The people behind the papers – Stanley Strawbridge Free

Stanley Strawbridge with new plants purchased with his wife.

Stanley, what questions are your lab trying to answer?

At the broadest level, my lab is trying to answer the question: why do early pregnancies fail? We know that a significant proportion of human pregnancies are lost at very early stages, often before the pregnancy is even known about. While genetic abnormalities, such as aneuploidy, account for some of these failures, many embryos with a normal genome still fail to implant or develop properly. My lab focuses on understanding the cellular and molecular events that occur during these earliest stages of development: how cells make decisions, how tissues self-organise and how these processes can go wrong.

How did you come to work in the lab and what drives your research today?

Like many people, I have had a nonlinear career path. I started out in analytical toxicology in Ben Capacio's group at the US Army Medical Research Institute of Chemical Defence (MRICD) in Maryland, USA, where I grew up. I began there as a 17-year-old summer intern, returned for two more summers and then joined full-time as a research assistant straight out of undergraduate study. I also spent a summer at the Plum Island Animal Disease Centre, USA, working on molecular virology, as part of a Department of Homeland Security scholarship that I received as an undergraduate at West Virginia University, USA. At that point, I was firmly on the fast track for a career in the defence sector.

However, while double majoring in Mathematics and Biochemistry, I took a graduate-level course in mathematical systems biology with Adam Halasz (West Virginia University, USA). It was the first time that I saw how mathematics could be used to explore living systems in a meaningful, mechanistic way. I had always known that I wanted to use mathematics in experimental research, but that course showed me how.

Eventually, I moved to the UK and completed a Master's degree in Mathematical Modelling and Scientific Computing at the University of Oxford, where I met my now co-author Alexander Fletcher, and subsequently joined Austin Smith's group at the Cambridge Stem Cell Institute, UK, for my PhD. There, I met Jenny Nichols (now at the University of Edinburgh, UK), who introduced me to the remarkable complexity of the embryo itself.

What drives my research today is the same sense of awe and curiosity I felt back then: how a single cell gives rise to a complete organism, how fate decisions are made and coordinated across space and time, and – more recently – why these processes sometimes go wrong. I'm motivated by the belief that combining quantitative approaches with experimental biology can not only deepen our understanding of fundamental developmental processes, but also help address real clinical challenges.

Can you tell us about the background of the field that inspired your work?

A striking phenomenon in mouse development is that if you inject enough embryonic stem cells (ESCs) into an 8-cell-stage embryo, while leaving the zona pellucida intact to physically constrain the system, you can generate a mouse entirely derived from the donor ESCs. Interestingly, this complete donor contribution doesn't occur in morula aggregations, in which the zona has been removed and the embryo can freely expand. I am interested in understanding the driving mechanisms behind this takeover of the embryonic compartment by the donor cells.

Can you give us the key results of the paper in a paragraph?

In this study, we investigated how ESCs influence early lineage decisions in the developing mouse embryo. By injecting embryos with donor ESCs, we found that donor cells can displace host cells into the extra-embryonic tissues. During the first cell-fate decision, host cells are physically pushed into the TE, the outer layer that forms the placenta. This displacement occurs even when donor cells lack FGF4, a signalling molecule known to drive primitive endoderm (PrE) formation. However, when donor cells express FGF4, they also bias the remaining host inner cell mass cells toward a PrE fate. This suggests a sequential mechanism: donor cells first outcompete host cells through physical interactions and then influence their fate through signalling. We tested this hypothesis using mathematical models, which showed that a two-step process involving crowding and FGF4 signalling best explains and, importantly, predicts the data. Together, our results reveal how ESCs can reshape lineage allocation in early embryos through a combination of mechanical and molecular mechanisms.

When doing the research, did you have any particular result or eureka moment that has stuck with you?

Surprisingly, our big ‘eureka’ moment came during the revision stage of the paper. We had built a mathematical model to describe how cells in the embryo make decisions over time, and, initially, we used a type of simulation that captures the average behaviour, where you take the average of embryonic behaviours from a population. However, a reviewer suggested that we also investigate whether the model could capture the variability observed among individual embryos, which is a much harder question. To address it, we ran a different type of simulation, one that preserved the same model interactions, but generated many individual embryos, each with slight differences, to reflect biological variation. When the results came out, they didn't just look good – they looked suspiciously good. So, I asked my co-author Alex to double-check my code, while I did the same. I was elated to find that my analysis was indeed correct and reproducible.

We had expected this outcome in theory, but seeing it play out so well in practice was something else. It reminded me how mathematics, at times, can feel like magic!

And what about the flipside: any moments of frustration or despair?

Accurate 3D nuclear segmentation was a constant challenge. At the time of this work, we didn't have reliable deep-learning tools to handle it. So, we had to either manually count individual nuclei, which is not scalable, or use traditional image-processing methods such as watershed segmentation, which still required a lot of manual correction.

This burden was shared by our brilliant co-author, and the first research student I had the privilege of supervising, Anna Schrattel, who brought remarkable skill and perseverance to both the experimental and computational sides of the project.

Thankfully, deep learning has since caught up. Tools such as Cellpose, which use convolutional neural networks, have made a huge difference. A new version of Cellpose was released last month (May 2025) that incorporates transformer-based architectures, some of the most powerful models in modern deep learning today. I'm really looking forward to using it in future work.

Why did you choose to submit this paper to Development?

Development was the clear home for this manuscript. The journal has played a defining role in the history of developmental biology, publishing many of the landmark studies that have shaped the field. It remains one of the most respected and trusted journals in our community, valued not only for the quality of its science but also for its commitment to supporting researchers and advancing the discipline. I was fortunate to receive a The Company of Biologists Travelling Fellowship, awarded through Development, which supported a visit to my collaborator, co-author and now colleague, Professor Alexander Fletcher (University of Sheffield, UK), to develop the modelling component of this work. I'm genuinely thrilled to see our work published in Development. It feels like a real career milestone, and I'm proud to contribute to a journal that has had such a profound and lasting impact on the field.

Where will this story take your lab next?

With this model in hand, my focus now shifts to the human system. Although much of what we've learned from mouse models is highly relevant to human development, comparable high-resolution, lineage-resolved time-series datasets for humans are still lacking. This includes datasets like those used in this study, generated by Nestor Saiz during his time in Anna-Katerina Hadjantonakis' lab (Sloan Kettering Institute, USA). These were instrumental in building accurate models of mouse development. In contrast, modelling efforts in human, particularly those aimed at predicting successful embryo development or pregnancy, have largely relied on brightfield time-lapse videos from embryoscopes. While these have been fruitful in correlating morphological features with reproductive outcomes, they do not explain why an embryo fails to progress, specifically whether it is due to aneuploidy, failed morphogenesis or incorrect lineage specification.

The next step for my lab is to generate precisely the kind of lineage-resolved, snapshot time-series data needed to build predictive models of human development. By comparing embryos that develop normally with those that do not, we aim to understand not only how things go right, but also when and why they go wrong. This work already forms an active branch of my research programme, in which we are making progress using published, but incomplete datasets. Crucially, published datasets often exclude lower-quality embryos, which limits our ability to model developmental failure. By building toward generating our own lineage-resolved datasets, we aim to fill this gap and ultimately provide clinically relevant insights to improve reproductive outcomes.

Finally, let's move outside the lab – what do you like to do in your spare time?

Public engagement is something I care deeply about. Over the years, I've organised Pint of Science events in both Oxford and Cambridge, and I co-founded a spin-out initiative called Creative Reactions, which pairs scientists with local artists to create works inspired by the research. I've also performed science-inspired stand-up comedy at events such as Science Showoff. These activities are not just fun, they are revealing. They show you what you don't know about your own subject and challenge you to communicate clearly and accessibly. After all, biologists and mathematicians might talk about solutions to the same problem, but they often do so in different languages.

That said, I do have a life beyond science. Music has always been a big part of it, everything from orchestra and chorus to an 8-year stint in a marching band. Since moving to the UK, I've focused on the double bass. I have played with the Academy of Great St Mary's in Cambridge from 2014, and, more recently, I joined the Sheffield Philharmonic Orchestra. Interestingly, both orchestras bring together a mix of local musicians and academics, which has made them especially welcoming and rewarding communities to be part of.

I also enjoy cooking, which pairs well with another hobby – gardening. My wife, Fiona, and I keep a mix of ornamental and edible gardens, complete with fruit trees, vegetable beds, and even a few egg-laying hens, dutifully watched over by our cat, Alfie. My office is a miniature jungle too, home to about two dozen plants of all shapes and sizes. Finally, I've loved rediscovering the hills since moving to Sheffield.