476. No GVHD, but Human Inflammatory “Cytokine Storm” and Mouse Macrophage Activation Upon Accelerated Development of Human CD4+ Effector Memory T Cells in Long-Term Humanized Mice

Abstract

Non clinical assessment of immunotoxicity of human cellular products in vivo is limited by the lack of availability of pharm-tox models enabling a full range of human innate and adaptive responses. We previously demonstrated improved development of human adaptive T and B cell responses in mice receiving human hematopoietic stem cell transplantation (HSCT) in combination with prime/boost immunization with induced dendritic cells (iDCs), i.e. monocytes derived the stem cell donor reprogramed with a lentiviral vector for co-expression of GM-CSF, IFN-alpha and the CMV antigen pp65 that self differentiate into DCs in vivo (Salguero et al, 2014 J. Immunology; Daenthanasanmak et al, 2015 Mol. Ther. Meth.). For these previous studies, mice were kept for 6-10 weeks after iDC administration, and we did not observe signs of tumorigenicity or graft-versus-host disease (GVHD). After consultation with the German regulatory authorities, we were requested to perform pharm-tox analyses 26 weeks after iDC immunization. We report here a pilot-feasibility study with humanized mice showing long-term (more than 33-36 weeks after HSCT) robust immune reconstitution with human T cells (control mice, n=2 Vs. iDC-immunized mice, n=6). Longitudinal analyses of human immune reconstitution in blood samples confirmed accelerated development of CD4+ effector memory T cells 5-10 weeks after iDC immunization. No weight loss, GVHD or malignancies were observed after iDC immunization. 3/6 immunized mice developed skin erythema and inflammation around 23 weeks after iDC immunization and were sacrificed earlier for analyses. Analyses of all mice were performed at autopsy by macroscopic observations, H&E histopathology, immunohistochemistry and flow cytometry of target organs. Positivity for anti-human nuclei antibody, anti-human HLA-DR and anti-human CD3 leucocytes were detected in several tissues (spleen, bone marrow, brain, lungs, skin, eyelids, liver, and kidneys), confirming the long-term persistency of human T cells in the humanized mice. In inflamed skin, abundant human CD4+ T cells were localized in the basal layers of the epidermis and hair follicles in the dermis. Human macrophages (CD68+) could be detected in several tissues, but were rare in inflamed skin. However, in the inflamed dermis areas of ulceration, mouse macrophages (F4/80+) were observed frequently with strongly positive cytoplasm. Mice with ulcerations had significantly higher levels of several human inflammatory cytokines in plasma (IFN-gamma, IL-6, GM-CSF and IL-8) than immunized mice with no skin inflammation or control mice. In summary, a long-term pilot study exploring a hybrid human-mouse model system was feasible and iDCs accelerated mature human T cell reconstitution with no signs of tumorigenicity or GVHD. However, accumulation of human CD4+ memory T cells and activated mouse macrophages in inflamed skin was associated with elevated levels of human inflammatory cytokines resembling a “cytokine storm”.