It can be challenging to understand the dynamics of gene expression patterns when embryos develop inside a uterus, making it impossible to image changes to gene expression in real time. A new paper in Development presents a pipeline where images of mouse embryos at discrete developmental timepoints can be used to reconstruct a continuous description of gene expression changes over time. To learn more about the story behind the paper, we caught up with first author Laura Aviñó-Esteban and corresponding author James Sharpe, Head of EMBL Barcelona, Spain.
From left to right: Heura Cardona-Blaya, Laura Aviñó-Esteban and James Sharpe. Image courtesy of Carla Manzanas.
James, what questions are your lab trying to answer?
JS: The main question of the lab is how a complicated organ is self-sculpted out of tens of thousands of cells, with all the right tissues geometrically positioned in the right arrangement. We know that the full process is an emergent interaction between components at different scales: molecules, genes and cells, as well as the tissue geometry and mechanics. But understanding how these complex interactions lead to the correct result is very difficult. We believe the only way forwards is to integrate our knowledge within multi-scale computer models, and so our goal is to create the first full computer simulation of limb development.
Our goal is to create the first full computer simulation of limb development
Laura, how did you come to work in the lab and what drives your research today?
LA-E: When I decided I wanted to do a PhD I looked through the labs in the ‘local’ institute. I saw James' lab with pictures of the limb and it immediately clicked for me. Since my undergrad, I've been deeply impressed by the evolutionary development (evo-devo) of the limb – it's fascinating how the same basic organ, with some modifications, can enable such diverse functions as flying, swimming, running, jumping, or even typing this. Right now, I'm working to understand how the structure of the limb is actually formed during development.
Tell us about the background of the field that inspired your work
JS & LA-E: Limb development is a paradigm model system that has led to many classic concepts in development. We believe the field is now ripe for a more integrative, quantitative approach, in which mathematical models help us understand the true control mechanisms – complex feedbacks across scales, which are essential for the organ to develop. This leads to the need for more comprehensive datasets, such as the dynamic gene expression patterns described in our paper.
Can you give us the key results of the paper in a paragraph?
JS & LA-E: This paper introduces a computational method for reconstructing continuous 2D+time gene expression patterns in the developing mouse limb. Instead of interpolating full 2D patterns, we track small tissue regions over time and interpolate their gene expression using splines, creating smooth and biologically meaningful trajectories. The method was successfully applied to key developmental genes such as Sox9, Hand2 and Dusp6. And it allows us to do a precise spatiotemporal mapping, even at previously unobserved time points. We also successfully tested the method in the developing neural tube. Overall, we believe this technique is very valuable for studying gene expression dynamics in systems where real-time imaging is not feasible.
We believe this technique is very valuable for studying gene expression dynamics in systems where real-time imaging is not feasible
Schematics of the interpolation pipeline using Sox9. Top row: raw in situ hybridizations. Middle row: interpolated expression for a specific moving piece of tissue. Bottom row: final spatiotemporal reconstruction integrating interpolations across all tissue movements.
When doing the research, did you have any particular result or eureka moment that has stuck with you?
LA-E: Yes, one moment that stands out is when I realized I could use B-splines for interpolation. I had been struggling with a clunky interpolation where the gene expression intensity changes weren't smooth over time, and it just didn't feel right. Then, while playing a game, I noticed how the camera in the intro moved so smoothly along an arbitrary trajectory. That was the thing I needed! After some research into how game developers achieve this, I discovered B-splines, which turned out to be the perfect solution. It was definitely a eureka moment for me.
Why did you choose to submit this paper to Development?
JS & LA-E: We were thinking about the audience. The primary breakthrough here is a computational method, and we wanted to communicate this to as much of the relevant developmental biology community as possible. Development is the best place for this, as it strongly showcases tool development to our target audience.
What is next for you after this paper?
LA-E: Now that I have a smooth reconstruction of how the skeletal marker pattern of the limb changes over developmental time, my next step is figuring out how it actually happens! So, I'm currently focusing on the mathematical modeling of gene expression patterns over time and space.
Where will this story take your lab next?
JS: This will be a central tool for the lab. We are trying to understand the gene networks that control different patterning tasks within the limb. The gene expression patterns are so dynamic that it would be impossible to describe them without these great trajectories over time and space created by Laura's technique. It will help many projects.
Finally, let's move outside the lab – what do you like to do in your spare time?
LA-E: I draw a lot in my spare time, especially human and animal anatomy studies. There's something about the classical, old-school sketches that I find timeless. It's not just about understanding anatomy, it's about recognizing the art and complexity of life. Plus, it's incredibly relaxing for me.
JS: Travelling is probably my favorite pastime – seeing new landscapes and new cultures – I can never get enough!
L.A.-E. & J.S.: European Molecular Biology Laboratory (EMBL-Barcelona), Barcelona 08003, Spain; Barcelona Collaboratorium for Modelling and Predictive Biology, Barcelona 08005, Spain.
J.S.: Institució Catalana de Recerca I Estudis Avançats (ICREA), Barcelona 08010, Spain.
E-mail: [email protected]; [email protected]
