Tipping the balance of the immune system toward the destruction of tumors

Cancer is immunogenic, that is, it is able to induce an immune response, like microbes or vaccines. Indeed, in patients with cancer, immune cells may be found that are directed against specific molecules - named tumor antigens - made by cancer cells. It has also been shown that the tumors are naturally infiltrated by a variety of immune cells.

These observations have stimuated the development of anti-cancer immunotherapy for the last 20 years: an approach that seeks to reinforce the activation of the immune system against the tumor.

The attitude of immune cells against a tumor is nevertheless often contradictory. Some of these cells attack the tumor while others spare it and even encourage its growth. The team of Anne Marie Schmitt-Verhulst is trying to better understand these phenomena in the hope of improving existing immunotherapies.


The role of the immune system in controlling tumor growth is in fact highly ambiguous. Certainly, for example, the precise location of some lymphocytes in colorectal tumors is a powerful predictor of patient survival, but the presence of other immune cells in other parts of the tumor is rather associated with a poor prognosis.

The tumor uses cells in its environment, including cells of the immune system, for its own benefit, either in direct support of its growth, or by inducing cells called "suppressor cells" capable of counteracting the effect of immune cells. Moreover, among the immune cells involved in the battle against the tumor, some retain their ability for long periods while others "falter" quickly after the first attack.

Dissect cancer-immune interactions...
from the mouse to the patient

"Animal models using transplanted tumors poorly duplicate the complexity of phenomena that are brought into play when the tumor begins to develop and grow, including the cellular interactions that are involved in this process" says Anne-Marie Schmitt-Verhulst. "Tumors are organized clusters of cancer cells installed in a stroma. This complex structure, irrigated by blood vessels induced by the tumor itself, is invaded by many types of immune cells: some react against the tumor, while others promote its proliferation, others have become indifferent toward the tumor. How do these interactions take place? The answer to this question determines our ability to tip the balance of the system towards the elimination of the tumor."

To decipher these phenomena, the team of Anne-Marie Schmitt-Verhulst has developed a unique model of melanoma in mice.

Melanoma is a skin cancer that has a poor prognosis because it is often diagnosed late and therefore has already metastasized. This cancer is well known for its immunogenicity. Historically, it is in the treatment of melanoma that most immunotherapy approaches were and still are being attempted, with varying success: molecules stimulating all lymphocytes, cell therapy based on injection of activated immune cells, factors blocking suppressor cells and even anti-tumor vaccines.

"With our inducible melanoma model in mice we were able to develop a tumor whose architecture resembles some "real" human tumors," says Anne-Marie Schmitt Verhulst. Because it contains all the types of cells that are actually involved, this tumor model allows:

  • To study how the immune system perceives the growth of the tumor in situ;
  • To better understand how signals sent by the tumor influence the development of immune cells;
  • To analyze the effects of "super" lymphocytes, whose signaling pathways have been manipulated, giving them increased power to eliminate tumor cells;
  • To interpret the biological effects induced by immunotherapy on cancer cells.

"Ultimately, comparing the profiles of immune cells in our model with those from biopsies of patients, we hope to contribute to the development of new immunotherapies against melanoma. These treatments may be capable of mobilizing more efficient immune cells and overcome the constraints induced by the tumors," concludes Anne-Marie Schmitt-Verhulst.