Original Article
Immuno-regulated common markers but different network signatures in two associated cancers: evidences from epigenetic treatment
Abstract
Background: Networks are now widely accepted inference tools in translational oncology. Besides providing agnostic model frameworks for complex data-driven clinical problems of diagnostic, therapeutic and prognostic impacts, networks mainly support insights, testable hypotheses and decision processes on the basis of their topological configurations and connectivity patterns.
Methods: The purpose of this study is to emphasize the role of both gene and network signatures in two specific cancers. Retinoblastoma (RB) and osteosarcoma are associated to some extent. It is known that patients who carry germline mutations in the RB1 gene, and who survive RB, are typically at an increased risk of early-onset second cancers, including osteosarcomas. Gene signatures are widely used, but also criticized for their partial lack of reproducibility. Network signatures include gene association dynamics by identifying modules or communities in which subsets of genes functionally belong.
Results: Two cancer cell lines (one per cancer type) were subjected to a similar epigenetic treatment regimen, using a demethylation agent (DAC, and including similar dose and time course administration). A minimal set of shared differentially expressed (DEG) genes was identified in cancer-specific cell lines from microarray analyses. However, the identified immune signatures were observed to translate into much diversified network signatures.
Conclusions: Our evidence is relevant to therapeutic developments, indicating that preference should be assigned to the assessment of bio-entities in a connected environment rather than considering single entities alone.
Methods: The purpose of this study is to emphasize the role of both gene and network signatures in two specific cancers. Retinoblastoma (RB) and osteosarcoma are associated to some extent. It is known that patients who carry germline mutations in the RB1 gene, and who survive RB, are typically at an increased risk of early-onset second cancers, including osteosarcomas. Gene signatures are widely used, but also criticized for their partial lack of reproducibility. Network signatures include gene association dynamics by identifying modules or communities in which subsets of genes functionally belong.
Results: Two cancer cell lines (one per cancer type) were subjected to a similar epigenetic treatment regimen, using a demethylation agent (DAC, and including similar dose and time course administration). A minimal set of shared differentially expressed (DEG) genes was identified in cancer-specific cell lines from microarray analyses. However, the identified immune signatures were observed to translate into much diversified network signatures.
Conclusions: Our evidence is relevant to therapeutic developments, indicating that preference should be assigned to the assessment of bio-entities in a connected environment rather than considering single entities alone.
