Integrated Analysis of Cancer Tissue and Vitreous Humor from Retinoblastoma Eyes Reveals Unique Tumor-Specific Metabolic and Cellular Pathways in Advanced and Non-Advanced Tumors

Abstract

Retinoblastoma (Rb) is a pediatric intraocular malignancy that is proposed to originate from maturing cone cell precursors in the developing retina. The molecular mechanisms underlying the biological and clinical behaviors are important to understand in order to improve the management of advanced-stage tumors. While the genetic causes of Rb are known, an integrated understanding of the gene expression and metabolic processes in tumors of human eyes is deficient. By integrating transcriptomic profiling from tumor tissues and metabolomics from tumorous eye vitreous humor samples (with healthy, age-matched pediatric retinae and vitreous samples as controls), we uncover unique functional associations between genes and metabolites. We found distinct gene expression patterns between clinically advanced and non-advanced Rb. Global metabolomic analysis of the vitreous humor of the same Rb eyes revealed distinctly altered metabolites, indicating how tumor metabolism has diverged from healthy pediatric retina. Several key enzymes that are related to cellular energy production, such as hexokinase 1, were found to be reduced in a manner corresponding to altered metabolites; notably, a reduction in pyruvate levels. Similarly, E2F2 was the most significantly elevated E2F family member in our cohort that is part of the cell cycle regulatory circuit. Ectopic expression of the wild-type RB1 gene in the Rb-null Y79 and WERI-Rb1 cells rescued hexokinase 1 expression, while E2F2 levels were repressed. In an additional set of Rb tumor samples and pediatric healthy controls, we further validated differences in the expression of HK1 and E2F2. Through an integrated omics analysis of the transcriptomics and metabolomics of Rb, we uncovered a significantly altered tumor-specific metabolic circuit that reduces its dependence on glycolytic pathways and is governed by Rb1 and HK1.

Keywords

• retinoblastoma
• transcriptomics
• metabolomics
• multi-omics
• glycolysis
• metabolism