Biomedical research institute
    You are here: Home / Institute / News / Mammary tumors: When TGF beta disarms type I interferons

    Mammary tumors: When TGF beta disarms type I interferons

    A study directed by Nadège Bercovici, team Emmanuel Donnadieu

    The group directed by Nadège Bercovici (team Emmanuel Donnadieu) publishes in Nature Communications, results of major importance.

    In cancer immunotherapy, studies on mouse models of transplanted tumors frequently lead to very promising results. Unfortunately, the same treatments often fail in human carcinoma and in murine models of spontaneous tumors. Why? The group of Nadège Bercovici at the Institut Cochin has just found a beginning of explanation. They have shown that TGFbeta, a molecule that is abundant in many human tumors, and associated with a poor prognosis, is also present in spontaneous murine mammary tumors, where it prevents the action of anti-cancer treatments.

    This team had previously shown, in a mouse model of transplanted tumors, that by mimicking a viral attack with a STING ligand (cytosolic receptor involved in the detection of cytoplasmic DNA of microbial origin), one can induce a strong decrease of the tumor mass. In this model, tumors were obtained after the injection of tens of thousands of tumor cells at once. Their regression, triggered by the STING ligand, was associated with the induction of a massive production of type I interferons alpha and beta (IFNa/b) by tumor infiltrating macrophages (Mφ). The STING ligand allowed the Mφ to acquire a cytotoxic activity, and to alert T lymphocytes which then amplified the anti-tumor Mφ attack. All this occurs as if the STING agonist set in motion an antiviral arsenal, used to attack the tumor.

     In addition to the transplanted tumor model, there is a spontaneous one in which mammary tumors appear several months after the birth of transgenic mice expressing an oncogene in this gland. Unfortunately, these spontaneous tumors, which are a better model for breast cancer, were almost insensitive to the STING ligand. How can one explain this?

    TGFbeta is a molecule that is required for the development of tissues and their repair, whether embryonic or tumoral. In a non-infectious context, in most tissues, this molecule is notably produced and used by Mφ to fulfill their function as garbage collectors that keep tissues clean and without inflammation. Contrary to what happens in transplanted tumors, TGFbeta is abundant in spontaneous tumors, where it promotes tumor growth. 

    The present study demonstrates that TGFbeta can also disable the antiviral alert system by turning off the STING-IFNa/b pathway: Precisely what is needed to allow an uncontrolled growth of the tumor ... Fortunately, by treating the mice with an anti-TGFbeta antibody, the team of Nadège Bercovici has shown that it was possible to reset the alert STING-IFNa/b system and thus to allow a STING ligand to trigger an anti-tumor response accompanied by a regression of spontaneous tumors.

    This result is of major importance. Indeed, in human tumors, similar mechanisms of inhibition of type I IFNs are likely to hinder the effectiveness of STING agonists, currently in clinical trials. The same antagonism between TGFbeta and IFNa/b could explain some phenomena of resistance to anthracyclines, whose anti-tumor activity also involves the production of IFNa/b in the tumor. Clearly, these interferons are not only antiviral, but can also be anti-tumoral.

    Thus, this work makes it possible to predict that the effectiveness of a number of anti-tumor therapies could be amplified by the transient inactivation of TGFbeta. It is known that, in human tumors, its abundance is of poor prognosis. These studies show that there will be a great interest in the coming period, to combine existing treatments with anti-TGFbeta or inhibitors of its signaling, which are already available.


    Figure Legend

    The presence of TGFbeta in spontaneous mammary tumors (top left) polarizes macrophages to an immunosuppressive phenotype. Under these conditions, the addition of a STING agonist (bottom left) does not induce the production of IFNa/b and neither tumor regression. In the presence of an anti-TGFbeta treatment (top right), the macrophages switch to an activated phenotype, which, following the addition of the STING agonist, allows the production of IFNa/b  and tumor regression.


    Learn more

    TGFbeta blocks IFNa/b release and tumor rejection in spontaneous mammary tumors. Guerin MV, Regnier F, Feuillet F, Vimeux L, Weiss JM, Bismuth G, Altan-Bonnet G, Guilbert T, Thoreau M, Finisguerra V, Donnadieu E, Trautmann A, and Bercovici N. Nature Communications, published online on Sept. 11, 2019


    Researcher contact