinical settings to see if the DRG-ERG score truly predicts treatment response in real-world patients.

Summary updated

Abstract

<title>Abstract</title> Background Hepatocellular carcinoma (HCC) is a common malignant digestive tract tumor worldwide. Disulfidptosis and entosis are two special cell death mechanisms. The purpose of this study was to investigate the prognostic and clinical applicability of these two cell death-related genes in HCC. Methods The R package "Deseq2" was used to conduct differential expression analysis, 22 disulfidptosis-related genes and 25 entosis-related genes were screened out to 15 differential genes, and 8 key genes were subsequently screened out by single-factor Cox regression analysis using the R package “survival”. Finally, LASSO regression analysis was performed with the R package “glment” to screen 2 prognostic genes. A prognostic prediction model was constructed based on this scoring system. Then, immunoinfiltration analysis was performed with the R packages “IBOR” and “estimate”. Drug sensitivity analysis was performed with the R package “OncoPredict” to screen for liver cancer-sensitive drugs. Results A hepatocellular carcinoma prognosis prediction model based on the KIF2C and SLC7A11 scoring system was constructed with good predictive ability, and the KIF2C and SLC7A11 scoring systems were applied to screen immunotherapy targets and sensitive drugs related to HCC.
BibTeX
@inproceedings{hu2024combined,
  author = {Shuai Hu and Kai Huang and Guangquan Zhang and Yongfei He and Junming Xu and Hang Zhai and Jicai Wang and Shengjie Hong and Fenfang Wu and Xianjie Shi},
  title = {{Combined analysis of entosis and disulfidptosis to predict the prognosis and tumor microenvironment (TME) landscape of hepatocellular carcinoma}},
  booktitle = {Research Square},
  year = {2024},
  doi = {10.21203/rs.3.rs-4662221/v1},
}
Figure