Huang Lab Overview
“The breadth and depth of complexity of living organisms require that we combine ‘entirety of analysis’ (-omics approaches) with ‘analysis of entirety’ (complex systems theory) – two distinct realms life science endeavor that are under one roof at the ISB”
–Sui Huang, MD, PhD, Professor
Visit Faculty Profile Page
Multicellular organisms comprise a defined set of phenotypically distinct cells, such as brain cells, liver cells, skin cells, blood cells etc., each of which contains the same genome with the very same set of genes. This diversity of stable phenotypes, all produced by one genome, poses an epistemic challenge to the biologists’ habit of seeking a direct mapping from one gene into one phenotype. Genes do not act solo but in concert.
The Huang Lab is interested in the fundamental principles that govern how the gene regulatory network (GRN) orchestrate the activities of the genes to produce the variety of stable cellular states, such as the known cell types of the body, which in turn undergo state transitions, such as the differentiation of a multi-potent stem cell to a specialized blood cell. The theoretical framework that guides research in the laboratory is that these stable cell states are “attractors” in the rugged epigenetic landscape that epitomizes the global non-linear dynamics of the gene regulatory network.
But much as genes, the cells also change their states not independently of each other. Cells communicate with each other to coordinate state transitions, thereby ensuring the “correct” composition of the various cell types in tissues and organs. Thus, cell population dynamics must also be considered in multicellular development.
Disruption of the normal coordination between the genes and between the cells can result in trapping of cells in a self-perpetuating mechanism that leads to cancer. Thus, an overarching theme of the research in Huang’s lab is the intricate interface between normal cell development and cancer development – seen through the new lenses of the formal concepts of gene networks and cell population dynamics.
The Huang group attacks the cancer problem by seeking to understand the very phenomenon of multi-cellularity in health and disease because cancer is the price we pay for having evolved a complex multicellular system. The goal is to arrive at a new level of understanding of cancer using the new powerful formal tools of the epigenetic landscape and by considering cell population dynamics. Specifically, one overarching hypothesis in the Huang laboratory is that cancer cells are cells trapped in a pathological attractor state of the network. Using state-of-the art single cell resolution measurement of gene expression patterns coupled with monitoring their phenotypic consequences, the Huang lab examines the generic properties of spontaneous cell state diversification and of regulated cell fate switching in multipotent cells and cancer stem cells. Such experiments unveil the non-genetic (“epigenetic”) aspects of tumor progression hat have been neglected.
The experiments, performed on cancer cells and stem cells in culture, are either driven by formal hypotheses rooted in theories of complex dynamical systems and statistical physics or are guided by specific biological questions. Yet they leave room for explorative discovery-driven science that is still a central pillar of systems biology. Thus, members of the Huang laboratory have backgrounds ranging from theoretical physics to molecular and systems biology and benefit from the technology developed or improved by other groups at the ISB.
One rarely articulated problem that pertains to the fundamental issue of genotype-to-phenotype mapping occupies much of the thinking in the Huang laboratory: The ubiquitous “wiring diagrams” that display which biomolecules in cells regulate which one’s activity are commonly used as a mental image of the web of causal relationships to explain a phenotype in terms of gene activities. Where is the problem with this picture? In using such network maps as explanatory tool one forgets that a regulatory network is a physical entity associated with a single cell and that the body contains of zillions of cells. Thus, when reading such network diagrams to explain a phenotype manifest at the level of the entire organism one undertakes a mental, subliminal multiplication, tacitly assuming that the “whole is the sum of the parts” (rather than ” more than the sum …” – as Aristotle taught us). In reality, as the Huang group and many others have noted, every single cell, even within a population of apparently identical (clonal) cells is in a distinct cell state with respect to its gene expression pattern – often for no apparent reasons. Such stochastic, non-genetic heterogeneity of cells adds another layer of complexity to how the genotype becomes a phenotype. Clearly, things don’t just “add up.” Instead, the inevitable heterogeneity of cell populations is important: It drives cell type diversification in multi-cellular organisms and may play a central role in the origin of cancer.
Experiments: Quantitative characterization of the dynamics of non-genetic variability of gene expression patterns between individual cells in progenitor and cancer stem cell populations combined with mathematical modeling. The goal is to obtain a detailed “statistical mechanics” description of the fluctuations in gene expression patterns that lead to the spontaneous heterogeneization of isogenic cell populations and of the controlled cell state transitions in order to understand how gene regulatory networks channel the fluctuations to produce the observed cell population behavior. The laboratory also applies such analysis to leukemic and breast cancer cells undergoing a state transition into a state of increased malignancy and drug resistance.
Theory and Modeling: Development of a formal framework for the notion of a quasi-potential surface that captures the global dynamics of gene regulatory circuits that produce multiple stable states (globally consistent “relative depth” of attractors). The goal is to use gene network information and single-cell gene expression data to build the true “epigenetic landscape” that affords prediction of developmental paths of cells and of the accessibility to nearby attractor states (“reprogrammability”) as cells change their gene expression patterns during differentiation and adaptation to external stress.
Bioinformatics: Analysis of existing genome and transcriptome data in the light of the concept of a global epigenetic landscape. The goal is to better understand how discrete gene sequence differences map into gradual quantitative differences of a phenotype that may manifest, so goes the hypothesis, as a distortion of the epigenetic landscape leading a shift of attractor states.