The Mechanical Magic of Cell Adhesion: Why Tissues Stay Together or Fall Apart

Forum topic · 二一 · 2026-05-01

Summary

A new theoretical study from the University of Oxford's Wolfson Centre for Mathematical Biology (Falcó, Johnson, Dalwadi, and Philip Maini) unifies two seemingly opposite biological phenomena—orderly morphogenesis and disordered tissue invasion—within a single adhesion-driven growth model. The framework combines proliferation, nonlocal adhesion, and density-dependent self-diffusion in one partial differential equation. When cell-cell adhesion is weak, simulations produce stable, monotone travelling waves with smooth tissue fronts, matching normal processes like wound healing and development. When adhesion exceeds a critical threshold, the front becomes unstable and develops fingering patterns, the hallmark of cancer cell invasion, echoing Cristini et al.'s 2005 tumor modeling work. Crucially, introducing density-dependent adhesion regulation fully suppresses the instability, restoring coherent front propagation. The model extends Steinberg's Differential Adhesion Hypothesis to growing, moving tissues, connects nanoscale adhesion molecules to tissue-scale morphology, and suggests that restoring adhesion regulation (e.g., E-cadherin expression) could be a therapeutic strategy against invasive cancer. Reference: arXiv:2604.26928.

The Mechanical Magic of Cell Adhesion: Why Tissues "Stick Together" or "Break Apart"?

*A new theory from Oxford's mathematical biology group: one equation unifies two strikingly different phenomena in development, regeneration, and cancer.*

---

1. A Puzzling Observation

Imagine a patch of cells growing outward. During normal embryonic development, these cells might form a finger, a segment of neural tube, or a smooth layer of skin—their edge neat, like a well-trained army advancing in formation.

But under other circumstances, the same cell population looks completely different: the front is no longer smooth but extends ragged "fingers," like oil spreading on frosted glass or ink branching as it disperses in water. Observed under the microscope, this is called a fingering instability—a classic hallmark of cancer cells invading surrounding tissue.

The question is: why do the same cells sometimes form beautiful patterns and sometimes become dangerous invaders?

For decades, biologists studied these two phenomena—orderly morphogenesis and chaotic tissue invasion—as separate problems. The former belonged to developmental biology; the latter to oncology. But new work by Falcó, Johnson, Dalwadi, and Philip Maini at the Wolfson Centre for Mathematical Biology, University of Oxford, reveals something surprising: they may be two sides of the same coin.

2. Adhesion: The "Glue" Between Cells

To understand this, we return to a fundamental social property of cells: adhesion.

Every cell's surface is covered with adhesion molecules, the most famous being E-cadherin. Think of them as tiny "hands" that cells extend; when two cells approach, these hands grip each other, sticking the cells together. Adhesion strength is not fixed—during development, cells tune the expression of adhesion molecules; in cancer, adhesion is often weakened (the so-called "E-cadherin switch"), making it easier for cells to break away and metastasize.

But what role does adhesion physically play?

In 1963, biologist Malcolm Steinberg performed a famous experiment: he dissociated cells from different embryonic tissues into single cells and mixed them together. Astonishingly, the cells spontaneously reorganized, eventually forming structures similar to the originals. Steinberg proposed the Differential Adhesion Hypothesis (DAH): cells behave like different kinds of liquids—cells with stronger adhesion tend to cluster together, pushing weakly adhesive cells to the periphery. This hypothesis was later refined by Brodland and others into the Differential Interfacial Tension Hypothesis (DITH), drawing an analogy between tissue rearrangement and fluid interfacial tension.

But Steinberg's experiment involved only static tissue rearrangement. What about collective behavior during growth and movement? That is the question Falcó and colleagues set out to answer.

3. The Story of One Equation

Mathematical physicists typically tackle such problems by writing partial differential equations (PDEs). It may sound abstract, but describing cells with an equation is as natural as describing water flow in a pipe with the Navier–Stokes equations—except the "fluid" here consists of living cells.

Falcó et al.'s model includes three key physical processes:

First, proliferation. Cells divide, producing new cells. This is the tissue's "thrust," like a pump continuously inflating a balloon.

Second, self-diffusion. Cells don't stay rigidly in place; they undergo small random movements. Interestingly, this diffusion is not ordinary Fickian diffusion—it is modulated by cell density and adhesion. When cells adhere tightly, their collective movement becomes harder, and the diffusion coefficient changes.

Third, adhesion. This is the core of the model. Adhesion between cells generates a "pulling back" force—if you try to pull a cell out of the population, adhesive forces drag it back. Mathematically, this force is modeled as a nonlocal term: the force each cell feels comes from the collective action of surrounding cells within a certain range.

Combining the three processes into one equation, schematically:

\frac{\partial \rho}{\partial t} = \text{proliferation} + \text{self-diffusion} + \text{adhesion effects}

It looks simple, but when you actually solve this equation, interesting things happen.

4. Weak Adhesion: The Orderly "Soldier Phalanx"

Suppose adhesion is weak—cells only loosely hold hands. Numerical simulations then show the tissue front forms a stable travelling wave: it advances like a flush battle line. The cell density profile is a smooth monotone curve, rising from zero at the front to a saturating density behind.

Mathematically, this is called a monotone travelling wave. It is stable because weak adhesion is insufficient to create perturbations at the interface. Any small bump is smoothed away by proliferation and diffusion—like a well-trained army where anyone slightly out of step is quickly pulled back by the group's rhythm.

This corresponds to many normal physiological processes: epidermal cell migration during wound healing, intestinal villus growth, even the expansion of certain benign tumors. The tissue maintains integrity and continuity, with no fragments breaking off.

5. Strong Adhesion: From "Phalanx" to "Fingering Invasion"

Now let's turn up the adhesion strength.

Something remarkable happens: when adhesion exceeds a certain critical threshold, the previously stable flush front suddenly becomes unstable. Any tiny perturbation along the front—perhaps just one or two extra cells in one spot—is no longer smoothed away but amplified. Protrusions emerge, keep growing forward, and eventually develop into finger-like structures extending from the main tissue.

It's like a group of people holding hands too tightly. Imagine climbers roped together: if the rope is too slack, everyone goes their own way; if it's just right, the team advances neatly; but if it's pulled too tight, one person stumbling over a rock sends violent waves down the rope, and the team may break into pieces.

In two dimensions, these finger-like structures evoke the morphology seen during cancer cell invasion—tumor cells extending "fingers" from the primary tumor into surrounding healthy tissue. In 2005, Cristini and colleagues' pioneering work first showed mathematically that tumor invasion could be viewed as a morphological instability, driven by competition between proliferation (which destabilizes shape) and adhesion (which stabilizes it). Falcó et al. now place this picture in a more universal theoretical framework: it is not only a feature of cancer, but a generic phenomenon that can arise in any adhesion-driven system.

6. The Saving Regulation: Density-Dependent Adhesion

If strong adhesion causes tissue fragmentation and dangerous fingering invasion, how do organisms solve this problem?

Nature's answer is: regulation. Falcó et al. introduce a key mechanism—density-dependent adhesion regulation. Simply put, when cell density is high (inside the tissue), adhesion can be reduced; at lower density (near the front), adhesion remains at normal levels.

The effect is striking: simulations show that even with strong basal adhesion, appropriate density-dependent regulation completely suppresses the front instability, restoring an orderly, coherent advance.

What biological reality does this correspond to? During normal development, cells indeed dynamically regulate adhesion molecule expression. In collective migration of neural crest cells, for example, "leader cells" at the front and "follower cells" behind have different adhesive properties. In wound healing, cells adjust their connection strength according to local density and environmental signals. These regulatory mechanisms keep tissue coherent and functional while growing and moving.

In cancer, this regulation often fails. E-cadherin downregulation and mutations in adhesion signaling pathways deprive cells of this fine balance—the result is disorderly invasion and metastasis.

7. Why It Matters

The significance of this work goes far beyond "explaining known phenomena with elegant equations."

First, it provides a unified framework. Previously, developmental biologists had one theory for morphogenesis and oncologists another for cancer invasion. Falcó et al. show both can be described by the same equation—the difference lies only in parameter values (adhesion strength). Weak adhesion → orderly morphogenesis; strong adhesion without regulation → chaotic invasion. It's like the van der Waals equation describing both gas and liquid phases, depending on temperature and pressure.

Second, it points to potential therapeutic targets. If fingering invasion in cancer arises from adhesion-mediated instability, restoring density-dependent adhesion regulation could be a treatment strategy. Studies have already shown that enhancing E-cadherin expression in certain cancers can suppress metastasis—fully consistent with the model's prediction that "regulating adhesion restores stability."

Third, it demonstrates the power of multiscale modeling. From single-cell adhesion molecules (nanoscale), to mechanical interactions between cells (micron scale), to tissue-level morphology (millimeter scale)—through systematic multiscale analysis, Falcó et al. link microscopic mechanisms to macroscopic phenomena. This is one of the most exciting directions in 21st-century mathematical biology.

8. Epilogue: Equations and Life

Philip Maini directs the Wolfson Centre for Mathematical Biology at Oxford, is a Fellow of the Royal Society, and one of the most influential researchers in this field. His invited lecture at the 2010 International Congress of Mathematicians was on "aspects of tumour modelling." For decades, he and his team have been translating the mysteries of life into mathematical language—from neural crest cell migration to angiogenesis, from wound healing to tumor evolution.

This latest work continues that tradition. It reminds us that even the most complex biological phenomena may, at the right level of abstraction, follow concise and elegant mathematical rules. Three terms in one equation—proliferation, diffusion, adhesion—are enough to encode whether a tissue's fate is to "grow together" or "invade and scatter."

The next time you see cells holding hands as they advance under the microscope, consider: how strong is their "grip"? If it were tightened just a little, would that neat phalanx suddenly sprain into stretching tentacles? And inside your body, every day, countless cells are using this precise mechanical balance to decide whether you are healing, developing, or fighting uncontrolled growth.

Perhaps the order and chaos of life hide in that invisible adhesive force.

---

Paper information: Falcó C., Johnson S.W.S., Dalwadi M.P., Maini P.K. *Theory of adhesion-driven self-organisation in growing tissues*. arXiv:2604.26928 (2026).