Core Coefficient quantifies transcription-factor functional dominance for therapeutic target selection

Abstract

Transcription factors (TFs) operate in regulatory networks to control gene expression. The functional importance or leverage of a TF depends on the specific regulatory network in which it operates. Commonly used measurements of TF biology do not quantify the functional importance of TFs in these networks, which makes target selection for therapeutic intervention suboptimal. We developed the Core Coefficient (CC), a graph theory measure of TF functional dominance to overcome this limitation. Analysis of networks constructed from patient-biopsy and human post-mortem datasets revealed that these regulatory networks are organized in a functional hierarchy that is dominated by an identifiable and limited group of TFs, which we refer to as the Core. Analysis also revealed that the functional dominance of a TF, or the capacity to influence a cellular outcome, is a property of its position in the network hierarchy rather than its identity. While the Core consists of a group of functionally dominant TFs that control the network, the individual TFs forming that Core vary with cell type and cellular state. During the course of a given disease, from cancer to neurodegeneration, the functional dominance of TFs relevant to that disease shifts from low hierarchical importance to Core positions, whereas the functional dominance of other TFs is reduced or unchanged. CC can measure this shift in functional dominance and predict Core membership as a disease progresses. From a therapeutic standpoint, CC can also track the alterations in network hierarchy in diseased cells produced by treatments in the clinic. For example, network analyses of cancer cells before and after treatment from the same patients show dominant Core regulators shift from positions of high importance to positions of low importance after treatment. Experimental testing of these network mechanics in cancer cell lines showed CC can predict Core re-composition, signaling pathway reorganization and the cellular outcome of blocking a specific TF. In contrast, differential TF expression was informative in individual cases in these experiments but did not provide a general rule for measuring TF functional importance. This body of work introduces the CC, which quantifies TF functional dominance and enables the identification of targets for reprogramming TF networks to achieve a therapeutic objective.