Graft-Versus-Host Disease Induces Immune Dysregulation through Irreversible Damage to Specialized Immune-Interacting Fibroblastic Stromal Cells in Secondary Lymphoid Organs

Abstract

Allogeneic hematopoietic cell transplantation (allo-HCT) is a potentially life-saving treatment for patients with leukemia and other hematological disorders. However, its success is limited by graft-versus-host disease (GVHD), a life-threatening syndrome caused by alloreactive donor T cells that induce damage in normal host tissues. Many patients who survive transplantation develop chronic GVHD which can be severe and has few effective therapeutic options. Chronic GVHD is characterized both by immune hyperreactivity, reminiscent of autoimmunity, and defective immune responses, including poor responses to vaccination and infectious challenges. The mechanisms underlying the complex immune dysregulation of chronic GVHD remain incompletely understood. Previous work indicates that pathogenic alloreactive T cells can damage lymph node fibroblastic stromal cells, a group of specialized cells increasingly recognized to regulate multiple aspects of immune homeostasis. Using mouse models of allo-HCT, we sought to identify the populations of specialized fibroblastic stromal cells and other stromal elements that were most sensitive to GVHD-mediated injury and to assess the functional impact of this damage during chronic GVHD. We used a sublethal allo-HCT model with a low T cell dose that allowed for long-term survival over 100 days after transplantation (B6 into lethally irradiated B6xBALB/c F1 mice, 2x 5.5 Gy). Unlike allo-HCT recipients of T cell-depleted bone marrow (no GVHD control), recipients of T cell-depleted bone marrow plus allo-T cells (1-5x106 splenocytes) developed low-grade systemic GVHD with mild reversible weight loss. At day 30-60 after transplantation, allo-T cell recipients showed modestly decreased thymopoiesis, consistent with the thymus being a sensitive target organ of GVHD, as previously reported. In contrast to this mild injury, peripheral lymph nodes (pLN) showed profound long-lasting depletion of CD31-/podoplanin+ fibroblastic stromal cells as assessed by flow cytometry. Among these cells, MAdCAM-hi marginal reticular cells (MRCs) and CD157-hi fibroblastic reticular cells were most severely affected, including most cells lineage-traced by a Ccl19-Cre transgene. Total blood endothelial cells and lymphatic endothelial cells (LECs) were only modestly decreased by GVHD at day 30, but the LEC subset characterized by high MAdCAM expression consistent with a floor LEC phenotype virtually disappeared. Thus, MRCs and floor LECs lining the marginal sinus were severely affected by GVHD, suggesting that antigen uptake and processing could be defective in these pLN. Single cell RNA-seq revealed a near complete loss of floor LECs, MRCs and Ccl19/Ccl21+ T zone FRCs in GVHD-affected pLNs. Instead, residual fibroblastic stromal cells showed accumulation of collagen-expressing Cxcl14+ fibroblasts with loss of immune-interacting properties. In terms of functional consequences, GVHD pLNs had a markedly decreased abundance of naïve B and T cells, the latter despite relatively preserved naïve T cell production in the thymus. Compared to pLNs, the spleen was less profoundly affected and remained more receptive than pLNs to homing of adoptively transferred T cells. Immune responses to SARS-CoV2 mRNA used as a potent and clinically relevant T-dependent antigen were profoundly impaired in GVHD pLNs. In summary, our data suggest that specialized fibroblastic stromal cells and LECs in secondary lymphoid organs are even more sensitive to GVHD than thymic epithelial cells, with little evidence of functional recovery. We speculate that subclinical damage to pLN stromal compartments may underlie many aspects of immune dysregulation in chronic GVHD, including persistently impaired responses to vaccination and loss of peripheral tolerance.