CAR-T Cells: A New Tool for Monitoring T-cell Alloreactivity?

Models and assays for testing alloreactivity: what are they good for? Are they helpful to mimic the clinical situation? Over the past 3 decades very many papers have been published reporting the outcomes of studies unraveling the recipient’s immune response to a transplanted organ, a response that is responsible for either rejection or tolerance. For these studies, a wide range of in vitro and in vivo models and tools have been used. After so many years of research, it is clear that a single biomarker assay often lacks sensitivity and specificity for predicting rejection or tolerance, whereas the translational feature of observations in models for testing alloreactive responses remains challenging.1,2 To improve our understanding of transplant immunology, which is also critical for the development of immunosuppressive drugs and therapeutic approaches, the availability of reliable and well-performing preclinical models is needed. For these studies, humanized immunodeficient mice models have been used regularly. These models allow us to study the behavior and effects of newly developed therapeutic agents—such as cellular therapies—on the human immune system. Humanized immunodeficient animals overcome the biological differences that exist between humans and other species.3 However, xenogeneic graft versus host disease (xGVHD) is a commonly seen complication in immunodeficient mice treated with human peripheral blood mononuclear cells (PBMCs). This mouse model facilitates short-term follow-up studies, but long-term study is not possible because after 3–5 wk, these animals develop xGVHD, which affects scrutiny of the allogeneic immune response. Development of xGVHD in immunodeficient mice is dependent on the relative engraftment of human T cells and driven by recognition of host major histocompatibility complex class I and class II molecules.4 Additionally, immunodeficient mice have difficulties engrafting human B cells that, together with the underdeveloped lymphatic tissues, impair the activation of and crosstalk between human T and B cells.3,5 Despite these limitations, numerous publications have shown that data from humanized small animal studies more closely resemble biological reality in humans compared with rodent models.3 At the same time, however, humanized animal models with better developed lymphatic tissues, absence of xGVHD and testing systems are required to study interactions between immune cells and to determine the functionality of novel immunosuppressive therapies. For example, the activity and interaction of chimeric antigen receptor-regulatory T cells (CAR-Tregs), with cells of the host immune system.6,7 In this edition of Transplantation, Ellis et al8 propose a new model: human A2-CAR T cells as a tool to study T-cell–mediated rejection of human islets transplanted into mice without the complication of xGVHD. CARs are genetically engineered receptors that provide specific properties to immune cells including T cells, Tregs, B cells, and natural killer cells.9 Today, with the large-scale production of CAR cells, cell therapy has arrived in the clinic. Engineered autologous Tregs are given to transplant recipients to dampen the allogeneic immune response or even to induce transplant tolerance,6,7 whereas CAR T cells therapy is successful in cancer treatment.10 Ellis et al8 show that human T cells engineered with HLA-specific CARs can also be used as a “human” preclinical model to study the process of allograft rejection in humanized immunodeficient mice without the complication of xGVHD.8 This newly developed method has clear advantages over currently used preclinical animal models. It provides the option for functional testing of human immune cell interactions in small animals while simultaneously eliminating some of the hurdles of animal studies by creating a “human” environment. Using a type 1 diabetes model, it was shown that A2-CAR T cells home to grafts and mediate rapid and synchronous rejection of human HLA-A2+ islets or embryonic stem cell–derived β-like cells transplanted under the kidney capsule or in the anterior chamber of the eye of non-obese diabetic scid gamma mice. xGVHD was only observed in mice that received human PBMCs. Importantly, the reader should be aware that this tool was developed as a better humanized mouse model of T-cell–mediated islet rejection for which the authors studied the ability of CAR T cells to reject human cell grafts; it is not an islet transplant model. Compared with the classical immunodeficient mice models engrafted with human PBMCs, this HLA-A2-CAR T cell model now opens the option to study in vivo, allogeneic immune response for a longer period as up to 12 wk the xGVHD complication is absent. Although it is greatly improved, this animal model is not yet perfect. A limitation is the lack of antigen-presenting cells. In conclusion, the promising observations reported in this study by Ellis et al8 on improving humanized mouse models helps us to better understand the biology of the T-cell–mediated rejection processes. This is of essential importance to further improve the success of transplantation, including islet-replacement therapies. These humanized mouse models also facilitate future invention and translation of novel therapeutic immunosuppressive drugs/approaches. It is, however, important to keep in mind the murine background of these humanized models.