ALBERT

Description: Abstract 3750 Graft-vs-Host disease (GVHD) is the major complication of allogeneic hematopoietic cell transplantation (HCT). Murine models have been critically important to define the biological mechanisms and potential pathways of intervention of GVHD prevention and treatment. Although it is well recognized that GVHD occurs in response to minor histocompatibility antigens, little is know about the kinetics of donor T cell proliferation and homing in minor mismatch models. This is in contrast to models across major histocompatibility barriers where the early development of GVHD has been more thoroughly characterized. In prior studies across major barriers, we have defined an initiation phase within the first 3 days where conventional CD4+ and CD8+ T cells (Tcon) home to secondary lymphoid tissues, proliferate and up-regulate key homing markers allowing for entry into GVHD target tissues during the effector phase (Beilhack, et al. Blood 106:1113, 2005). Since minor models are more similar to clinical HCT, it is critical to understand the timecourse of GVHD development across minor histocompatibility barriers. Since the manifestations of GVHD in recipients of minor mismatch transplants are delayed, it is possible that disease development has altered kinetics. To investigate the temporal and spatial events of donor T cell activation and homing, side-by-side transplants were conducted using T cell depleted bone marrow (TCD BM) and Tcon from donor C57BL/6 (H2b) mice into either major mismatched Balb.c (H2d), or minor mismatch Balb.b (H2b) recipients. Balb.c mice received 1×106 Tcon while Balb.b mice were given 15×106 Tcon, based on previous titration experiments. Recipient mice were regularly scored for GVHD symptoms and monitored for at least 100 days for survival. Additionally, donor Tcon proliferation and migration were monitored longitudinally using in vivo and ex vivo bioluminescent imaging (BLI) by quantitating photons emitted by luciferase (luc+) expressing donor Tcon isolated from luc+ transgenic mice. Donor Tcon were also labeled with CFSE to determine proliferation kinetics at selected timepoints. The upregulation of T cell activation and tissue specific homing markers was examined using flow cytometric analysis of donor CD4+ and CD8+ T cells re-isolated from the secondary lymphoid tissues of transplanted mice. In both models, T cells initially home to secondary nodal sites by 3 days post-transplant, with an exodus into the tissues by day 6, albeit to a lesser extent in recipients of minor mismatch transplants. Additionally, similar levels of global donor CD4+ and CD8+ T cell proliferation between the models were observed using both BLI and CFSE staining as early as 3 days after transplant (BLI, p〉0.05, n=9). More noticeable reductions in minor mismatch recipients were apparent by day 6 (BLI p

Description: Abstract 1809 Various imaging platforms are well established in hematology research. Nevertheless, the three-dimensional architecture of the bone marrow and tumor growth within this microenvironment remain largely uncharacterized. To date the major hindrance to microscopically image tumor engraftment and the immune response in the bone marrow on a single cell level is the compact structure of the bone that is almost impossible to image through. Therefore, we developed a novel bioluminescent mouse model that recapitulates the clinical characteristics of MM using the new human UMM3 cell line (CD38+, CD56+, CD138+, CD19−, CD20−), from the pleural effusion of a patient with an IgG lambda myeloma (ISS stage I) as well as the well-characterized RPMI8226 cell line transduced to express eGFP and firefly luciferase (UMM3eGFPluc and RPMIeGFPluc). 1×106 MM cells were injected intravenously into NOD.Cg-Prkdcscid IL2rg (NSG) mice and disease progression and bone marrow (BM) engraftment were monitored twice weekly by in vivo bioluminescence imaging. Both cell lines homed to the BM compartment, reflecting MM pathophysiology. Histological analysis confirmed BM engraftment and showed multiple osteolytic lesions for both UMM3 and RPMI cells. Since we were interested in imaging the interactions between human MM cells and the bone marrow microenvironment on a single cell level, we employed the multi-color LSFM after decalcification, specific deep-tissue antibody staining and clearing of the bone structures. With this innovative microscopy technique, we were able to establish a novel tool to display tumor cell engraftment in the bone marrow compartment in three dimensions through the intact bone. We recorded 1500 optical sections for three individual channels each (488, 532, and 647 nm) with an increment of 5μm which allows scanning the whole bone marrow compartment of the sternum within minutes in single-cell resolution. Using higher magnification enabled us to even visualize subcellular components within the bone marrow. Moreover, tissue autofluorescene, recorded mainly in the 488 nm channel, displayed detailed microanatomical structures which allowed for the localization of individual cells within their anatomical context. We could establish protocols for various fluorophore-coupled antibodies and successfully stained CD138+ cells in relation to CD3+ cells and to the microenvironment in the bone marrow. The CD138-positive cells infiltrated the bone marrow in a number of small clusters and comprising about 15% of cellular elements in total. Ex vivo bioluminescence imaging of the sternum from UMM3 tumor-bearing mice revealed massive infiltration of luciferase-expressing cells into the bone marrow compartment. This could also be confirmed by flow cytometrical analysis of bone marrow cells which showed eGFP+hCD138+ cells. We have successfully introduced a novel technique to study MM cell engraftment and progression in a humanized mouse model. We were able to track the tumor cells both in the living animal by in vivo bioluminescence imaging and on single-cell resolution by multi-color LSFM within the intact bone. Our model may lead to better insights into the pathogenesis of MM and could serve as a model for preclinical testing of new therapeutic approaches for the treatment of MM patients. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 1896 The curative potential of MHC-matched allogeneic bone marrow transplantation (BMT) is in part due to immunologic graft-versus-tumor (GvT) reactions mediated by donor T cells that recognize host minor histocompatibility antigens. Immunization with leukemia-associated antigens, such as Wilm's Tumor 1 (WT1) peptides, induces a T cell population that is tumor antigen specific. We determined whether BMT combined with immunotherapy using WT1 peptide vaccination of donors induced more potent anti-tumor activity when combined with allotransplantation. WT1 peptide vaccinations of healthy syngeneic or allogeneic donor mice with a 9-mer WT1 peptide (amino acids 126–134, the WT1 9-mer which has the highest binding affinity for H-2Db) and Incomplete Freund's Adjuvant induced CD8+ T cells that were specifically reactive to WT1-expressing FBL3 leukemia cells. We found that compared to vaccination with IFA alone, four weekly WT1 vaccinations induced an increased percentage of WT1-tetramer+CD8 T-cells (0.15% vs. 1%) in the peripheral blood 28 days following the first vaccination (Figure A *p

Description: Abstract 1000 Natural killer (NK) cells exhibit in vitro cytotoxicity against many tumor cell types and have an important role in controlling tumor growth, as depletion of NK cells from tumor-bearing mice hastens tumor growth and impairs survival. These data, in combination with results from clinical trials of haploidentical killer immunoglobulin-like receptor (KIR)-ligand mismatched bone marrow transplantation, led to interest in the use of adoptive NK immunotherapy for the treatment of malignancy. Recent clinical results have shown that allogeneic NK cells can be safely administered after chemotherapy and/or irradiation but have also demonstrated limited persistence of the infused NK cells without clear evidence of efficacy. We traced the fate of adoptively infused NK cells in order to delineate the barriers to successful NK immunotherapy using several NK-sensitive murine tumor models. Mice bearing established lymphoma or leukemia received an intravenous infusion of 0.5–1.0×106 (approximately 2.5–5×107/kg body weight) NK cells from allogeneic or syngeneic donors after or concurrently with total body irradiation and bone marrow rescue. Using luciferase +ve NK cells, we first showed that in animals bearing subcutaneous tumors, NK cells homed to lymphoid organs in the first week, followed by progressive localization to and accumulation within the tumor site (Figure 1). In contrast, in non-tumor bearing animals NK cells homed to lymph nodes, spleen and liver with maximal proliferation at the end of the second week. These observations indicated that NK cells fail to eradicate the tumor despite prior demonstration of in vitro sensitivity, successful homing and local accumulation. As expected from these findings, survival was not prolonged by the NK cell infusion. Reisolation of donor NK cells within 18 hours of transfer showed enhanced cytotoxicity (14% vs 2%, p = 0.004) and IFNγ production (48% vs 22%, p = 0.04) compared with naive resting NK cells. In contrast, NK cells isolated from tumor-bearing mice at later time-points beginning d+5 showed loss of IFNγ production (48% early vs 3% late, p = 0.01), decreased expression of the activating receptor NKG2D, and impaired cytotoxicity in chromium release assays. These observations did not relate to over-stimulation through NKG2D, as NK cells from NKG2D−/− animals were also susceptible to acquired dysfunction (although their baseline cytotoxicity was lower than WT NK toward A20 lymphoma).Fig. 1Adoptively transferred luciferase-transgenic NK cells accumulate within the tumor over timeFig. 1. Adoptively transferred luciferase-transgenic NK cells accumulate within the tumor over time Eomesodermin and T-bet are transcription factors with important roles in effector functions of CD8+ T cells and NK cells. In T cells T-bet downregulation has been shown to correlate with exhaustion (Kao et al, Nat Imm 2011). Flow cytometry of reisolated NK cells revealed downregulation of Eomesodermin (naive splenic control, d+1 reisolated and d+10 reisolated cells showing 85%, 96%, and 29% expression, respectively) and T-bet (naive splenic control, d+1 reisolated and d+10 reisolated cells showing 82%, 99%, and 59% expression, respectively), correlating with loss of IFNγ production. The phenotype described herein was most dramatic within the tumor and within mice carrying high tumor burdens, but was also present in NK cells reisolated from non-tumor bearing animals that received NK cells, suggesting that homeostatic proliferation after transfer of mature NK cells could also contribute to exhaustion. CFSElo proliferated NK cells showed the most dramatic loss of effector function (chromium release = 42% in unproliferated vs 18% in proliferated cells, p = 0.03) and transcription factor expression (Eomesodermin positive 83% in unproliferated cells vs 18% in proliferated cells, p = 0.002). Collectively, our results suggest that the success of NK cell immunotherapy is limited by an acquired dysfunction that occurs within days after homeostatic proliferation and target encounter and that may be related to the downregulation of transcription factors required for NK effector function. These findings illuminate a previously unappreciated phenomenon and explain why short-term in vitro killing assays have limited utility in predicting the in vivo behaviour of transferred NK cells. Hence, these findings suggest that transferred NK cells become dysfunctional in vivo and that novel approaches may be required in order to circumvent the described dysfunction phenotype. Disclosures: No relevant conflicts of interest to declare.