IFN-γ: The T cell’s license to kill stem cells in the inflamed intestine

See allHide authors and affiliations

Science Immunology  06 Dec 2019:
Vol. 4, Issue 42, eaaz6821
DOI: 10.1126/sciimmunol.aaz6821


IFN-γ produced by T cells directly induces intestinal stem cell death upon inflammation-induced intestinal injury (see the related Research Article by Takashima et al.).

Intestinal regeneration upon tissue damage is fueled by intestinal stem cells (ISCs) residing in the crypt bottom of the epithelium and marked by the gene Lgr5 (1, 2). There is growing evidence that tissue repair is at least partially mediated by a regenerative inflammatory response (3, 4). How inflammation-induced intestinal injury influences ISCs and their microenvironment (stem cell “niche”) remains poorly understood. In this issue of Science Immunology, Takashima et al. (5) explore the changes in the ISC niche in vivo upon T cell–mediated injury as a model of graft-versus-host disease (GVHD) and in vitro using organoid T cell cocultures. Although earlier studies already implicated interferon-γ (IFN-γ) as a negative regulator of intestinal epithelial homeostasis (68), Takashima et al. now demonstrate that IFN-γ directly acts on ISCs by triggering apoptosis.

In an allogeneic bone marrow transplant (BMT) model, Takashima and colleagues found that ISC numbers per intestinal crypt were markedly reduced in mice receiving bone marrow alone or bone marrow and T cells when compared with normal control mice. While the ISCs in the mice receiving only bone marrow recovered 7 days later, the ISC numbers remained reduced in those mice also transplanted with donor T cells. Of note, Paneth cell numbers were also reduced after ISC depletion. The numbers of organoids established from the intestines of mice 10 days after BMT recovered back to that of control mice, whereas the organoid forming capacity from crypts of mice after combined transplantation of bone marrow and T cells remained significantly lower. Similar in vivo and in vitro results were obtained when autoreactive T cells were transplanted, pointing to a common feature of T cell–mediated intestinal injury.

As seen by three-dimensional confocal microscopy, intraepithelial T cells (CD3+ IELs) preferentially localized to the villus region, whereas lamina propria–associated T cells (CD3+ LPLs) were equally distributed along the crypt-villus axis of control mice (Fig. 1A). Conversely, mice receiving bone marrow and allogeneic T cells showed a progressive increase in the density of both CD3+ LPLs and CD3+ IELs in the crypt region.

To identify signaling molecules that cause the loss of ISCs in this model, Takashima and colleagues performed several elegant murine and human epithelial organoid coculture experiments. Murine naïve allogeneic T cells did not impair murine intestinal organoid numbers, whereas alloreactive T cells effectively reduced organoid numbers. Likewise, human allogeneic cytotoxic T cells robustly inhibited human intestinal organoid forming efficiency. Even bead-activated autologous T cells suppressed human intestinal organoid growth. The authors then proceeded to screen for potential pathways mediating cytotoxicity. Organoids cocultured with T cells in the presence of anti–IFN-γ neutralizing antibodies showed normal growth. Although IFN-γ receptor (IFN-γR)–depleted T cells were still able to affect organoid viability, IFN-γR–depleted organoids were resistant to T cell–mediated killing. Organoid toxicity by IFN-γ was also observed in the absence of T cells. Live imaging confirmed the progressive ISC depletion upon organoid exposure to IFN-γ. Treatment of organoids with the immunosuppressive JAK1/2 inhibitor ruxolitinib robustly preserved numbers of both organoids and ISCs in the presence of IFN-γ, irrespective of whether the organoids were cultured alone or together with T cells. The authors additionally demonstrated that JAK1-depleted organoids are resistant to IFN-γ treatment. Further downstream, ruxolitinib prevented STAT1 phosphorylation by IFN-γ in intestinal crypts, and, in line, STAT1-depleted organoids were resistant to growth suppression in response to IFN-γ treatment.

IFN-γ–treated organoids showed reduced expression of ISC marker genes. ISCs underwent apoptosis in vitro in a direct response to IFN-γ. Next, the authors confirmed in vivo that ISC numbers did not change upon transplanting allogeneic bone marrow and T cells when treating mice with IFN-γ neutralizing antibodies. Likewise, ruxolitinib treatment protected ISCs from T cell–mediated killing in vivo. Donor T cells, particularly T helper 1 cells, were activated and IFN-γ+. Transplanting IFN-γ–depleted allogeneic T cells robustly reduced the ISC loss and allowed epithelial cell proliferation to increase.

Takashima and colleagues lastly investigated whether IFN-γ directly induces ISC apoptosis. Using tissue-specific depletion of IFN-γR1, the authors found that epithelial loss of the receptor protects from the immune-mediated GVHD phenotype. IFN-γR1 is expressed by both ISCs and Paneth cells, the epithelial component of the ISC niche (9). However, Paneth cell–deficient organoids remained sensitive to both IFN-γ– and allogeneic T cell–mediated cytotoxicity. Likewise, T cells were able to reduce the number of organoids containing IFN-γR1–deficient Paneth cells, whereas organoids containing IFN-γR1–deficient ISC were protected from cytotoxicity. The authors demonstrated in further experiments that IFN-γ directly induces ISC apoptosis independent of Paneth cells (Fig. 1, B and C).

The study by Takashima et al. extends our knowledge on signaling between ISCs and immune cells, identifying ISCs as direct targets of IFN-γ secreted by T cells in immune-mediated intestinal damage (as caused by GVHD). In the 2015 study by Lindemans et al., this group already identified that interleukin-22 (IL-22) secreted by group 3 innate lymphoid cells (ILC3s) directly stimulates ISCs to proliferate and regenerate the intestinal epithelium upon inflammation-induced intestinal injury (4). Modulating the effects of T cell–derived IFN-γ on ISC, for instance, by suppressing JAK/STAT signaling via ruxolitinib treatment, may provide a new therapeutic avenue to reducing GVHD-induced damage of the intestinal epithelium (10).

Fig. 1 In immune-mediated intestinal damage, T cell–produced IFN-γ directly induces ISC apoptosis.

(A) ISCs maintain adult homeostasis of the intestinal epithelium. T lymphocytes patrol the intestine. (B) Takashima et al. show that in GVHD as modeled by BMT and aberrant activation of T lymphocytes, T cell–derived IFN-γ directly acts on ISCs and induces apoptosis via JAK/STAT signaling. (C) Disease progression results in marked intestinal damage due to loss of ISCs and their niche.

Credit: A. Kitterman/Science Immunology


Acknowledgments: Funding: K.K. is a long-term fellow of the Human Frontier Science Program Organization (LT771/2015). Competing interests: H.C. and K.K. are named inventors on patents or patents pending on Lgr5 stem cell–based organoid technology.

Stay Connected to Science Immunology

Navigate This Article