Clonal memory in human embryonic stem cells biases fate potential during endoderm differentiation
By integrating lineage tracing with single-cell multi-omics, this study reveals that heritable chromatin accessibility states in pluripotent human embryonic stem cells, rather than extrinsic signals alone, drive clonal memory that biases fate decisions and generates heterogeneity during directed endoderm differentiation.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Life begins with a single cell that holds the potential to become any part of the human body, from a beating heart to a thinking brain. As this cell divides, its descendants must make a series of critical choices, deciding whether to become skin, blood, or nerve tissue. Scientists have long understood that these decisions are guided by chemical signals from the cell's environment, much like a teacher giving instructions to a classroom. However, a persistent mystery has remained: why do cells that receive the exact same instructions sometimes choose different paths? Even when grown in a perfectly controlled dish with identical nutrients and chemical cues, populations of stem cells often produce a messy mix of outcomes, with some cells turning into the desired tissue and others becoming the wrong type entirely. This unpredictability has been a major hurdle for researchers trying to grow specific tissues for medical treatments, as it suggests that something inside the cells themselves, rather than just the outside environment, is influencing their destiny.
A team of researchers has now uncovered a hidden layer of memory within human embryonic stem cells that helps explain this variability. By tracking individual cells and their families over time, they discovered that cells carry a heritable record of their past that biases their future choices, even before they receive any signals to change. This phenomenon, which the researchers call clonal memory, means that cells related by birth tend to make the same fate decisions, not because they are reacting differently to the outside world, but because they started with a different internal landscape. The study reveals that this bias is written in the way the cell's DNA is packaged and accessed, a physical state that can be passed down through many cell divisions, long before the cells are asked to specialize.
To investigate this, the scientists used a method that allowed them to tag individual stem cells with a unique genetic barcode, essentially giving each cell and its future descendants a distinct name. They started with a small group of human embryonic stem cells, added these barcodes, and then let the cells grow and divide for two weeks. During this time, the cells multiplied into millions, and the original barcodes were copied into every new cell, creating large families of related cells. The researchers then split these families into separate culture dishes and guided them to become definitive endoderm, a specific type of tissue that eventually forms the gut and liver. By sampling the cells at different stages of this process, they could trace which family each cell came from and see what it had become.
The results showed that cells from the same family consistently made the same choices. If a particular family tended to become the correct gut tissue, almost all its members did so. If another family tended to become the wrong type of tissue, such as muscle, its members followed suit. This pattern held true even when the families were separated into different dishes and grown in isolation, proving that the tendency was not caused by the local environment or random chance. The researchers found that these fate-biased families were already present in the starting population, long before the cells were told to change.
What was most surprising was that these different families looked identical when the scientists examined their active genes. At the stage where the cells were still stem cells, there was no detectable difference in the genes being turned on or off between the families that would later succeed and those that would fail. This ruled out the idea that the bias was caused by a visible difference in the cells' current instructions. Instead, the answer lay in the cell's epigenome, which is the system that controls how tightly the DNA is wrapped and which parts are accessible for reading. The researchers found that the families destined to become the wrong tissue type already had their DNA wrapped in a way that made it easier to access the genes for muscle development. In contrast, the families that successfully became gut tissue did not have this pre-existing access to muscle genes.
This pre-accessibility acted like a hidden predisposition. When the cells received the chemical signals to become gut tissue, the families with the open muscle genes were hypersensitive to the signals and accidentally turned on the muscle program instead. The study identified that this bias was particularly strong for genes involved in a specific signaling pathway known to drive muscle formation. The researchers found that the families prone to making mistakes had more open spots near these muscle-related genes, making them more likely to respond to the growth signals by becoming muscle rather than gut. This suggests that the "noise" or inconsistency often seen in stem cell experiments is not just random error, but a reflection of a real, inherited diversity in how different cell families are primed to respond.
The findings challenge the view that all stem cells in a dish are functionally identical until they receive a signal. Instead, the study suggests that even in a culture that looks uniform, there are distinct subgroups of cells with different internal histories and different potentials. This has important implications for how scientists grow tissues for medicine. If the starting culture contains a mix of these pre-biased families, the final product will inevitably be a mix of the right and wrong cell types, no matter how carefully the chemical signals are tuned. The researchers propose that to create pure, reliable tissues for therapy, scientists may need to find ways to reset these internal memories or select only the families that are truly ready to become the desired tissue.
The study also highlights the power of looking at cells as individuals with lineages rather than just as a collective mass. By tracking the families, the researchers could see patterns that would have been invisible if they had only looked at the average behavior of the whole group. They found that while the cells passed through a stage where they looked like they were preparing to become gut tissue, the families that would later fail had already begun to drift toward a muscle fate at the level of their DNA structure. This drift happened silently, without changing the active genes enough to be seen by standard tests, but it was enough to steer the cells down the wrong path once the differentiation process began.
Ultimately, this work provides a concrete explanation for why stem cell differentiation is often imperfect. It shows that the source of this variability is not a failure of the external instructions, but a feature of the cells themselves. The cells carry a memory of their past divisions in the form of their chromatin structure, a physical state that influences how they interpret the world around them. While this memory can lead to errors in a laboratory setting, the researchers note that in a developing embryo, such diversity might actually be useful, allowing different groups of cells to respond to gradients of signals and form complex patterns. In the controlled environment of a lab dish, however, this same mechanism creates the heterogeneity that researchers have struggled to eliminate. The study does not offer a simple fix, but it does offer a clear target: understanding and managing the epigenetic landscape of the starting cells may be the key to unlocking more reliable and consistent results in regenerative medicine.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.