ALBERT

Description: Background: Multiple myeloma (MM) is characterized by widespread involvement of the bone marrow (BM) at diagnosis, implying a continuous (re) circulation of the MM cells in the peripheral blood and (re) entrance into the BM. The normal process of B cell homing is regulated by cytokines and receptors such as SDF-1, CXCR4, VLA4, LFA-1, VCAM-1 and ICAM-1. In order to better understand the role of homing in MM, we developed an in vivo model which allows the continuous real-time imaging of MM cells as they home and adhere to the BM, as well as quantifying the numbers of cells in the circulation. Methods: MM.1S (2 ×106/ml) were fluorescently labeled by incubation with the dialkylcarbocyanine membrane dye “DiD” (Molecular Probes) 1uM dye for 15 min at 37°C. Cells were i.v injected in Balb/c mice. Appropriate arterioles in the ear pinnae of the mice were chosen for obtaining measurements, and the fluorescence signal on the MM cells was excited as the labeled cells passed through a slit of light (from a 632 nm He:Ne laser) focused across the vessel. Cell counts were obtained every 5 min from the time of injection. MM cell homing to bone marrow vasculature of the skull was analyzed using fluorescence confocal microscopy. A small incision was made in the scalp so as to expose the underlying dorsal skull surface. The mouse was placed on a warmed microscope stage. Imaging duration was 1–3 hours per session. DiD was excited with a 635 nm diode laser. High-resolution images with cellular details were obtained through the intact mouse skull at depths of up to 250 um from the surface of the skull. Images from several depths were obtained and z-stacking was performed to merge the images. Quantitative evaluation was made by dividing the bone marrow into pre-determined quadrants (areas 1 to 4) and counting numbers of fluorescent cells per field. To demonstrate that that this new model identifies changes in homing of MM cells, we used the CXCR4 inhibitor AMD3100 and anti-VLA-4 antibody to inhibit homing to the BM. MM cells were pre-incubated with AMD3100 (50 uM overnight, Sigma, MO) or anti-VLA-4 antibody (1hr incubation with 10 ug/ml, BD Pharmingen, CA) or control PBS under the same conditions. In vivo flow cytometry and confocal imaging were then performed on control and AMD3100 treated mice. Results: The number of cells in the control group decreased dramatically (86% decrease) after 1 hour indicating homing, whereas there was only a 47% reduction in the cells at 1 hour in the AMD3100 treated cohort, (p=0.002). Similarly, we demonstrated that the number of cells present in the perivascular bone marrow niches of the skull was significantly higher in the control mice as compared to the AMD3100-treated group at 1 hour after injection. The mean cell count in the AMD3100 treated mice decreased to 38% as compared to controls, p=0.01 Likewise, the use of anti-VLA-4 antibody demonstrated significant retention of MM cells in the circulation 1 hr after injection (4% reduction in circulating cell count vs 82% for control). Conclusion: We describe a new model that detects in vivo real-time homing of MM cells from the peripheral circulation into BM niches, which can be used to study the trafficking of MM cells into and out of the BM, as well as the effect of novel agents on this dynamic process.

Description: Abstract 63 Anti-tumor effects of allogeneic hematopoietic cell transplantation have been attributed to engrafted donor lymphocytes and long-term donor engraftment is thought to be essential to cure hematological malignancies. However, following non-myeloablative hematopoietic cell transplantation (HCT), we have observed that some patients are cured of advanced hematologic malignancies despite rejection of their donor hematopoietic cell graft. This led us to hypothesize that host-versus-graft (HVG) reactions may induce host-derived anti-tumor effects. In a murine model, we established that HVG responses can indeed promote host-anti-tumor responses that are dependent on host-type CD8 T cells, host-type NKT cells, IFN-g, and initial donor engraftment (Rubio et al, Blood 2003 102:2300 and J Immunol 175:665,2005; Saito et al. manuscript submitted). Host CD8 T cells in the bone marrow secrete high levels of IFN-g even after peripheral donor chimerism has totally disappeared (Saito et al. manuscript submitted). Thus, we hypothesized that alloresponses may persist in the bone marrow, driven by the presence of alloantigens, either due to sustained engraftment of the donor cells (direct alloantigen presentation) or to presentation of alloantigens by host antigen-presenting cells (APC) (indirect alloantigen presentation) in the bone marrow. To assess this hypothesis, we used dynamic in vivo microscopy to directly visualize events in living animals. Two-photon microscopy allowed visualization of the bone marrow environment 150 mm below the skull bone surface of anesthetized mice. With these tools, we were able to track fluorescently labeled donor protein together with bone and blood vessels in real time. We established stable mixed allogeneic bone marrow chimeras using GFP+ C57BL/6 mice as donors and BALB/c mice as recipients using non-myeloablative conditioning. This system allows tracking of GFP as a model donor antigen and discrimination of donor cells from host cells engulfing donor proteins by examining both GFP staining and expression of specific major histocompatibility complex (MHC) alleles. Recipient leukocyte infusion (RLI) was performed to induce HVG reactions 7 to 10 weeks after initial allogeneic BMT in order to induce rejection of engrafted donor cells. GFP expression in the bone marrow was examined in stable mixed chimeras and in recipients of RLI. In stable mixed chimeras, clusters of GFPhigh donor cells were detected as green areas that were evenly distributed and located mainly around the blood vessels in the bone marrow. Fast-moving and rolling GFPhigh cells were observed inside the vessels. Even 11 to 22 days after RLI administration, we still observed static GFPdim cells surrounding the vessels, when donor chimerism in peripheral blood had become undetectable. When we gated out most of the RBCs on the basis of forward scatter, the majority of GFPdim cells in the bone marrow at 22 days post-RLI were Mac1+ cells, while a smaller proportion were of the B cell lineage. When we examined the origin of the Mac1+ GFPdim cells using donor and recipient MHC-specific antibodies, they were found to be recipient-derived Mac1+ cells. Our studies revealed the persistence of host-derived myeloid cells containing donor protein and residing around the bone marrow blood vessels after peripheral blood chimerism was lost. The results suggest that persistent presentation of donor-derived antigens on host APC may maintain alloreactive host CD8 T cell activation in the bone marrow, promoting the anti-tumor effect induced by HVG reactions. Engraftment of donor bone marrow cells and remaining cells after rejection. GFP+ cells are shown in green, bone in blue and blood vessels in red. GFP+ cells are visualized in vivo inside the skull bone of chimeras (A) and of chimeras that received RLI 11 days earlier (B). Arrowheads point to GFP+ cells around the blood vessels and arrows point to GFP+ cells in the blood vessels. Data from one representative mouse of two mice are shown. (A) GFP expression among Mac1+, B220+, CD4+, and CD8+ bone marrow cells in RLI recipient. (B) GFP positive leukocytes in the bone marrow are stained for H-2Db and H-2Dd. Data from representative mouse from two mice are shown. Disclosures: No relevant conflicts of interest to declare.

Description: A crucial player in immune regulation, FoxP3+ regulatory T cells (Tregs) are drawing attention for their heterogeneity and noncanonical functions. For example, specific subsets of Tregs in the adipose tissue control metabolic indices; muscle Tregs potentiate muscle repair, and lung Tregs prevent tissue damage. These studies, together with a previous finding that Tregs are enriched in the primary site for hematopoiesis, the bone marrow (BM), prompted us to examine whether there is a special Treg population which controls hematopoietic stem cells (HSCs). We showed that HSCs within the BM were frequently adjacent to distinctly activated FoxP3+ Tregs which highly expressed an HSC marker, CD150. Moreover, specific reduction of BM Tregs achieved by conditional deletion of CXCR4in Tregs, increased reactive oxygen species (ROSs) in HSCs. The reduction of BM Tregs further induced loss of HSC quiescence and increased HSC numbers in a manner inhibited by anti-oxidant treatment. Additionally, this increase in HSC numbers in mice lacking BM Tregs was reversed by transfer of CD150high BM Tregs but not of CD150low BM Tregs. These results indicate that CD150high niche-associated Tregs maintain HSC quiescence and pool size by preventing oxidative stress. We next sought to identify an effector molecule of niche Tregs which regulates HSCs. Among molecules highly expressed by niche Tregs, we focused on CD39 and CD73, cell surface ecto-enzymes which are required for generation of extracellular adenosine, because 1) CD39highCD73high cells within the BM were prevalent among CD150high Tregs and 2) HSCs highly expressed adenosine 2a receptors (A2AR). We showed that both conditional deletion of CD39 in Tregs and in vivo A2AR antagonist treatment induced loss of HSC quiescence and increased HSC pool size in a ROS-dependent manner, which is consistent with the findings in mice lacking BM Tregs. In addition, transfer of CD150high BM Tregs but not of CD150low BM Tregs reversed the increase in HSC numbers in FoxP3cre CD39flox mice. The data indicate that niche Treg-derived adenosine regulates HSCs. We further investigated the protective role of niche Tregs and adenosine in radiation injury against HSCs. Conditional deletion of CD39 in Tregs increased radiation-induced HSC apoptosis. Conversely, transfer of as few as 15,000 CD150high BM Tregs per B6 mouse (iv; day-1) rescued lethally-irradiated (9.5Gy) mice by preventing hematopoiesis failure. These observations indicate that niche Tregs protect HSCs from radiation stress. Finally, we investigated the role of niche Tregs in allogeneic (allo-) HSC transplantation. Our previous study showed that allo-hematopoietic stem and progenitor cells but not allo-Lin+ cells persisted in the BM of non-conditioned immune-competent recipients without immune suppression in a manner reversed by systemic Treg depletion1. This observation suggests that HSCs have a limited susceptibility to immune attack, as germline and embryonic stem cells are located within immune privileged sites. Because the study employed systemic Treg depletion and non-conditioned recipients, it remains unknown whether niche Tregs play a critical role in immune privilege of HSCs and in allo-HSC engraftment following conditioning. We showed here that the reduction of BM Tregs and conditional deletion of CD39 in Tregs abrogated allo-HSC persistence in non-conditioned immune-competent mice as well as allo-HSC engraftment following nonmyeloablative conditioning. Furthermore, transfer of CD150high BM Tregs but not of other Tregs (15,000 cells/recipient; day -2) significantly improved allo-HSC engraftment. This effect of niche Treg transfer is noteworthy given that 1-5 million Tregs per mouse were required in case of transfer of spleen or lymph node Tregs. These observations suggest that niche Tregs maintain immune privilege of HSCs and promote allo-HSC engraftment. In summary, our studies identify a unique niche-associated Treg subset and adenosine as regulators of HSC quiescence, numbers, stress response, engraftment, and immune privilege, further highlighting potential clinical utility of niche Treg transfer in radiation-induced hematopoiesis failure and in allo-HSC engraftment (under revision in Cell Stem Cell). 1 Fujisaki, J. et al. In vivo imaging of Treg cells providing immune privilege to the haematopoietic stem-cell niche. Nature474, 216-219, doi:10.1038/nature10160 (2011). Disclosures No relevant conflicts of interest to declare.