This page is archived

Links to external sources may no longer work as intended. The content may not represent the latest thinking in this area or the Society’s current position on the topic.

Interdisciplinary approaches to dynamics in biology

24 - 25 February 2020 09:00 - 17:00

Theo Murphy international scientific meeting organised by Dr Ruben Perez-Carrasco and Dr Berta Verd.

This meeting brought together an interdisciplinary group of scientists interested in the timing of biological processes. Biology is dynamic at every scale, from transcription in a single cell to evolution. Increasingly, technological advances allow us to temporally quantify biological processes. The challenge we now face, which this meeting addressed, is to extract mechanistic understanding from the dynamics we measure.

Information on the speakers and programme is available below. Speaker abstracts are also available below. Recorded audio of the presentations is available below. Meeting papers will be published in a future issue of Interface Focus

Attending this event

This meeting has taken place. 

Enquiries: contact the Scientific Programmes team

Organisers

  • fbt

    Dr Berta Verd, University of Cambridge, UK

    Dr Verd is a somewhat unusual breed of biologist.   A mathematician by training, she spent a year studying an MSc in Sociology of Science at Kings College London before starting to work in biology during her MRes at Imperial College. She received her PhD in 2016 from the Centre for Genomic Regulation (CRG) in Barcelona, where she was introduced to the field of EvoDevo by her supervisor,  Dr Johannes Jaeger. During her PhD, she used data-driven mathematical modelling to study pattern formation during segment determination in Drosophila and other species of flies. In her work, she develops and employs mathematical tools to characterise gene expression dynamics, in the hopes that this will give us a better understanding of how gene regulatory networks drive developmental processes and shape their evolution. In October 2017, Dr Verd joined the Steventon Lab in the University of Cambridge as a Herchel-Smith Postdoctoral Fellow, where she studies the mechanisms underlying axial elongation in zebrafish embryos.

  • Dr Ruben Perez-Carrasco, Imperial College London, UK

    Ruben Perez-Carrasco is a lecturer in Theoretical Systems Biology in the Department of Life Sciences at Imperial College London. His research focuses on creating mathematical and computational models to understand how cells make decisions in autonomous or collective scenarios. In order to do so he brings together tools from dynamical systems, Bayesian inference, and stochastic calculus with the aim to generate predictive models able to synergize with experimental data. This has resulted in interdisciplinary collaborations such as the creation of synthetic gene regulatory circuits with specified functions or understanding tissue precision patterning in developmental biology. He held the competitive Clifford Fellowship for applied mathematics at UCL in 2017 and he has been the secretary for the Cell and Developmental Biology group of the Society for Mathematical Biology since 2020. Currently he leads the theoretical systems biology group at Imperial College, where in addition to research he is training the new generation of interdisciplinary scientists in mathematical and computational biology.

Schedule

09:20 - 09:55 Complex dynamics of ecological systems 

Many analyses of ecological systems are still based on the idea that natural (or managed) systems are essentially close to stable equilibria of corresponding simple ecological models. Moving away from this paradigm is key for understanding both natural and managed systems and designing management strategies. The presence of strong density dependent interactions, multiple time scales including changing environmental conditions, spatial interactions, and high dimension all contribute to making the inclusion of the potential for complex dynamics key for understanding ecological systems. Here complex dynamics includes all the dynamical features, including chaotic dynamics, complex basin boundaries, consequences of bifurcations, stochasticity, transient dynamics, as well as other aspects. Recent progress in the kinds of analyses that can be used will be reviewed. The results will be applied to various specific examples from a variety of ecological systems.

Professor Alan Hastings, University of California, Davis and Santa Fe Institute, USA

09:55 - 10:30 Dynamical models and comparative biology: homology and analogy of processes

This talk explores the role of dynamical models of developmental processes in comparative biology. Dynamical models of processes such as insect segmentation and vertebrate somitogenesis can explain how these processes might plausibly work, or how they actually work, depending on their level of mechanistic detail. As a consequence, these models can inform phylogenetic hypotheses about homology, parallelism, and convergence by determining the probabilities of transformation between character states. An interesting question concerns how dynamical processes themselves should be described and classified in order to function this way. Should they be classified (1) independently of phylogenetic considerations eg, in terms of ahistorical properties such as phase portrait geometries or, (2) in terms of phylogeny? The first option may be useful at certain stages of research, especially when combined with mechanistic research approaches. But ignoring phylogeny threatens to resurrect some of the classic problems with the pre-Darwinian programs of “rational taxonomy” or “idealistic morphology.” Pursuing the second option gives occasion to re-examine the unresolved conceptual issue of “process homology” from the point of view of dynamical modeling, and to assess its relation to the classical criteria of homology, such as topology, complexity, and congruence. The central question concerns how to productively combine formal and historical approaches to understanding developmental processes.

Dr James DiFrisco, Centre for Logic and Philosophy of Science, Leuven, Belgium

10:30 - 11:00 Coffee
11:00 - 11:35 Dynamical process in biology: on the role of mathematical modelling

Biology is fundamentally dynamic, and mathematical modelling plays an increasingly prominent role in understanding dynamical behaviours of living systems. What does a mathematical representation bring to the study of such systems? A quantitative description can enhance the clarity of assumptions made about the constitution of the system, and open up possibilities for more precise prediction. However, can the use of mathematical models do more than this? In addition to providing a precise deductive framework, mathematics also provides a language for abstract reasoning, with logical structures that do not necessarily correspond in any simple way to the physical elements of a system. Given the complexity of biological phenomena, tracing the route from system composition to system behaviour is challenging: the language of dynamical models provides a natural framework for exploring this dynamic process.

Professor Nick Monk, University of Sheffield, UK

11:35 - 12:10 Dissecting transcriptional dynamics in development one burst at a time

Over the last few decades in vitro and in situ approaches have revealed the identity of the molecular players driving transcription in eukaryotes. Yet, these studies are virtually silent on the precise timing of the recruitment of each of these players to the promoter, and on how this recruitment determines output transcriptional dynamics in vivo. Here, we present a new method for simultaneously measuring local input transcription factor concentration at target loci and the resulting output transcriptional activity of these loci in single living cells. Specifically, we study how the Dorsal activator, a key transcription factor in the development of the fruit fly Drosophila melanogaster, is recruited to the promoter of its target gene snail in order to drive transcriptional bursting. We found that transient surges in Dorsal concentration coincide with, but do not precede, the onset of transcriptional bursts. Interestingly, these surges are not maintained throughout the duration of the bursts and subside before the promoter transitioned back into a transcriptionally inactive state. Instead, we discovered that the amplitude of the transient Dorsal concentration surges at the start of transcriptional burst, and not surge duration, dictates transcriptional burst duration. We speculate that Dorsal delivers a “package” of downstream players to the promoter (eg, a cluster of RNA polymerase molecules) that sustains the transcriptional burst until this package is exhausted. Thus, our tool sets the stage for uncovering the precise timing and ordering of the diverse molecular players that drive the transcriptional process.

Dr Hernan Garcia, UC Berkeley, USA

13:10 - 13:45 Dynamics of neural patterning in Drosophila

The genetic and molecular substrates of many developmental processes have been identified, yet how spatial patterns dynamically emerge, and the relative contribution of positional cues and self-organization, often remain unclear. Neural patterning in Drosophila is commonly described as a two-step process, where prepattern factors drive localized proneural gene expression, then Notch-mediated lateral inhibition singles out neural precursors. In the dorsal thorax of the adult fly, small sensory bristles are arranged in five rows. A combination of genome engineering, imaging and mathematical modelling revealed that this ordered layout is organized by Notch, which establishes a series of proneural stripes that resolve into regular rows of sensory organ precursor cells. Patterning is initiated by a gradient of Delta ligand, and progresses through inhibitory signalling between and within the stripes. A simple mathematical model captures this temporal sequence, and suggests that the interplay between cell-intrinsic dynamics and signalling is essential for the gradual refinement of the pattern. It is anticipated that a similar logic could be at play in other contexts, with different initial or boundary conditions producing different arrangements of neural precursors. This possibility is explored in ongoing work on patterning in the fly eye, where a regular lattice of R8 photoreceptor cells emerges in the wake of a travelling differentiation front.

Dr Francis Corson, CNRS Ecole Normale Supérieure, France

13:45 - 14:20 The dynamics of spinal cord development

The generation of the correct cell types at the appropriate position and time is the first step in the assembly of functional tissues. One well-studied example is the development of the vertebrate spinal cord. In this tissue, distinct neuronal subtypes are generated in a precise spatial order from progenitor cells arrayed along the dorsal-ventral axis of the neural tube. Underpinning this organization is a complex network of extrinsic and intrinsic factors. Particularly well understood is the mechanism that determines the generation of different neuronal subtypes in ventral regions of the spinal cord. In this region of the nervous system, the secreted protein Sonic Hedgehog (Shh) acts in graded fashion to organize the pattern of neurogenesis. This is a dynamic process in which exposure to Shh generates progenitors with successively more ventral identities. At the same time tissue growth alters the arrangement of cells and the proportions of cell types and contributes to the elaboration of pattern. A gene regulatory network composed of transcription factors regulated by Shh signalling play an essential role in determining the response of cells. The dynamics of this network controls the pattern, pace, precision and proportion of the forming neural tube. Thus, accurate development of the neural tube and the specification of neuronal subtype identity relies on the interplay of cellular and molecular processes.

Dr James Briscoe FMedSci FRS, The Francis Crick Institute, UK

14:20 - 14:55 Oscillations and noise: partners in neural cell state transitions in vivo

In recent years, our understanding of how cells make cell state transitions has been transformed by discovery of short-time scale dynamics in gene expression, which can be oscillatory or noisy. Ultradian oscillations are exemplified by the expression of HES/Her transcription factors in neural progenitors but it is not known whether oscillations occur in vivo what is the role of noise. This talk will describe a quantitative and dynamic single cell level imaging approach to analyse ultradian oscillations and noise during cell state transitions during mouse and zebrafish neural development. Using reporter protein fusion knock-ins, complex HES5 dynamics are revealed in mouse neural tissue, with both short and long term trends. Contrary to expectations, HES5 expression becomes more frequently periodic as cells transition to differentiation. Coupled with an overall decline in HES5 expression this creates a transient period of oscillations with higher fold expression change, which presumably increases the decoding capacity of HES5 oscillations. Mathematical modelling suggests that noise has an oscillatory “priming” function because the HES5 oscillator operates close to its bifurcation boundary where stochastic conversions between dynamics are enabled. CRISPR/Cas9 mutagenesis of a miR-9 binding site in a knock-in Her6 reporter in zebrafish shows that changes in Her6 dynamics, including an increased high frequency noise has the opposite effect of locking cells in a progenitor state which is normally transitory. These findings show how oscillations operate during cell state transitions at the tissue level and reveal that noise optimization by miR-9 is necessary for transitions to occur. 

 

Professor Nancy Papalopulu, University of Manchester, UK

14:55 - 15:25 Tea
15:25 - 16:00 Unveiling the dynamics of commitment and fate choice in the developing nervous system

An ambitious goal of single-cell analyses is to describe dynamical biological processes and shed light on gene regulation. However, the destructive nature of transcriptomics measurements constitutes an immediate challenge to this purpose. “RNA velocity” is an analysis framework that helps to overcome this fundamental limit. It uses a single-cell transcriptomic snapshot measurement to estimate the time derivative of gene expression for a single cell. We applied the analysis framework to study different processes occurring during development, including differentiation, commitment, and cell cycle dynamics. In this talk, we will showcase the insights that the method is giving in the developing nervous system and the retina. In particular, our analysis captures details of differentiation dynamics in the Dentate Gyrus, cell cycle dynamics in retinal progenitors, and commitment biases. We discuss how to ensure that the estimation is not confounded by technical factors and on how to apply the framework to larger and more complex datasets.  

Dr Gioele La Manno, Swiss Federal Institute of Technology Lausanne (EPFL), Switzerland

09:00 - 09:35 Dynamic gene regulation at the single-cell level

Gene expression can be surprisingly dynamic and heterogeneous. This variability in gene expression, even in a clonal population of cells grown under the same condition, has been observed in diverse organisms, from mammalian stem cells to bacteria. It remains unclear how this variability is generated and what function it can serve. By using a combination of single cell time-lapse microscopy, mathematical modelling and synthetic biology techniques, the Locke group are attempting to understand how gene circuits generate dynamic gene expression, and how this dynamic information is transferred to downstream processes and other pathways. They work on simpler model systems such as cyanobacteria and B. subtilis, where it is possible to have precise single cell control of gene regulation, but have also extended this approach to the model plant Arabidopsis. In Arabidopsis, they are building on their work on simpler systems to examine the functional role of dynamic, and even stochastic, gene regulation in development.

Dr James Locke, Sainsbury Laboratory, University of Cambridge, UK

09:35 - 10:10 Our first choices: decoding signals during developmental transitions (lessons from quantitative biology)

During early development, extrinsic triggers prompt a collection of pluripotent cells in the blastocyst to begin the dramatic and long process of differentiation that gives rise to the tissues of the three germ layers (endoderm, mesoderm and ectoderm). Precise temporal control during these early fate-choices is paramount and impacts on the success of differentiation. Changes in morphology, gene expression signatures and epigenetic patterns and cell division cycles are believed to mark the point of no-return in fate choices. However, when and how cells irreversibly commit to differentiation is a fundamental, yet unanswered question. Poised to differentiate, embryonic stem (ES) cells are an invaluable model to address this question. Given appropriate differentiation cues, ES cells can recapitulate in vitro all the hallmark events that occur during differentiation and are our system of choice to understand fate decisions at the single cell level. During Silvia’s talk she will share two stories that illustrate how Silvia’s lab combines single cell imaging, genomic approaches and mathematical modeling to study how cells encode fate choices during early development.

Dr Silvia Santos, The Francis Crick Institute, UK

10:10 - 10:40 Coffee
10:40 - 11:15 Comparative analysis of multi-cellular dynamics and pattern emergence in the vertebrate tailbud

A major question in developmental biology is how cell fate decisions are coordinated precisely in space and time to generate emergent patterns of gene expression. The timing of cell transitions is regulated at multiple length scales in a highly integrated manner. Within individual cells, gene regulatory networks (GRNs) drive cell state transitions with an inherent timing. These are in turn patterned across cell populations by extracellular signals that diffuse or are transported between cells. This talk focuses on the multi-tissue level and considers how multicellular ensembles shift relative to each other to influence the timing at which competent cells receive signals. Such ‘tissue tectonics’ provide a causal relationship between processes at higher levels of biological organization (ie the relative positioning of signalling and responding tissues) to the control of processes at lower levels (ie cell biology and the activity of GRNs). As such, it is an example of downward causation and provides a multi-scale feedback mechanism to enable the self-organisation of developmental processes. The vertebrate tailbud offers a unique system to explore the relationship between multi-tissue morphogenesis and cell fate coordination in vivo. This will be discussed together with recent work showing how the inhibition of convergence and extension within multi-cellular aggregates of embryonic cells disrupt the shaping of BMP and Wnt/beta-catenin signalling gradients. This leads to alterations in the anterior-posterior patterning of neural marker expression, demonstrating a role for tissue tectonics in pattern formation.

Dr Ben Steventon, University of Cambridge, UK

11:15 - 11:50 On clocks and timers in development

Some biological oscillators function throughout the life of an organism, for example the circadian clock, whereas others have a more restricted duration, particularly in embryogenesis. The “segmentation clock” is a multi-cellular patterning system of genetic oscillators thought to control the rhythmic and sequential formation of the vertebrate embryo's body segments. Individual oscillating cells are synchronized with their neighbours, forming a coherent wave pattern of gene expression. How these wave patterns arise and how they are regulated during embryogenesis is not clear. Dr Oates will describe recent progress in understanding the behaviour of individual cells from the zebrafish as they slow their oscillations and differentiate during segmentation, and discuss how this gives rise to the tissue-level wave patterns. Central to this understanding is the concept of a timer that regulates the duration of a clock. This perspective reveals what part of the oscillatory cycle is changing as the cells slow and stop.

Dr Andy Oates, Swiss Federal Institute of Technology Lausanne, Switzerland

11:50 - 12:25 Human time vs mouse time with stem cell differentiation

Different species have different tempos of embryonic development: larger animals tend to grow more slowly than smaller animals. Dr Ebisuya's group has been trying to understand the molecular basis of interspecies differences in developmental time by using in vitro segmentation clock as a model system. The segmentation clock is the oscillatory gene expressions that regulate the timing of body segment formation from presomitic mesoderm (PSM) during embryogenesis. The group has recently succeeded in inducing PSM from both human iPS cells and mouse ES cells, recapitulating the oscillation and travelling wave of segmentation clock in vitro. Interestingly, the oscillation period of human segmentation clock was 5-6 hours while that of mouse was 2-3 hours. Taking advantage of our in vitro system, the group measured several biochemical reaction parameters of the core gene of the oscillation mechanism, Hes7, finding out that the degradation and production processes of Hes7 are 2-3 times slower in human PSM cells compared to mouse cells. The mathematical model quantitatively explained how the slower biochemical reactions in human cells give rise to the longer oscillation period in the human segmentation clock.

Dr Miki Ebisuya, EMBL Barcelona, Spain

13:25 - 14:00 Dynamics of segment patterning in arthropods

Segmentation in arthropods is established during development, often concurrently with posterior elongation. Much about this patterning process is still poorly characterised, but recent years have seen considerable progress towards grasping its basic regulatory structure and logic. Many of the genes that affect segmentation appear to be linked into small networks that form dynamical components such as timers, oscillators, and switches. Additional cross-regulation between these components creates a dynamical interplay that is fundamental to producing the segmental pattern. Several of these dynamical components appear to be strikingly similar to ones involved in patterning the nervous system, suggesting perhaps that they evolved within a more ancient developmental context and were only later recruited to a role in body segmentation. Alternatively, body segmentation might have originated from an ancestral neuromuscular patterning process which involved these components from the start. Simple models that capture the essential dynamical properties of the segmentation network show how linking developmental processes that create traveling waves (velocities) with others that measure time is sufficient to generate segmental patterns, even in the absence of instructive initial conditions. In addition, through adjusting the relative timescales of these components, a given system can produce a range of different output patterns, demonstrating how different body forms might easily evolve. On the other hand, the example of Drosophila shows that instructive positional gradients can also be incorporated into such a system, replacing a robust but relatively slow self-organising capacity with finely-tuned top-down control. Even in this extreme case, however, it is clear that dynamical processes still form a fundamental, ancestral, core.

Dr Erik Clark, Harvard Medical School, USA

14:00 - 14:35 Waves, flows and droplets: physical organisation of embryogenesis and regeneration

Embryonic development and tissue regeneration both require a high degree of spatiotemporal coordination. Dr Di Talia will describe my lab efforts to understand the synchronization of the cell cycle in early Drosophila embryos and the coordination of tissue growth in zebrafish bone regeneration. Dr Di Talia will show how synchronization of the cell cycle in Drosophila embryo is linked to precise nuclear positioning, which is in turn driven by cytoplasmic flows generated by actomyosin cortical contractions. Cortical contractions are regulated by oscillations in the activity of mitotic phosphatase PP1, thus forming a self-organized mechanism through which nuclei drive their positioning and synchronization. Dr Di Talia will also show how the cell cycle oscillator is interfaced with zygotic histone biogenesis by mediating the formation of the histone locus body by phase separation. Finally, Dr Di Talia will show that excitable waves of growth factor signalling drive cell and tissue growth in the osteoblast population controlling zebrafish scale regeneration, thus highlighting a novel principle of large-scale coordination for regeneration of large adult structures. 

Stefano Di Talia, a professor in the department of cell biology, poses in his lab

Dr Stefano Di Talia, Duke University Medical Center, USA

14:35 - 15:10 Mapping space and time in developing model systems

During development, sequential cell fate decisions generate diversity by carefully orchestrating their organisation across space and time. Often the dynamics of cell fate decisions directly relate to spatial location, such as in the case of somitogenesis where decisions are organised along the embryonic anterior-posterior axis. Advances in single-cell transcriptomics have enabled the characterisation of numerous cell states to be identified with high-resolution, yet these methods typically require cellular dissociation which disrupts such spatial information. In silico techniques, such as pseudotime approaches, often aim to ‘reintroduce’ dynamics by using information inherent in the data but make several assumptions. Instead, the group sought to identify spatial information inherent in such single-cell datasets using a parallel spatially-defined bulk RNA-sequencing and complementary single-cell data. The group developed a pipeline to recover spatial information at single-cell resolution, which allows them to reconstruct cellular arrangements that closely reflect the original organisation of source tissues. This method has potential implications for transcriptomic assays, where preserving spatial information in high-resolution data could allow the identification of novel spatial and dynamic trends along embryonic axes.  

Dr Naomi Moris, University of Cambridge, UK

15:10 - 15:40 Tea
15:40 - 16:15 Using avian models to understand developmental tempo

Embryonic development is a process of orchestrated progressive steps, which lead to the generation of patterned structures. While the sequence of developmental events is highly conserved, the speed at which they occur or “tempo” (“embryonic tempo” as a term was defined by Ebisuya and Briscoe, 2018) is a unique feature of a species and is highly variable within chordata and even between closely related species. Embryonic tempo is a mostly unexplored field presenting an opportunity to investigate and develop a fundamental area of biology with far reaching implications for many fields, including stem cell research, developmental biology, life-long health, ecology and food production. Variability in tempo, suggests that changes can be effected through few pleiotropic changes and there is evidence that the control of tempo is genetic. Dr Davey will describe the preliminary steps we are taking to define embryonic tempo and understand the genetic control of tempo using commercial duck species and chicken breeds and other avian species which demonstrate highly regulated embryonic tempo.

Dr Megan Davey, University of Edinburgh, UK

16:00 - 17:00 Closing remarks