Interview with Hervé Turlier

" Biology on which equations can be laid down is physics "

In 2021, the Paoletti Prize, awarded to talented young researchers in the life sciences, went to Hervé Turlier. Interview with this CNRS and Collège de France research fellow, who has distinguished himself through interdisciplinary, but above all fundamental, research at the interface between the physics and biology of embryonic development.

What was your first contact with physics?

As a child, I did a lot of building and electronics activities. I loved taking things apart around the house to try and understand how they worked... but I couldn't put them back together again!

Then, in my final year of high school, I did an exciting internship with a startup specializing in liquid crystal displays. Liquid crystals, like polymers, foams and even certain biological structures, are part of what we call soft matter. Their states are neither solid nor liquid, and therefore difficult to classify. This field was initiated by Pierre-Gilles de Gennes, who, before winning the Nobel Prize, was a professor at the Collège de France.

I then followed a very traditional curriculum: preparatory classes and the École Polytechnique, with a focus on physics and mechanics. There, I again studied soft matter in practical work, and my final year was devoted entirely to the physics of liquids and soft matter. During this period, I unexpectedly met Prof. Jean-François Joanny, whose research focused on cell biophysics at the Institut Curie. It was so obvious, I said to myself: "This is the person I want to do my thesis with". He himself had done his doctorate with Pierre-Gilles de Gennes: a combination of circumstances that reveals a certain continuity in my career path since high school.

You completed your thesis in Soft Matter and Biophysics in 2013..

The theme was cell division in animal cells, and more specifically cytokinesis - the way in which a cell physically deforms and separates during cell division. Cell division is characterized by the separation of chromosomes from the mother cell into two future daughter cells, known as mitosis. On the other hand, these two daughter cells have to be generated, which means cutting the cytoplasm in two: this is cytokinesis.

This PhD was based on mechanical studies and numerical simulation. The latter required computer programming skills, and I was lucky enough to be trained by a competent researcher in the field, as I'd never coded before.

After that, I wanted to immerse myself more in biology, as my colleagues were primarily physicists. I went on to do a postdoctorate at the European Molecular Biology Laboratory (EMBL) in Germany. There, I worked between two research teams, in embryonic development biology and digital simulation. My role was again to carry out simulations, this time focusing on a slightly larger scale, of the order of a few cells.

As this theme appealed to me enormously, in 2017 I submitted a research program to the Collège de France on early embryogenesis, and this program was accepted.

What is early embryogenesis?

It's about how embryos start to form from a fertilized egg cell and up to a hundred embryonic cells. It's a fairly old subject, as biology historically began with the observation of embryos in marine biology stations. Today, this is less studied, as most researchers are interested in living organisms at much more advanced and complex stages.

In my team, made up of physicists, mathematicians and mechanical and computer engineers, we're going back to basics. We collaborate with developmental biologists - from the Institut Curie and the Observatoire Océanologique de Villefranche-sur-Mer in particular - who provide us with real microscopy and gene expression data from early embryos.

How do you use this data?

To determine how these cells self-organize, i.e., in the sense of physics, how structure will emerge as a result of interactions between the elements that make it up. In this case, the structure is the embryo and the elements are the cells.

Cells need to know which way to divide, which tissue to become, and so on. They can only evolve because they sense their environment, i.e. the other cells around them. It is their interactions, together with the expression of each cell's own genes, that enable self-organization.

Our aim is to reproduce these phenomena in numerical models. This involves working out our equations together on a whiteboard, then programming on the computer in C++ or Python.

How do you translate this physical and digital approach?

The idea is to find out what ingredients are necessary and sufficient to reproduce a particular behavior of the early embryo.

The number of species studied remains fairly limited for various reasons, essentially linked to whether or not the embryos can be seen under the microscope. Only transparent embryos can be imaged. This creates a bias. We only use data from animal species, such as mice, starfish, sea urchins and the small worm Caenorhabditis elegans.

Our starting point is that an embryonic cell performs the same functions whatever its species, and whatever its genetic information. All cells are capable of dividing, polarizing, deforming, adhering to their neighbors... And we imagine that it is possible to model these functions generically by varying very few physical parameters. Numerical simulations help us to specify cell geometry and mechanics, as well as communications and genetic and molecular regulatory networks.

Eventually, the aim will be to launch "Virtual Embryo", a virtual platform for digital modeling of animal embryos, in 3D and as a function of time, which can be used by the scientific community and later become collaborative.

You recently published a paper in Sciencemagazine..

With our partners at the Institut Curie, we analyzed the last stage in the development of the early mouse embryo. At this stage, a cavity is formed inside the embryo by an influx of water from the surrounding environment. What has been observed experimentally is that, in reality, thousands of cavities are created at every point of contact between embryonic cells. These contacts are then broken up to form the large cavity.

We proposed a simple physical model: each of the tiny cavities is subject to surface tension, like a soap bubble. The fluid in a bubble is under pressure: the slightest contact allows it to escape. The smaller the bubble, the greater the pressure inside.

In a bubble bath, the bubbles get bigger and bigger over time, as the smaller ones disappear, emptying their contents into the larger ones. It's the same for the cavities between the cells of the embryo: as soon as they're close enough, they exchange their fluid, and knowing that they'll always have a disparity in size, the small ones will empty into the big ones. In the end, the only solution is a single cavity. In soft matter physics, this phenomenon is known as "Ostwald ripening".

How does your work differ from that of your peers?

My research is extremely interdisciplinary. It goes way beyond physics, and some physicists even think I'm doing biology! But as far as I'm concerned, quantitative biology - that is, biology on which we can place equations - is physics. However, the objects we observe are very different from classical physics. There are no great unified principles, and this calls for innovative methods. Mechanics, applied mathematics, computer science and even machine learning are all tools we use to analyze our data. It's not easy to cross fields that are sometimes a little hermetic, but I find it absolutely fascinating.

How would you sum up your experience in the academic world?

One of the major difficulties today, and it's not just me who has to deal with it, is finding a permanent position in France. The Collège de France has shown confidence in me by allowing me to set up my own team. I benefit from enviable working conditions, the presence of colleagues of a high scientific level, and access to a wide range of multidisciplinary lectures.

I'm extremely privileged, but I know that people more deserving than me don't necessarily have the same opportunities. And I worry about the next generation... I do my best to support young people in their scientific projects, and I can't help wondering about their future. Even for the best of them.

But watching them progress and mature gives me great pleasure. The moments I appreciate the most are discussions with a student: sitting down in front of the blackboard, taking the time to think together about a scientific question... Finally, I think the first thing that motivates me is to continue learning, and above all learning from others. Science is all about sharing.

---------

Hervé Turlier is a CNRS research fellow and head of the Multi-scale Physics of Morphogenesis team at the Collège de France's Interdisciplinary Biology Research Center.

In addition to prestigious funding from the ATIP-Avenir program, the Bettencourt-Schueller Foundation and the European Research Council, he has been awarded the Paoletti 2021 prize. This award recognizes the quality of his work and his team's ambitious research project, which aims to understand the principles of cell organization and communication in an embryo at the very earliest stages of development.

Interview by Océane Alouda