Beyond cancer genes: colorectal cancer as robust intrinsic states formed by molecular interactions.

Ruoshi Yuan,Xiaomei Zhu,Suzhan Zhang,Runtan Cheng,Yanqin Huang,Demin Lu,Ping Ao,Leroy Hood,Jiekai Yu,Sui Huang,Shu Zheng

doi:10.1098/rsob.170169

Abstract

Colorectal cancer (CRC) has complex pathological features that defy the linear-additive reasoning prevailing in current biomedicine studies. In pursuing a mechanistic understanding behind such complexity, we constructed a core molecular–cellular interaction network underlying CRC and investigated its nonlinear dynamical properties. The hypothesis and modelling method has been developed previously and tested in various cancer studies. The network dynamics reveal a landscape of several attractive basins corresponding to both normal intestinal phenotype and robust tumour subtypes, identified by their different molecular signatures. Comparison between the modelling results and gene expression profiles from patients collected at the second affiliated hospital of Zhejiang University is presented as validation. The numerical ‘driving’ experiment suggests that CRC pathogenesis may depend on pathways involved in gastrointestinal track development and molecules associated with mesenchymal lineage differentiation, such as Stat5, BMP, retinoic acid signalling pathways, Runx and Hox transcription families. We show that the multi-faceted response to immune stimulation and therapies, as well as different carcinogenesis and metastasis routes, can be straightforwardly understood and analysed under such a framework.

Highlights

Colorectal cancer (CRC) is a leading cause of cancer deaths in the USA [1] and around the world [2,3]
We investigated CRC under the framework that cancer is a robust state(s) evolutionarily formed from the underpinning endogenous molecular –cellular interaction network [7,8]
The nonlinearity of the interaction network dynamics is directly responsible for the formation of the robust CRC subtypes and their differential responses to intervention

Summary

Introduction

Colorectal cancer (CRC) is a leading cause of cancer deaths in the USA [1] and around the world [2,3]. The current research on combined molecular-targeting agents [4], the need for better risk models to incorporate genetic, lifestyle and environmental effects [5], and the development for early detection method [6] all require a better understanding at the molecular level. A causal and quantitative model may be used as a ‘dry-experiment’ platform to recapitulate carcinogenesis and metastasis routes, and to test efficacy of drug combinations. Towards such a goal, we investigated CRC under the framework that cancer is a robust state(s) evolutionarily formed from the underpinning endogenous molecular –cellular interaction network [7,8]. The nonlinearity of the interaction network dynamics is directly responsible for the formation of the robust CRC subtypes and their differential responses to intervention

Methods

Results

Conclusion