Annotation and functional characterization of long noncoding RNAs deregulated in pancreatic adenocarcinoma

Unveiling the role of PYGB in pancreatic cancer: a novel diagnostic biomarker and gene therapy target

PURPOSE

Pancreatic cancer (PC) is a highly malignant tumor that poses a severe threat to human health. Brain glycogen phosphorylase (PYGB) breaks down glycogen and provides an energy source for tumor cells. Although PYGB has been reported in several tumors, its role in PC remains unclear.

METHODS

We constructed a risk diagnostic model of PC-related genes by WGCNA and LASSO regression and found PYGB, an essential gene in PC. Then, we explored the pro-carcinogenic role of PYGB in PC by in vivo and in vitro experiments.

RESULTS

We found that PYGB, SCL2A1, and SLC16A3 had a significant effect on the diagnosis and prognosis of PC, but PYGB had the most significant effect on the prognosis. Pan-cancer analysis showed that PYGB was highly expressed in most of the tumors but had the highest correlation with PC. In TCGA and GEO databases, we found that PYGB was highly expressed in PC tissues and correlated with PC's prognostic and pathological features. Through in vivo and in vitro experiments, we found that high expression of PYGB promoted the proliferation, invasion, and metastasis of PC cells. Through enrichment analysis, we found that PYGB is associated with several key cell biological processes and signaling pathways. In experiments, we validated that the MAPK/ERK pathway is involved in the pro-tumorigenic mechanism of PYGB in PC.

CONCLUSION

Our results suggest that PYGB promotes PC cell proliferation, invasion, and metastasis, leading to poor patient prognosis. PYGB gene may be a novel diagnostic biomarker and gene therapy target for PC.

A novel sphingolipid metabolism-related long noncoding RNA signature predicts the prognosis, immune landscape and therapeutic response in pancreatic adenocarcinoma

Sphingolipid metabolism affects prognosis and resistance to immunotherapy in patients with cancer and is an emerging target in cancer therapy with promising diagnostic and prognostic value. Long noncoding ribonucleic acids (lncRNAs) broadly regulate tumour-associated metabolic reprogramming. However, the potential of sphingolipid metabolism-related lncRNAs in pancreatic adenocarcinoma (PAAD) is poorly understood. In this study, coexpression algorithms were employed to identify sphingolipid metabolism-related lncRNAs. The least absolute shrinkage and selection operator (LASSO) algorithm was used to develop a sphingolipid metabolism-related lncRNA signature (SMLs). The prognostic predictive stability of the SMLs was validated using Kaplan-Meier. Univariate and multivariate Cox, receiver operating characteristic (ROC) and clinical stratification analyses were used to comprehensively assess the SMLs. Gene set variation analysis (GSVE), gene ontology (GO) and tumor mutation burden (TMB) analysis explored the potential mechanisms. Additionally, single sample gene set enrichment analysis (ssGSEA), ESTIMATE, immune checkpoints and drug sensitivity analysis were used to investigate the potential predictive function of the SMLs. Finally, an SMLs-based consensus clustering algorithm was utilized to differentiate patients and determine the suitable population for immunotherapy. The results showed that the SMLs consists of seven sphingolipid metabolism-related lncRNAs, which can well determine the clinical outcome of individuals with PAAD, with high stability and general applicability. In addition, the SMLs-based consensus clustering algorithm divided the TCGA-PAAD cohort into two clusters, with Cluster 1 showing better survival than Cluster 2. Additionally, Cluster 1 had a higher level of immune cell infiltration than Cluster 2, which combined with the higher levels of immune checkpoints in Cluster 1 suggests that Cluster 1 is more consistent with an immune 'hot tumor' profile and may respond better to immune checkpoint inhibitors (ICIs). This study offers new insights regarding the potential role of sphingolipid metabolism-related lncRNAs as biomarkers in PAAD. The constructed SMLs and the SMLs-based clustering are valuable tools for predicting clinical outcomes in PAAD and provide a basis for clinical selection of individualized treatments.

Comprehensive analysis of basement membrane and immune checkpoint related lncRNA and its prognostic value in hepatocellular carcinoma via machine learning

Background

Hepatocellular carcinoma (HCC), which is characterized by its high malignancy, generally exhibits poor response to immunotherapy. As part of the tumor microenvironment, basement membranes (BMs) are involved in tumor development and immune activities. Presently, there is no integrated analysis linking the basement membrane with immune checkpoints, especially from the perspective of lncRNA.

Methods

Based on transcriptome data from The Cancer Genome Atlas, BMs-related and immune checkpoint-related lncRNAs were identified. By applying univariable Cox regression and Machine learning (LASSO and SVM-RFE algorithm), a 10-lncRNA prognosis signature was constructed. The prognostic significance of this signature was assessed by survival analysis. GSEA, ssGSEA, and drug sensitivity analysis were conducted to investigate potential functional pathways, immune status, and clinical implications of guiding individual treatments in HCC. Finally, the promoting migration effect of LINC01224 was validated via in vitro experiments.

Results

The multiple Cox regression, receiver operating characteristic curves, and stratified survival analysis of clinical subgroups exhibited the robust prognostic ability of the lncRNA signature. Results of the GSEA and drug sensitivity analysis revealed significant differences in potential functional pathways and response to drugs between the two risk groups. In addition, the risk level of HCC patients was distinctly correlated with immune cell infiltration status. More importantly, LINC01224 was independently associated with the OS of HCC patients (P < 0.05), suppressing the expression of LINC01224 inhibited the migration of HCC cells.

Conclusion

This study developed a reliable signature for the prognosis of HCC based on BM and immune checkpoint related lncRNA, revealing that LINC01224 might be a prognostic biomarker for HCC associated with the progression of HCC.