From bench to code: collaboration unlocks insights into Alzheimer’s disease
How a chance encounter at a scientific conference resulted in a fruitful cross-Atlantic research collaboration
In 2023, EMBL-EBI Postdoctoral Fellow Iguaracy Pinheiro de Sousa presented his bioinformatics work on how cells communicate within the heart, as part of the Network Biology meeting in Cold Spring Harbour, New York.
“I met Yifei Cai during the opening keynote by Dana Pe’er; we happened to sit next to each other and immediately connected over how exciting the talk was,” said Pinheiro de Sousa. “At the time, she was a postdoc at the Yale School of Medicine, working on understanding how communication between cells in the nervous system happens – and how it breaks down in Alzheimer’s disease. During the poster session, we had a great discussion about how I could adapt my approach, originally developed for the heart, to get new insights into her Alzheimer’s brain datasets.”
This was the beginning of an exciting collaboration that brought together neuroscientists at Yale with bioinformaticians at EMBL-EBI. Together, they developed new methods for studying this devastating disease, uncovered new insights, and identified a promising therapeutic target.
Communication breakdown in the brain
At Yale, Cai and her colleagues were trying to understand how cellular swelling interferes with neurons’ ability to communicate, particularly in Alzheimer’s disease patients.
Neurons use axons – thin fibres that connect them to other neurons – to communicate. Sometimes, abnormal clumps of misfolded proteins can accumulate between neurons. These are called amyloid plaques, and for decades, they have been considered a promising therapeutic target for Alzheimer’s disease. Despite this, anti-amyloid therapies haven’t made headway in treating the condition.
The lab that Cai was working in was exploring something called axonal spheroids – a prevalent hallmark in Alzheimer’s disease. Axonal spheroids are bubble-like structures on axons; they form due to swelling induced by amyloid plaques. Cai wanted to integrate publicly available single-cell data with the lab’s own subcellular proteomics data from axonal spheroids to gain new insights. To do so, she needed additional bioinformatics expertise, which Iguaracy and EMBL-EBI were able to provide.
Bridging the wet and dry lab divide
The two groups started collaborating remotely, and shortly after, Cai was awarded a Corporate Partnership Programme (CPP) visitor fellowship through EMBL’s Scientific Visitor Programme. This is designed to enable PhDs and postdocs from other organisations to benefit from new collaborations and technologies at EMBL. This meant Cai could spend a month at EMBL-EBI, working closely with Pinheiro de Sousa in Evangelia Petsalaki’s research group, and sharpen her bioinformatics skills in the process. Together, they developed a robust data analysis pipeline that Cai now uses to analyse her data.
Their results showed that axonal spheroids represent a potentially important pathological process. “We believe that targeting these spheroids could be an important future avenue for treating Alzheimer’s disease by improving the overall electrical conduction and brain circuits, rather than just simply removing the amyloid plaques,” said Jaime Grutzendler, Physician-Scientist at Yale School of Medicine Department of Neurology, and senior author of the paper outlining this work, which was published in the journal Nature Aging.
This was only the beginning. Continuing the collaboration, Cai also worked with EMBL-EBI colleagues to explore the molecular interplay between axons and the myelin sheath, which is the fatty insulating layer that wraps around axons, much like the rubber coating on electrical wires. In a second study, they integrated subcellular proteomics data from the axon-myelin interface with published single-nucleus RNA sequencing data to analyse cell-to-cell communication.
Their goal was to investigate how this critical interface is disrupted in Alzheimer’s disease. The study revealed that the proteins present between axons and their myelin sheaths differ between individuals with Alzheimer’s and those without. These disruptions in molecular crosstalk were linked to impaired function in both axons and their insulating myelin. The findings were published in the journalNature Neuroscience.
A multidisciplinary approach
“Working with EMBL-EBI has been a very fruitful collaboration,” said Cai. “The opportunity to visit for a month, work with their bioinformatics experts on my specific research questions and datasets, and to immerse myself in the environment was a great learning experience. As a scientist, you can’t know everything, so working with colleagues who have complementary skills is vital.”
“Our lab is computational, but we work very closely with colleagues in experimental settings,” explained Evangelia Petsalaki, Group Leader at EMBL-EBI. “Our aim is to understand what drives cell responses in different health conditions. This requires analysing large and complex datasets, which often include multi-omics data. Just like building a bridge requires architects, structural and civil engineers, urban planners and builders, understanding disease also requires a wide range of skills and expertise. This is a great example of how basic medical research and bioinformatics complement each other to push the boundaries of our knowledge, and to help us understand a disease that affects so many.”
Read more about the EMBL Scientific Visitor Programme and the Corporate Partnership Programme visitor fellowship.
Source article(s)
Subcellular proteomics and iPSC modeling uncover reversible mechanisms of axonal pathology in Alzheimer’s disease
Cai, Y. et al
Nature Aging 10 March 2025
10.1038/s43587-025-00823-3
Myelin–axon interface vulnerability in Alzheimer’s disease revealed by subcellular proteomics and imaging of human and mouse brain
Cai, Y. et al
Nature Neuroscience 13 June 2025
10.1038/s41593-025-01973-8
