Not all TGF-betas are the same.
More than five years after her death, developmental biologist Kathryn Anderson's final study has emerged from Memorial Sloan Kettering Cancer Center, carried to completion by colleagues who refused to let it dissolve into grief. The work reveals how WNT, a molecular signal present at the very dawn of life, guides embryonic cells through layered decisions — not by issuing a single command, but by integrating spatial cues across a developing body to determine what each cell will ultimately become. In tracing this choreography of identity, the research also illuminates how cancer hijacks the same ancient machinery to spread and destroy. It is a reminder that science, like memory, can outlast the person who began it.
- When Kathryn Anderson died in 2020, her lab's unfinished research on embryonic cell identity risked disappearing entirely — a loss that colleagues refused to accept.
- The central tension in the science itself mirrors this fragility: embryonic cells begin in a state of pure potential, and the question of what forces them to commit to a single identity has long resisted clean answers.
- The team discovered that WNT does not act as a simple on-off switch but operates in sequential layers, reading spatial gradients of BMP and NODAL to steer each cell toward its final form.
- The same molecular process — epithelial-to-mesenchymal transition — that gives embryonic cells the freedom to move and specialize is co-opted by cancer cells to break away and metastasize, accounting for roughly nine in ten cancer deaths.
- Recognizing that BMP and NODAL, though related members of the TGF-beta family, push cells in opposite directions opens a potential path toward more precise therapies targeting metastasis.
- The paper finally reached publication through a patchwork of shared effort across multiple labs and institutions — a collective act of scientific and personal devotion to a colleague's unfinished work.
More than five years after Kathryn Anderson's death in November 2020, her laboratory at Memorial Sloan Kettering Cancer Center has published what may be her final scientific contribution. Appearing in Developmental Cell, the work traces how a single molecular signal called WNT orchestrates one of biology's most fundamental moments: when embryonic cells abandon infinite potential and commit to becoming something specific.
Anderson had spent her career studying early mammalian development, eventually chairing MSK's Developmental Biology Program and earning a reputation for landmark discoveries at every stage of her work. When she fell ill during the pandemic, some projects were abandoned. This one was not.
About a decade ago, postdoctoral researcher Rocio Hernández-Martínez created mouse embryos missing two genes — Axin1 and Axin2 — that normally keep WNT from becoming overactive. Without them, WNT locked permanently on, producing embryos that could form only a limited range of tissues, missing the cells that give rise to the heart, head, and front-facing structures. Hernández-Martínez sensed the project held something important and made it her own.
What the team ultimately found was that WNT works in layers. It first nudges cells out of pure plasticity, then integrates two distinct spatial signals — BMP and NODAL — that run in opposite directions across the developing embryo like a gradient. Cells reading high BMP are directed toward the back of the body; cells reading NODAL toward the front. Each cell combines its location on this spectrum with the WNT signal to determine its final identity. Single-cell sequencing allowed the researchers to map these decisions with remarkable precision.
The finding carries an unexpected implication for cancer. BMP and NODAL belong to the same molecular family — TGF-beta — yet push cells in opposite directions. This matters because the same process the study examines in the embryo, called epithelial-to-mesenchymal transition, is what allows cancer cells to break away and spread. As many as nine in ten cancer deaths are attributed to metastasis, and TGF-beta is known to drive it. The new research suggests that treating TGF-beta as a single signal may be too simple — understanding which member is active, and how it interacts with WNT, could point toward new ways to slow cancer's spread.
After Anderson's death, Hernández-Martínez joined the lab of Anna-Katerina Hadjantonakis, who succeeded Anderson as program chair, and spent two more years advancing the project before moving to UC San Francisco. Without dedicated funding, the work became a shared responsibility — squeezed in around other commitments by Hadjantonakis' team members Sonja Nowotschin and Ying-Yi Kuo, with genomics support from Bertie Göttgens' lab at the Cambridge Stem Cell Institute. For Hadjantonakis, seeing it through was both scientific and personal — a way of honoring a dear friend and respecting the fundamental questions Anderson had been asking. The science, she felt, was too beautiful to abandon.
More than five years after Kathryn Anderson's death in November 2020, her laboratory at Memorial Sloan Kettering Cancer Center has published what may be her final scientific contribution. The work, appearing in Developmental Cell, traces how a single molecular signal called WNT orchestrates one of biology's most fundamental processes: the moment when embryonic cells abandon their infinite potential and commit to becoming something specific—a heart cell, a bone cell, the tissue of a head.
Anderson had spent her career studying early mammalian development, the period when a cluster of highly flexible cells must somehow know what to become and where to go. She joined MSK in 1996 and eventually chaired the Developmental Biology Program, earning a reputation as a scientist who made landmark discoveries at every stage of her work. When she fell ill during the pandemic, the lab's research scattered. Some projects were abandoned. This one was not, though it took years of determined effort by colleagues who refused to let it disappear.
The story of the science itself begins about a decade ago, when a postdoctoral researcher named Rocio Hernández-Martínez was working in Anderson's lab. Using genetic techniques, the team created mouse embryos missing two related genes—Axin1 and Axin2. Normally, these genes act as a dimmer switch, keeping WNT from becoming too active. Without them, WNT gets locked in the "on" position. The result was striking: embryos that could form only a limited range of tissues, missing the cells that would give rise to the heart, head, and front-facing structures. Hernández-Martínez adopted the project as her own, sensing it held something important.
What the team eventually discovered was that WNT doesn't simply flip a switch. It works in layers. First, it nudges cells out of their initial state—a condition of pure plasticity where they could theoretically become anything—and starts them down a path toward becoming muscles, bones, organs, or connective tissue. But the specific identity a cell acquires depends on a second layer of instruction, which also involves WNT. The researchers used single-cell sequencing to read gene activity in individual cells and map how WNT shapes the earliest decisions about cell fate. What emerged was a picture of remarkable precision: WNT integrates two distinct molecular signals—BMP and NODAL—across a spatial landscape. These signals run in opposite directions across the developing embryo like a gradient. Cells that encounter high levels of BMP are directed toward the back of the body. Cells that encounter NODAL are directed toward the front. Each cell reads its location on this spectrum like a point on a topographic map and combines that information with the WNT signal to determine what it will become. BMP plus WNT yields one outcome; NODAL plus WNT yields another.
The finding carries an unexpected implication: BMP and NODAL belong to the same family of signaling molecules, called TGF-beta. They are related, yet they push cells in opposite directions. This distinction matters for cancer research. For a tumor to spread, cancer cells must do something normal cells cannot—break away and travel. The molecular program that enables this movement is called epithelial-to-mesenchymal transition, or EMT, and it is the same process the study examines in the embryo. In a healthy embryo, this freedom to move is by design. In cancer, it marks the final and most aggressive stage of disease. As many as nine in ten cancer deaths are attributed to metastasis. TGF-beta is known to drive EMT in cancer, but the new study suggests that treating it as a single signal may be too simple. Understanding which TGF-beta is active, and how it interacts with other signals, could point to new ways to slow or stop cancer's spread.
After Anderson's death, her remaining lab members transitioned to other labs across the institution. Hernández-Martínez joined Anna-Katerina Hadjantonakis, who succeeded Anderson as chair of the program, and spent two more years advancing the WNT project before taking a position at the University of California, San Francisco. Without dedicated funding and with Hernández-Martínez moving on, the project became a shared responsibility—squeezed in around other work. Two members of Hadjantonakis' lab, senior research scientist Sonja Nowotschin and senior research assistant Ying-Yi Kuo, stepped in to conduct further experiments. A collaboration with Bertie Göttgens' lab at the Cambridge Stem Cell Institute provided essential genomics expertise. It was everyone's side project, but they were determined to see it through. For Hadjantonakis, shepherding the work to publication was both a scientific commitment and a personal one—a way of honoring a dear friend and colleague, and of respecting that Anderson had been pursuing fundamental questions about how WNT directs cell identity as a baseline for understanding its role in other contexts, including disease. The science story was too beautiful to abandon.
Notable Quotes
Anderson was an extraordinary scientist who made landmark discoveries at every stage of her career.— Cell journal obituary
It was everyone's side project, but we were determined to see it through.— Anna-Katerina Hadjantonakis
The Hearth Conversation Another angle on the story
Why does it matter that WNT works in layers rather than all at once?
Because it means the signal isn't just a yes-or-no switch. A cell needs to be pushed out of its flexible state first, then it needs to read its location and integrate that information. If you only understand the first part, you miss how cells actually make decisions.
And the BMP-NODAL discovery—why is that surprising to researchers?
Because they're cousins in the same molecular family, yet they point cells in completely opposite directions. It means you can't just say "TGF-beta is active" and predict what happens. You have to know which one. That's the kind of distinction that changes how you think about a problem.
How does this connect to cancer spreading?
Cancer cells need to break free and move, just like embryonic cells do during development. But in cancer, that movement is chaotic and destructive. If we understand the subtleties of how these signals work in a healthy embryo, we might find ways to interfere with that same process in tumors.
What does it say about the lab that they finished this work after Anderson died?
It says something about scientific culture, I think. This wasn't a grant-funded priority. It was a side project. But they kept going because the science was good and because it mattered to honor her. That kind of persistence doesn't always make it into the papers.
Did Anderson know the work would be published?
No. She fell ill during the pandemic, and they didn't realize how sick she was because everyone was working remotely. By the time they learned she'd passed, the project was unfinished. They had to choose whether to let it go or carry it forward.