Abstract
Background: Diabetes mellitus is frequently associated with microvascular and macrovascular lesions, where hyperglycemia-induced vascular endothelial cell damage serves as a critical initiating factor. Methods: This study investigated long non-coding RNAs (lncRNAs) and mRNAs affected by hyperglycemia-induced damage using human umbilical vein endothelial cells (HUVECs). Cells were cultured under high (25 mmol/L) or normal (5 mmol/L) glucose conditions for 6 days, followed by lncRNA and protein-coding transcript profiling via RNA-seq. Results: Among 40,379 lncRNAs screened, 214 were upregulated and 197 were downregulated in response to high glucose. Of 28,431 protein-coding genes, 778 were upregulated and 998 were downregulated. A total of 945 lncRNA/mRNA pairs were identified, including 126 differentially expressed lncRNAs predicted to target 201 mRNAs. The resulting lncRNA-mRNA network comprised 354 lncRNA nodes, 1,167 mRNA nodes, and 9,735 edges. Several hub lncRNAs, such as ENST00000600527 and NONHSAT037576.2, were identified as potential key regulators. Conclusion: These findings provide novel insights into the regulatory molecules and pathways underlying hyperglycemia-related endothelial dysfunction in diabetes-associated vascular disease.
Introduction
Diabetes mellitus is notorious for damaging blood vessels, leading to both microvascular complications, such as retinopathy and nephropathy, and macrovascular complications, including coronary artery disease and stroke. At the core of these pathologies is endothelial dysfunction—the progressive breakdown of the cells lining the blood vessels due to chronic hyperglycemia. While the roles of protein-coding genes in this process have been extensively studied, the "dark matter" of the genome—long non-coding RNAs (lncRNAs)—remains less understood. lncRNAs are transcripts longer than 200 nucleotides that do not code for proteins but act as critical regulators of gene expression. This study aims to map the lncRNA landscape in vascular endothelial cells under high-glucose stress to identify potential therapeutic targets.
Materials and Methods
Human Umbilical Vein Endothelial Cells (HUVECs) were utilized as the primary experimental model. Cells were cultured for six days in either normal glucose (5 mmol/L) or high glucose (25 mmol/L) environments. Mannitol was employed as an osmotic control to ensure that observed cellular changes were specifically due to glucose toxicity rather than osmotic pressure.
Total RNA was extracted and subjected to deep RNA-sequencing (RNA-seq). The bioinformatics pipeline included FastQC for quality control, Hisat2 for mapping, and Stringtie for transcript assembly. Differentially expressed (DE) transcripts were identified using a threshold of |log2(fold-change)| > 1 and a False Discovery Rate (FDR) < 0.05. To elucidate the regulatory relationships, Pearson correlation coefficients (PCC) were calculated to construct lncRNA-mRNA co-expression networks. Finally, quantitative real-time PCR (qRT-PCR) was used to validate the expression levels of key transcripts.
Results
Transcriptomic Landscape and Pathway Enrichment
The RNA-seq analysis revealed a massive shift in the transcriptomic profile of HUVECs exposed to high glucose. Out of over 40,000 lncRNAs screened, 411 were significantly altered (214 upregulated, 197 downregulated). Similarly, 1,776 mRNAs showed differential expression (778 upregulated, 998 downregulated).
Functional enrichment analysis indicated that the affected genes were heavily involved in the AGE-RAGE signaling pathway, a well-documented driver of diabetic vascular complications. Other enriched pathways included cell cycle regulation and various metabolic processes, suggesting that high glucose disrupts fundamental cellular homeostasis.
Regulatory Networks and Hub Identification
The study distinguished between cis-regulation (acting on nearby genes) and trans-regulation (influencing distant genes). Trans-regulation appeared more prevalent, involving 715 specific interactions. The constructed lncRNA-mRNA network was remarkably dense, consisting of 354 lncRNA nodes and 1,167 mRNA nodes connected by 9,735 edges.
Several "hub" lncRNAs were identified based on their high degree of connectivity, suggesting they act as master switches. Notable hubs included DAPK1-IT1 and ENST00000600527. These molecules are likely central to the orchestration of inflammatory responses and programmed cell death in the endothelium. Validation via qRT-PCR for six selected lncRNAs, including ENST00000444438 and NONHSAT108582.2, showed high concordance with the RNA-seq data, confirming the reliability of the high-throughput findings.
Discussion
This comprehensive profiling highlights the complexity of the non-coding regulatory response to hyperglycemia. The identification of specific lncRNA-mRNA networks provides a roadmap for understanding how the "dark matter" of the genome contributes to diabetic vasculopathy.
However, several limitations remain. As an in vitro study using HUVECs, the results may not fully capture the physiological complexities of the human vasculature, which includes hemodynamic forces and interactions with other cell types like smooth muscle cells and leukocytes. Furthermore, while the co-expression networks suggest regulatory links, functional validation using CRISPR-mediated knockouts or over-expression assays is required to prove causality. Future research should also explore whether these differentially expressed lncRNAs can serve as circulating biomarkers for early detection of vascular damage in diabetic patients.
Summary updated