Modeling Gene Expression Noise: From Transcriptional Dynamics to Cell Growth and Division

summary

The gist

Gene expression is inherently noisy, leading to substantial cell-to-cell variation in mRNA and protein levels The paper reviews models of stochastic gene expression in growing and dividing cells…

In short

Gene expression is noisy, causing cell-to-cell variation in mRNA and protein levels. Standard static models like the telegraph model fail because they ignore a cell's life history, including growth and division. The paper advocates for mechanistic models that couple gene expression dynamics with the cell cycle to accurately capture biological noise.

Key concepts

Telegraph Model
This is a simple two-state model used to describe transcription noise. It assumes genes switch between two states: active or inactive. While useful for snapshot data, it ignores more complex promoter behaviors and the cell's dynamic state.
Cell Cycle Dynamics
This involves incorporating the cell's life cycle—birth, DNA replication, growth, and division—into gene expression models. This is crucial because growth and division introduce significant noise through random molecule partitioning and changes in transcription parameters.
Cell Size Control Strategies
These are different ways cells maintain stable mRNA levels during growth. The paper identifies three main strategies: the 'sizer' (dividing at a size threshold), the 'timer' (dividing after a fixed time), and the 'adder' (dividing after adding a fixed volume).
Mechanistic Models
These are advanced models that link gene expression directly to cell cycle progression. By coupling gene expression with cell cycle models, researchers can build more realistic simulations that account for how cellular history affects molecular fluctuations.

Terminology used across episodes

This episode discusses

The paper

Modeling Gene Expression Noise: From Transcriptional Dynamics to Cell Growth and Division · Read on arXiv

Augustinas Sukys, Lucy Ham, ¡Kaan Öcal, James Holehouse, Ramon Grima

School of BioSciences, University of Melbourne · School of Maths & Stats, University of Melbourne · Department of Biology, Washington University in St. Louis · Laufer Center for Physical and Quantitative Biology, Stony Brook University · Department of Applied Mathematics & Statistics, Stony Brook University

Gene expression is inherently noisy, leading to substantial cell-to-cell variation in mRNA and protein levels. The telegraph model and the simpler bursty model provide good fits to single-cell count distributions and have been used extensively to understand transcriptional kinetics and estimate the kinetic parameters. However, these models all describe a static cell. Here, we review models of stochastic gene expression in growing and dividing cells which aim to understand how a cell's own life history, from its birth, through copying its genome, to its division, leaves a distinct fluctuation signature on molecule numbers. We discuss the biochemical and biophysical processes captured by these models, their analytical tractability, the concept of cyclo-stationary behavior, when this is well approximated by the steady-state of static models and how the distributions of molecular numbers measured under these conditions sensitively depend on the experimental setup. This is followed by a discussion of how these models are fitted to single-cell measurements, how biological variability can be separated from measurement noise, and which computational advances make this feasible for thousands of genes. We finish by highlighting the open problems in the field.

Transcript

Introduction to the show: ident: Genomics Radio. Generated commentary on the latest computational biology and genomics papers.

Ines: We're looking at the paper: "Modeling Gene Expression Noise".

Marcus: Here's the gist of the paper. Gene expression is inherently noisy, leading to substantial cell-to-cell variation in mRNA and protein levels The paper reviews models of stochastic gene expression in growing and dividing cells which aim to understand how a cell’s own life history, from its birth, through copying its genome, to its division, leaves a distinct fluctuation signature on molecule numbers Limitations of Static Models The standard stochastic model of transcription is the telegraph, or two-state, model Both the telegraph model and the bursty model predict steady-state distributions of mRNA numbers that often fit snapshot data The simplicity of these models comes at the cost of neglecting key biological detail Transcription may involve multiple active or inactive promoter states, particularly in eukaryotes Furthermore, they assume the cell is static, i.e.

Ines: they exclude a cell-cycle description involving birth, DNA replication, growth and division A good statistical fit of the steady-state solution of these models to single-cell data does not guarantee biophysically realistic predictions or meaningful kinetic parameters Incorporating Cell Cycle Dynamics Cell growth and division add several sources of biological noise Cell growth and division add several sources of biological noise: random partitioning of molecules at division; the doubling of gene copy number upon DNA replication, partially offset by dosage compensation; dependence of transcription parameters on cell size and cell-cycle stage; variability in interdivision times and growth rates Mechanistic models of gene expression include coupling gene expression to a model of the cell cycle, for example a multi-state Markov model of cell-cycle progression These can be analyzed along a single lineage or, when population snapshots are of interest, incorporated into a branching process tracking all descendants Modeling Cell Size Control and Noise Sources Cell size control is commonly classified into three phenomenological strategies

2611.06537: : the sizer, in which division occurs once a threshold size is reached; a timer, in which it occurs a fixed time after birth; and an adder, in which it occurs once a fixed volume is added since birth Cell size-dependent transcription is thought to be an important feature of eukaryotic transcription, as it helps to maintain stable mRNA concentrations during growth The three strategies can be distinguished, assuming exponential growth, by the slope of added size over the cycle against birth size Inference and Future Directions Most inference from single-cell data has thus far relied on conventional models which assume the existence of a steady-state The accuracy of parameter estimation using these simpler models implicitly relies on how well their steady-state can approximate the age-averaged distribution of the more complex cell-cycle models The gold standard for overcoming these limitations is to fit mechanistic models of gene expression that explicitly incorporate cell growth and division to cell-cycle-resolved single-cell data Beyond inference, the scope of current models needs to widen, including multi-state promoters, regulatory interactions, nuclear processing and protein dynamics Exact likelihoods for cell-cycle models remain difficult to derive and computationally expensive to evaluate Spatial transcriptomics now enables the measurement of expression variability at the subcellular and tissue levels, but mechanistic models that combine stochastic gene expression with intracellular transport, cell-cell coupling and growth and division remain largely undeveloped The telegraph model and its bursty limit remain the most widely used descriptions of stochastic gene expression A good fit, however, does not mean that the inferred kinetic parameters are correct or even physically meaningful, as both models ignore the cellular life cycle To understand cellular heterogeneity and noise in gene expression, we need models that couple gene expression to growth and division These models allow us to determine when the static cell description of the telegraph and bursty models provides a good approximation to the underlying biology Fitting these models to cell-cycle resolved single-cell data is a promising avenue to understand gene regulation at a genome-wide level References The authors acknowledge funding through the ARC Centre of Excellence for the Mathematical Analysis of Cellular Systems L.H.

Marcus: acknowledges funding through the ARC Centre of Excellence for the Mathematical Analysis of Cellular Systems K.Ö and A.S acknowledge funding through an Australian Research Council Laureate Fellowship (FL220100005) J.H.

Ines: would like to acknowledge the support of the National Science Foundation Grant Award Number EF-2133863 R.

Paper summary: Ines: So, looking at this paper "Modeling Gene Expression Noise: From Transcriptional Dynamics to Cell Growth and Division," they conclude that we need models that couple gene expression dynamics with cell growth and division.

Marcus: That means moving away from just the telegraph or bursty models when trying to understand molecular noise because those static descriptions miss the dynamic process of life.

Ines: They suggest that fitting these mechanistic models to single-cell data is a promising avenue for understanding gene regulation across the genome at a genome-wide level.

Yuki: For us, it means that to truly understand cellular heterogeneity, we have to look at the entire history of the cell, not just one moment in time.

Ines: The paper essentially shows that when you incorporate growth and division into the model, you can determine when those simpler static descriptions are actually a good approximation for what's happening biologically.

Conclusion: Ines: So, to wrap up this part of our talk on gene expression noise, these authors are looking at how cell growth and division add noise to the whole picture.

Marcus: Yeah, they're taking those standard models—the telegraph model and its bursty limit—which we always use for mRNA numbers in a single cell, but they’re saying that snapshot data isn't enough.

Ines: Exactly. They argue that if you only look at the steady state, you miss the whole life history of the cell from birth to division.

Yuki: From my perspective, it's about understanding how a cell’s very existence—growing bigger and splitting into two—creates a unique signature in those molecule counts.

Marcus: It sounds like they are pushing toward coupling gene expression models directly to a model of the cell cycle progression, maybe using something like a multi-state Markov model for that.

Ines: Right. And the core idea is that this isn't just adding extra noise; it's about figuring out when our simple static descriptions actually work and when we need this more complex coupling to get right kinetic parameters.

Yuki: It means we stop assuming a cell is static, and start modeling the birth, DNA replication, growth, and division as active parts of the expression dynamics.

Marcus: The implication for us in the cohort analysis is that if we use these more mechanistic models, our inference on transcription rates or burst timing might be much more accurate for real biology.

Ines: It definitely puts a new benchmark out there: fitting those coupled models to cell cycle resolved data would be the gold standard for understanding genome-wide regulation.

Yuki: It opens up a whole new avenue for population genetics because it lets us see how lineage history influences the noise we measure across different cells.

Marcus: So, while fitting these full mechanistic models is hard computationally, it seems like the next logical step to bridge the gap between simple statistical fits and actual biological reality.

More episodes

← Home