In the multicellular organism, cells develop from a single pluripotent/stem cell and differentiate into different cell types during proliferation. The gene regulatory network (GRN) plays an important role in this process. To understand the cell differentiation, C.H. Waddington proposed a metaphor of the epigenetic landscape in 1960s that cell differentiation was like a ball rolling down the hill reined by the GRN.
As the interaction between different genes is complicated and highly nonlinear, the energy landscape theory utilizes a non-equilibrium potential to depict the steady state distribution of the considered GRN system with noise, and construct an approximated gradient system which is illustrative and easy to describe the differentiation process. There are mainly two approaches to construct the energy landscape for biosystems. One is data-based, which includes the population balance analysis (PBA) and the landscape of differentiation dynamics (LDD). The other is model-based, including Wang’s landscape and quasi-potential theory.
In a new research article published in National Science Review, scientists from Chinese Academy of Sciences, Peking University, the University of Tokyo, and Fudan University present a novel energy landscape decomposition (ELD) theory under the consideration of cell birth and death. ELD decomposes the dynamics of cell differentiation into two important energy landscapes and a curl part term. The cell-type landscape U(x) can characterize different cell types by its metastable states, while the pluripotency landscape V(x) can indicate the pluripotency/stemness by its values and depict the differentiation direction by its negative gradient. Feasible numerical methods and mean-field approximation for high dimensional systems to make the decomposition are also presented in the article. Successful applications to drift-diffusion process, 2-gene fate decision system and T-cell differentiation process validate the ELD theory and method.
This unified theory of energy landscape for potential landscapes and PBA can effectively characterize the dynamical behavior of a cellular system, and the proposed ELD in this study can help understand systems with cell proliferation and death, beyond pure reactions. This dynamical analysis tool can be also applied to various similar biological systems.