Keeping up with the ′omics: non-equilibrium models of gene regulation
© Pincus; licensee BioMed Central. 2015
Published: 10 February 2015
Non-equilibrium processes are vital features of biological systems. Despite this universally accepted fact, gene regulation is typically formalized into models that assume thermodynamic equilibrium. As experimental evidence expands the repertoire of non-equilibrium genome regulatory mechanisms, theoreticians are challenged to devise general approaches to accommodate and suggest functions for non-equilibrium processes. Ahsendorf et al. provide one such framework, which is discussed in the context of the growing complexity of eukaryotic gene regulation.
Life is an excursion away from equilibrium. From the concentration gradients cells establish both with their environments and within themselves, to chaperone-assisted protein folding, energy-consuming cellular processes serve to fend off equilibrium. Such non-equilibrium processes are essential for life: as Ahsendorf et al. state in their paper in the current issue of BMC Biology, ‘we are only at equilibrium when we are dead’ .
The fundamental hallmark of an equilibrium process is reversibility. Biosynthetic processes, however, rely on a combination of reversible, passive steps and irreversible, energy-consuming steps. These irreversible steps allow biological systems to attain arbitrarily high degrees of fidelity, and as such are vital for information processing and propagation . In particular, each step of gene expression relies on non-equilibrium proofreading mechanisms to ensure fidelity: transcription, splicing and translation all rely on both thermodynamic equilibrium steps and irreversible kinetic steps. For example, the initial binding of a tRNA anti-codon to the codon in the decoding center of the ribosome is an equilibrium event, while the GTP hydrolysis-dependent proofreading is a non-equilibrium process required for fidelity .
There are three main reasons we tend to model gene regulation with an equilibrium framework. The first is historical: the original computational model of gene regulation, which focused on the transcriptional switch at the heart of the lytic/lysogenic fate decision of phage λ, treated transcription factors binding to and dissociating from naked DNA - an equilibrium process - as the function that determined transcriptional output . The equilibrium formalism was perfectly suited to this case of gene regulation and captured the dynamics of the process so well that it has been extended to other gene expression circuits that are far beyond its intended reach.
The second reason is that, despite the complexity of eukaryotic transcription, the equilibrium approximation works quite well. In many circumstances, within the precision of experimental capabilities, the equilibrium model can fit the data. This of course begs the question of whether it is necessary to bother accounting for non-equilibrium processes at all. Moreover, while we can list dozens of non-equilibrium processes that relate to gene regulation, we frequently have no idea which events are important to make explicit in a given situation, what the function of a particular energy-consuming event is or what the cost of neglecting a reaction might be. Thus, in the name of parsimony and because the equilibrium approximation gives results that are good enough to match the level of resolution to which we have experimental data, we ignore the non-equilibrium processes.
The third reason we still rely on equilibrium models is that we lack a straightforward theoretical framework to model non-equilibrium complexity. Without such a framework, we risk relegating the role of models to recapitulation. That is to say, as equilibrium approximations, the chief function of gene regulation models is to test if our assumptions for how a circuit is operating are sufficient to explain the experimental results we observe. If, on the other hand, we would like to use models to gain insight into the functions of aspects of gene regulation that we do not fully understand, such as post-translational modification of transcription factors, then we need methods that accommodate such processes.
To demonstrate the utility of this framework, the authors apply their approach to three case studies of gene regulatory models in published literature to gain insight into the effects of non-equilibrium processes. In the first application, to transcription driven by nuclear hormone receptors, the authors show that, despite the known role of non-equilibrium ubiquitination in hormone receptor regulation, the non-equilibrium effects remain hidden and first order equilibrium models suffice. Next, the authors use their framework to cast doubt on the stability of inherently bounded chromatin domains that was suggested by numerical simulation - any stable chromatin mark that can be propagated will eventually lead to unbounded spreading unless non-equilibrium dissipative mechanisms are involved. In this case, the framework allowed first principle-based conclusions for the effects of non-equilibrium processes to be drawn without numerical simulation. Finally, by applying their framework to the regulation of the yeast PHO5 promoter, the authors show clear effects of non-equilibrium processes in the complexities of the gene regulation function that have yet to be understood molecularly or revealed experimentally.
As increasingly sophisticated experimental techniques continue to reveal novel complexities involved in eukaryotic gene regulation, theoretical formalisms that allow us to model gene expression need to keep pace. Techniques such as ChIP-seq have revealed non-random localization of histone modifications, massive complexes of transcription factors termed ‘super-enhancers’ and poised/bivalent developmentally regulated genes [5,6]. More recently, experimental techniques enabled by high-throughput sequencing technology that map the three-dimensional architecture of the genome, such as Hi-C and ChIA-PET, have revealed long-range interactions and cohesin-bounded megabase-scale chromatin domains . Needless to say, these structures and organized domains are established and maintained through non-equilibrium processes. The graph-based framework and operations described by Ahsendorf et al. provide a theoretical toolkit to formally accommodate such processes, and will hopefully enable us not only to more realistically model gene regulation, but also to gain insight into the currently opaque functions of non-equilibrium processes involved in chromatin organization and regulation.
- Ahsendorf T, Wong F, Eils R, Gunawardena J. A framework for modelling gene regulation which accomodates non-equilibrium mechanisms. BMC Biol. 2014;12:102.PubMed CentralPubMedView ArticleGoogle Scholar
- Hopfield JJ. Kinetic proofreading: a new mechanism for reducing errors in biosynthetic processes requiring high specificity. Proc Natl Acad Sci U S A. 1974;71:4135–9.PubMed CentralPubMedView ArticleGoogle Scholar
- Blanchard SC, Gonzalez RL, Kim HD, Chu S, Puglisi JD. tRNA selection and kinetic proofreading in translation. Nat Struct Mol Biol. 2004;11:1008–14.PubMedView ArticleGoogle Scholar
- Ackers GK, Johnson AD, Shea MA. Quantitative model for gene regulation by lambda phage repressor. Proc Natl Acad Sci U S A. 1982;79:1129–33.PubMed CentralPubMedView ArticleGoogle Scholar
- Whyte WA, Orlando DA, Hnisz D, Abraham BJ, Lin CY, Kagey MH, et al. Master transcription factors and mediator establish super-enhancers at key cell identity genes. Cell. 2013;153:307–19.PubMed CentralPubMedView ArticleGoogle Scholar
- Lesch BJ, Dokshin GA, Young RA, McCarrey JR, Page DC. A set of genes critical to development is epigenetically poised in mouse germ cells from fetal stages through completion of meiosis. Proc Natl Acad Sci U S A. 2013;110:16061–6.PubMed CentralPubMedView ArticleGoogle Scholar
- Dowen JM, Fan ZP, Hnisz D, Ren G, Abraham BJ, Zhang LN, et al. Control of cell identity genes occurs in insulated neighborhoods in mammalian chromosomes. Cell. 2014;159:374–87.PubMedView ArticleGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.