Project Abstracts - CMAD

Remi Creusot, PhD, Columbia University
Holger Russ, PhD, University of Florida

Start Date: December 15, 2025

---

Abstract

Genetic predisposition to T1D, in addition to HLA, includes >200 non-HLA variants that each add small incremental risk. Inbred NOD mice represent only one genotype and therefore cannot capture diversity imposed by the highly variable human T1D-conferring genotypes. We can replicate this genetic variability in vivo in a human immune system (HIS) mouse model in which T1D patient- and healthy control (HC)-specific immune systems are constructed de novo from CD34+ hematopoietic stem and progenitor cells (HSPCs) of volunteer donors in immunodeficient mice that also receive human thymic tissue under the kidney capsule. T1D-derived “Personalized Immune (PI)” mice containing patient-specific T cells, B cells and myeloid antigen-presenting cells demonstrate HSC-intrinsic defects in thymic selection of autoreactive T cells, in peripheral T cell homeostasis, and in B cell negative selection. We now propose to combine the PI mouse model with stem cell technology by generating autologous induced pluripotent stem cells (iPSCs) and differentiating β-like cells (sBCs) and thymic epithelial cell progenitors (sTEPs)/thymic epithelial cells (sTECs) from them. Our novel toolset includes transduction of known human islet autoreactive (IAR) T cell receptors (TCRs) into HSCs, which reveals differences in intrathymic selection in HC and T1D-derived immune systems. We have used fetal porcine thymic tissue to support robust human thymopoiesis and demonstrated the impact of integrated human sTEPs on human thymocyte selection in vivo. We have generated a novel immunodeficient mouse line with improved lymphoid structure (due to preserved lymphoid tissue inducer [LTi] cells) compared to immunodeficient NSG mice. We propose a new multidisciplinary collaboration, including expertise in HIS mice, stem cell biology, genetic engineering, and human TCR repertoire analysis, to use this toolset to investigate previously inaccessible aspects of human T1D. Specifically, we will: Aim 1: Use iPSC-derived sTEP/TECs from T1D and HC donors to assess the role of genetically-controlled TEC-intrinsic differences in selection of autoreactive TCR repertoires in T1D. We will evaluate the impact on TCR repertoire, including IAR T cell selection, of integrating HC- vs T1D-derived iPSC-derived sTEP/TECs into pig thymus grafts in PI mice and in T1D and HC iPSC-derived thymic organoids in vitro and in vivo; and Aim 2: Develop isogenic T1D patient models resulting in β cell graft destruction. Using the new recipient mouse strain with functional LTi cells, we aim to achieve increased β cell infiltration/destruction by human T cells. By incorporating genetic engineering, we will model autologous β cell destruction in T1D PI mice with isogenic TECs and sBC grafts. We will use these models to compare T cell responses in T1D and HC PI mice to a novel form of antigen-specific immunotherapy. Collectively, these studies will result in autoimmune β cell destruction in HIS mice with completely isogenic immune, thymic and β cells. These models will allow generation of novel insights into the HSC- and TEC-intrinsic abnormalities that drive T1D pathogenesis and into their potential responses to immunotherapy.