Despite being offered multiple other PhD positions in different fields, I joined this project because I wanted to understand resilience better. Why, after exposure to seemingly similar stressful experiences, do some individuals recover relatively well while others develop persistent behavioural changes?
What I did not anticipate was that, somewhere along the way, resilience would stop being merely the object of my study and that the project would put our own resilience to the test.
Studying resilience requires some resilience
Like most long-running research projects, experiments failed. Approaches had to be reconsidered. For example, at the time of our experiments, we were not aware that injection of a drug (tamoxifen), used to capture task-activated populations, could influence the deciding test—the social interaction test—used to classify resilience. Through trials and tribulations, we solved the problem, coincidentally supporting a scientific result published by another group some months later.
But that was just the beginning.
The multidisciplinary nature of the project brought its own challenges. Behavioural neuroscience, epigenetics, transcriptomics and bioinformatics do not always look at a biological question through the same lens. Managing different scientific perspectives, analytical approaches and collaborative disagreements became part of the work itself. Then came the COVID-19 lockdowns, adding another layer of uncertainty to experimental research.
Over time, I began to appreciate that resilience means adapting the path while retaining the question that made you start walking in the first place. Resilience was not about being at 100% every day, but about retaining the goal even on days when I was functioning at 10%—like a string being pulled back before the arrow can be shot forward.
Molecular snapshots of resilience: From one layer to another
Scientifically, the project developed into two complementary studies. First, we investigated transcriptional differences in neurons activated during stress recall. Next, the project extended the question into the epigenome, integrating DNA methylation with transcriptional information to ask whether resilience and susceptibility were associated with distinguishable molecular states in these recall-activated cells and whether the two molecular layers were altered simultaneously.
What emerged was not simply a picture of “more” or “less” stress biology.
Resilient and susceptible animals showed distinct molecular patterns. In particular, integrating the molecular layers pointed us towards pathways involving synaptic organisation and GTPase signalling linking synapses to the nucleus with differential guanine nucleotide exchange dynamics, suggesting possible routes through which experiences could be embedded in the nucleus for longer-lasting modulation of neuronal function and, ultimately, behaviour. Here, I answered part of the question I had been curious about.
Unexpectedly, however, looking across two molecular layers also revealed results suggesting that molecular states following stress could change dynamically. Find more here. https://www.nature.com/articles/s41380-026-03834-7
These findings point towards viewing resilience and susceptibility not as fixed destinations, but as biological trajectories unfolding across time. For future studies seeking therapeutics against stress susceptibility and depression, this could mean challenging the concept of “one size fits all” and introducing timing- or molecular-state-dependent treatment strategies.
Although our findings were strong and the work was an adjoining study to our first published paper, the path towards its online publication today came with a difficult challenge.
The Gate of Reviewer 2
As happens in many scientific stories, Reviewer 2 eventually became one of the main protagonists.
In the icy winter of December 2025, after years of experiments, analysis, integration and writing—and 18 months of revision—the manuscript was rejected and, figuratively, dumped in the bin. One can imagine the impact this had on the first author as well as the co-authors. A process that had seemed straightforward went through unexpected twists and turns.
For a while, the blow was tremendous, until a trusted senior co-author encouraged me to find the strength to appeal to the editor.
So, I returned to the reviews—not simply asking how we could argue against the points raised by Reviewer 2, but where misunderstandings might have arisen and where our own presentation could have contributed to them. We wrote to the editor around the end of January 2026, explaining the scientific reasoning and addressing the points we believed deserved reconsideration.
We waited anxiously, with bated breath.
The appeal was accepted after a month-long wait. We were back in the game, with more revisions at play. Ultimately, Reviewer 2 was satisfied, and we could progress to the next round: publication.
In retrospect, there is something wonderfully appropriate about a paper on resilience having such difficulty reaching the finish line.
Resilience is not remaining unchanged: What can future studies do?
Resilience is sometimes described as “bouncing back”, as though a resilient system simply returns to exactly where it started. But biological systems rarely experience stress without being changed by it. The more interesting question may be how they adapt—and why those adaptations lead different individuals towards different trajectories.
Towards the end of this project, I became more convinced that resilience research needs greater investment, particularly in longitudinal studies. Single snapshots can tell us which molecular states distinguish individuals at a particular moment. They cannot fully tell us how those states arose, whether they persist, when they reverse, or when an apparently adaptive response becomes maladaptive. Following individuals across time—and integrating molecular, behavioural and environmental information—may bring us closer to understanding resilience as a process rather than a label.
I began this project wanting to understand what makes an individual resilient to stress and, somewhat unwillingly, was made to fortify my own resilience through the journey of the project. Understanding its relevance, I started a science communication platform to explore how we might build our capacity to cope with stress—our resilience—through basic daily routines, food habits and maintaining healthy sleep schedules. These ideas are communicated through a small, fluffy, pink mascot: Stresspunsch, who explains scientific findings about stress and resilience through understandable stories.
Thanks to all the co-authors for exhibiting their own resilience throughout this journey, especially Dewi, Beat and Hristo, without whom the paper could not see publication.