Quantitative, real-time, single cell analysis in tissue reveals expression dynamics of neurogenesis

Cerys S Manning, Veronica Biga, James Boyd, Jochen Kursawe, Bodvar Ymisson, David G Spiller, Christopher M Sanderson, Tobias Galla, Magnus Rattray, Nancy Papalopulu

Preprint posted on July 20, 2018

Article now published in Nature Communications at

Cell state transitions in real time: stochastic and oscillatory dynamics of neural progenitors revealed through live cell imaging of Hes5

Selected by Teresa Rayon


In the developing spinal cord, neural progenitors convert into neurons and glia in multiple waves of differentiation. These cell fate transitions are controlled by the dynamic changes in gene expression in response to signaling

But how do cell fate choices occur in single cells? Studying cell fate transitions at single-cell resolution is complex, since gene expression varies from cell to cell and measurements in single cells are noisy. In addition, the expression of some of the key transcription factors involved in cell fate transitions fluctuates over time. The best example of this is the segmentation clock, that rhythmically and sequentially subdivides the elongating axis, such that the period defines the length of the forming somites. Oscillations have only been recently discovered in neural tissues, by performing time-lapse imaging of bioluminescence reporters and quantification of brain slices [1]. In this preprint, the authors study the dynamic expression of the Notch target gene Hes5 at single cell resolution over time in the neural tube. By analyzing the behavior of hundreds of cells with ad hoc statistical tools and through computational modeling, Manning et al. show that early progenitors are noisy and primed to enter a transient oscillatory phase as the cells differentiate.


Why I chose the paper:

Live-cell imaging through fluorescent reporters has taught us that protein dynamics of transcription factors is quite heterogeneous within populations of cells. However, what this means for cell fate choices in single cells is still unresolved. In this preprint, the Papalopulu lab reports how the dynamic expression of HES5 influences cell fate transitions over time in its real context in single cells. Their quantitative approach is really exciting for a number of reasons:

  1. To avoid any masking of single-cell heterogeneity by non-quantitative methods, the authors perform absolute quantification of HES5 and find a high degree of variability in expression (ranging from 26nM to 319nM). The challenge of providing absolute numbers of venus::HES5 expression allows them to define cell-to-cell heterogeneity in terms of long-term (sustained) and short-term dynamics (fluctuations).
  2. Their statistical analysis on hundreds of tracked cells in movies enables the authors to assess the dynamics of HES5 expression in different positions in the neural tube, and to study the role of oscillations in cell fate transitions unbiasedly. The authors confirm that the levels of expression in proliferating progenitors fluctuate. Unexpectedly, oscillations were more commonly found in the progenitors transitioning to differentiate. This finding challenges the current view that oscillations are crucial to maintain the progenitor state. Instead, Manning et al. propose that oscillations coupled to a general decrease in levels allow downstream genes to interpret their response accordingly, since oscillations increase a higher fold-change in expression levels (see gifs below).

    A clever way to decode what normally is a very shallow oscillator. According to Manning et al., a ball bouncing down steps (oscillatory expression) undergoes greater height drops than a ball rolling down a ramp (aperiodic expression).
  3. Finally, thanks to their stochastic model of genetic auto-repression, they propose how HES5 expression dynamics can be generated: noisy expression of HES5 in the progenitors can trigger oscillations if the intrinsic stochastic noise operates in the region where small parameter changes can cause a transition between aperiodic and oscillatory.

How this work moves the field forward:

Cells of the same type show heterogenous levels in gene expression, and there are various reasons as to why this would be: from real noisy expression, to cells being in a different state or a limitation in the detection technique. To understand the role of fluctuations in gene expression, it is the key to study single cell behavior in real-time.  Furthermore, mathematical models are key to interpret complex dynamic gene expression patterns, since models give explanations as to how the identified patterns could be originated, and permit a full exploration of parameters. In this preprint, Papalopulu’s lab couples state-of-the-art quantitative techniques -such as fluorescence correlation spectroscopy, to stochastic and deterministic mathematical models. This allows them to carefully characterize gene expression dynamics in neural tube progenitors. Overall, this preprint advances our understanding on how noisy dynamics in gene expression can account for the establishment of periodic oscillations that might be then relevant for cell fate choice.

References and further reading:   

  1. Imayoshi, I. et al. Oscillatory control of factors determining multipotency and fate in mouse neural progenitors. Science 342, 1203–8 (2013).

Questions to the authors:

  1. Do the authors think that the oscillatory mechanism identified in progenitors transitioning to differentiate might be a common feature of cell fate choices?
  2. If we could look at more than one of the transcription factors involved in the fate choice, do the authors think that other HES or bHLH factors that have been described to oscillate will fluctuate coherently in the progenitors, or would they only synchronize in the differentiation zone away from the ventricle?
  3. The authors do not rule out an effect of cell cycle on HES5 fluctuations. From their data, it seems very tempting to find a direct link between the two since dividing progenitors show the highest proportion of noisy cells. Have the authors considered validating this finding combining their venus:HES5 reporter with cell cycle sensors (Fucci)?






Tags: computational modeling, live cell imaging, notch, quantitative, spinal cord

Posted on: 9th August 2018

Read preprint (No Ratings Yet)

  • Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here

    preLists in the cell biology category: