How pancreatic cancer cells take membrane fragments from neighbors

In the dense tissue of pancreatic tumors, pancreatic cancer cells take membrane fragments from neighboring fibroblasts to obtain cholesterol. Those fragments can help the cancer survive in a place where nutrients are scarce. They may also explain an apparently separate problem: why some T cells inside the tumor struggle to carry out an immune attack.

Pancreatic ductal adenocarcinoma is often embedded in a thick mass of connective tissue. Much of it is built by cancer-associated fibroblasts, cells that help organize the tissue around a tumor but can also support its growth.

This dense environment contains surprisingly few functional blood vessels.

That creates a puzzle. Cholesterol and other lipids are essential for making new cell membranes, so where do rapidly dividing cancer cells get them?

A detour through a neighboring cell

Charline Ogier and colleagues traced fluorescently labeled cholesterol carried in low-density lipoprotein particles in mice with pancreatic tumors.

The label appeared first in cancer-associated fibroblasts and later in malignant cells, particularly where the two cell types touched. Cancer cells immediately adjacent to fibroblasts accumulated about twice as much labeled cholesterol as cells farther away.

The transfer depended on physical contact. When fibroblasts and pancreatic cancer cells grew together in culture, the cancer cells rapidly acquired pieces of fibroblast membrane. Separating the two cell types with a barrier that still allowed soluble molecules to pass almost eliminated that transfer.

This form of cellular exchange is called trogocytosis, from a Greek root meaning "to gnaw."

Rather than engulfing an entire neighboring cell, one cell takes a small piece of its surface. In this case, the acquired membrane provides ready-made lipids, including cholesterol.

To find out whether the transfer actually mattered, the researchers used cancer cells whose ability to synthesize their own cholesterol had been experimentally disrupted.

In lipid-poor conditions, the cells died unless cholesterol was supplied. Direct contact with fibroblasts rescued their survival and supported further growth. Fibroblast-conditioned medium did not.

That result reinforced the conclusion that the crucial route was contact-dependent rather than simply the release of nutrients into the surroundings.

The signal hidden in a touch

The membrane exchange begins with a molecular response inside the fibroblast.

Contact with a cancer cell triggers calcium entry through Orai channels. The rise in calcium activates ANO6, or anoctamin 6, a protein that moves phospholipids between the two layers of the cell membrane.

One of those lipids, phosphatidylserine, normally sits mostly on the inner side. ANO6 activity brings it to the outer surface, where it helps initiate the uptake of a membrane fragment by the cell touching the fibroblast.

The donor is still alive. This is not the usual disposal of a dead cell, but a direct exchange of membrane material between living neighbors.

Reducing ANO6 in the fibroblasts almost abolished membrane transfer to pancreatic cancer cells. Restoring a functional copy of the protein brought the process back.

The work therefore connects an unusual way of feeding a tumor to a specific calcium-dependent mechanism.

The same exchange can disable immune cells

There is another recipient for these membrane fragments.

Cytotoxic T cells, whose job is to kill malignant cells, can also take pieces of membrane from nearby fibroblasts. In doing so, they acquire cholesterol. In the tumor environment, excess cholesterol is associated with T-cell dysfunction and exhaustion.

When the researchers suppressed ANO6 in fibroblasts, cultured T cells took up fewer membrane fragments and less cholesterol. They produced more interferon-gamma and became better at killing target cells.

In mice bearing pancreatic tumors, treatment with clofazimine was associated with lower cholesterol levels in tumor-infiltrating T cells, stronger signs of activation, and fewer markers of exhaustion.

The team also tested clofazimine and niclosamide, existing antimicrobial drugs known to interfere with ANO6 activity. In mouse models, three weeks of treatment reduced pancreatic tumor weight. Tumor grafts also grew more slowly in mice lacking ANO6.

Analyses of human pancreatic cancer datasets found that higher ANO6 expression was associated with poorer overall survival. That association is consistent with the proposed mechanism, but cannot establish that ANO6 itself causes worse outcomes.

These findings are preclinical. Neither drug acts exclusively on ANO6, and the protein affects other tumor-supporting properties of fibroblasts, including the surrounding extracellular matrix.

The experiments cannot neatly separate the effects of reduced lipid delivery from those of broader changes in fibroblast behavior. They also do not demonstrate an effective cancer treatment in patients.

What makes the mechanism striking is its double use.

In a tumor with limited access to blood-borne nutrients, a fibroblast can serve as an intermediary delivering cholesterol to cancer cells. Through the same kind of membrane contact, it can also transfer lipids to T cells and diminish their ability to fight the tumor.

Two advantages for the cancer begin at the surface of a single neighboring cell.