Search

Hematopoietic stem cell (HSC) transplantation is a lifesaving therapy for malignant and nonmalignant hematologic diseases and metabolic disorders. Although successful, hematopoietic transplantation can be hindered by inadequate stem cell number or poor engrafting efficiency. To overcome these deficits, we and others have previously reported the HSC-enhancing ability of a short-term exposure of prostaglandin E2 (PGE2); this strategy has now progressed to phase 1 clinical trials in double cord blood transplantation. To further analyze the short- and long-term effects of HSC exposure to PGE2, we followed the repopulation kinetics of PGE2-treated hematopoietic grafts through 5 serial transplantations and compared inherent long-term competitiveness in a HSC head-to-head secondary transplantation model. Treatment with PGE2 did not result in a long-term increase in HSC competitiveness, lineage bias, or enhanced proliferative potential, demonstrating that pulse exposure to PGE2 results in transient increases in HSC homing and engraftment potential.

Networking between hematopoietic stem cells (HSC) and cells of the hematopoietic niche is critical for the maintenance of stem cell renewal and function. HSC maintenance in the hematopoietic niche is considered to be the product of intimate interactions between cellular and soluble elements of the niche and stem cells. Among the cellular components of the niche participating in this function are a group of specialized bone-resident macrophages known as osteomacs (OM). Previously, we established the importance of osteoblasts (OB) in hematopoiesis and quite recently, we described the importance of OM and their interactions with OB and megakaryocytes (MK) in sustaining HSC function. We have also illustrated that CD166 is a critical functional marker of stem cell function and competence of the hematopoietic niche. Interestingly, immature OB which are CD166+ mediate the highest level of hematopoietic enhancing activity. We report here the importance of CD166 on calvarie-resident OM (identified as CD45+F4/80+ cells) and outline how these cells require cooperation from MK to increase CD166 expression and sustain HSC function. Bone resident-osteomacs, which are phenotypically similar but functionally different from bone marrow-derived macrophages, were collected by the enzymatic digestion of neonatal calvarial cells (NCC) or long bones of adult mice. Transplantation assays indicated that OM are relatively radioresistant and survive several weeks post lethal radiation. However, they eventually deplete and are replenished by progeny of donor HSC. To understand the importance of OM-OB-MK interactions in maintaining HSC function in the niche, we performed 3D cytometry on fixed and stained bone marrow sections that revealed intimate spatial interactions between OM, OB, MK and HSC. To assess changes in gene expression observed due to these interactions, we cultured NCC for 16hr in the absence or presence of MK prepared from fetal liver followed by sorting out OM from each group. These cells were then captured as single cells and sequenced to identify potential targets through which OM enhanced hematopoietic activity. Strikingly, several genes involved in the hematopoietic stem cell differentiation pathway including lmo2, fli1 and ikzf1 were upregulated in OM cultured in the presence of MK. Other genes that were upregulated were embigin and PF-4, both of which have been implicated in the maintenance of HSC function. Interestingly, OM express embigin, angiogenin and IL-18 mRNA similar to proximal osteolineage cells which we previously described as HSC regulators. To investigate changes at the translational level, we performed single cell proteomics using CyTOF. NCC were cultured for 2 days in the absence and presence of MK followed by staining for a panel of 29 surface and intracellular markers. Expression of markers such as CD166, embigin, mac-2 and STAT3 amongst others was elevated on OM cultured with MK compared to OM cultured without. These data informed our decision to focus our future investigations on CD166 and embigin. Next CD166+OM and CD166-OM were isolated by cell sorting and used in co-culture assays with OB to support the production of clonogenic cells in vitro. Only the CD166+ fraction of OM maintained hematopoietic activity similar to unsorted OM, implicating CD166 as one of the mediators of OM function. These results were validated using recombinant CD166 protein to substitute for OM function. Under these conditions, recombinant CD166 supported the hematopoietic enhancing activity of OB in the absence of OM. Recombinant Angiogenin and IL-18 were unable to augment the CD166-mediated support of hematopoiesis. Interestingly, CD166 knockout OM were unable to mediate the same hematopoietic enhancing activity observed with WT OM regardless of the presence or absence of MK in culture. In vivo transplantation studies to corroborate these findings have been initiated and are very early to yield meaningful conclusions. These data demonstrate that CD166 is one of the important mediators through which OM maintain HSC function. However, CD166-OM mediated HSC function is only maintained in conjunction with OB-MK interactions. Our data indicate the importance of crosstalk between OM, OB and MK which leads to the expression of novel mediators such as CD166 to support HSC function. Disclosures No relevant conflicts of interest to declare.

Multiple myeloma (MM) is characterized by monoclonal plasma cells that induce devastating bone disease, by markedly increasing osteoclastic bone destruction, and severely suppressing bone formation. Although the contributions of bone marrow stromal cells, immune cells, and osteoclasts to tumor growth and bone disease in MM have been extensively studied, little information is available on the role of osteocytes (Ots) in MM, which are the primary regulators of bone homeostasis and comprise 95% of bone cells. We recently showed that MM cells and Ots physically interact in vivo through the osteocytic lacunar-canalicular network. This network allows direct cell-to-cell communication and distributes Ot-secreted molecules among cells in bone and the bone marrow to regulate osteoblast and osteoclast activity. Further, apoptotic Ots recruit osteoclast precursors to specific areas of bone for localized bone resorption. In the current study we report that Ot-MM cell interactions activate bidirectional Notch signaling between Ots and MM cells which induce: 1) Ot apoptosis, that is maintained by MM-derived TNFα; 2) stimulate Ot and MM cells Rankl production, and 3) markedly increase MM cell growth. For these studies we used co-cultures of murine Ot-like MLO-A5 cells with murine and human MM cell lines or primary patient MM cells in vitro, 6wk-old female SCID mice injected intratibially with human JJN3 MM cells or saline in vivo, and a novel ex vivo calvarial organ culture system with murine 5TGM1 MM cells to provide a more realistic bone microenvironment with an intact osteocytic network. We found that the prevalence of Rankl-expressing Ots was significantly higher in JJN3-injected mice compared to controls. Similarly, increased Rankl expression in osteocytic cells was observed in co-cultures of Ots with JJN3, 5TGM1 or MM patient cells. Further, inhibition of MM-induced Ot apoptosis with DEVD significantly reduced Rankl levels as did anti-TNFα treatment. We next determined the effects of Ot apoptosis on osteoclast precursor recruitment. Conditioned media (CM) from JJN3 and MLO-A5 cells co-cultured in direct contact enhanced osteoclast precursor chemotaxis by 50% compared to CM from MLO-A5 or JJN3 cells cultured alone. This effect was inhibited by blocking Ot apoptosis with DEVD or by a combination of the Notch inhibitor GSIXX and anti-TNFα. Taken together, these results show that MM-induced Ot apoptosis increases Ot Rankl expression and potentiates Ot-mediated recruitment of osteoclast precursors. Importantly, we found that the Ot-MM Notch signaling was bidirectional. Direct contact of MM cells with Ot-like cells increased the expression of the Notch target genes Hes1/Hey1 in MM patient and 5TGM1 cells, which was abolished by the Notch inhibitor GSIXX. Moreover, expression of Rankl was increased in 5TGM1 and MM patient cells cultured in direct contact with MLO-A5 cells. Further, MM-Ot-like cell-to-cell contact specifically increased Notch receptor (R) 3 expression in patient MM cells, and Notch R1-3 as well as induced R4 in 5TGM1 cells. Finally, cell-to-cell contact between MM patient cells or 5TGM1 cells with MLO-A5 cells increased MM cell proliferation that was blocked by GSIXX. Culture of 5TGM1 cells in an ex vivo bone organ system containing authentic murine Ots also activated MM cell Notch signaling, with increased expression of Hey1, and increased the expression of Notch R3 and Rankl. All these effects were inhibited by the addition of Notch inhibitor GSIXX to the ex vivo cultures. These results indicate that bidirectional Notch signaling between MM cells and Ots increases MM tumor growth, Ots apoptosis and the expression of molecules that regulate osteoclast activity and chemotaxis. They further show that Ots can regulate tumor growth and osteoclast activity in MM and suggest that development of pharmacologic agents that selectively target NotchR3-mediated signaling in primary MM cells should decrease tumor growth and avoid the serious side-effects associated with generalized Notch inhibition. Disclosures No relevant conflicts of interest to declare.

Multiple myeloma (MM) is a plasma cell malignancy characterized by multiple lytic lesions throughout the skeleton, suggesting that trafficking of MM cells from the bone marrow (BM) and lodgment of these cells at secondary sites is important in disease progression. CD38+CD138- MM cells were previously characterized as putative MM stem cells (MMSC, Cancer Res. 2008; 68(1):190-7.). We analyzed CD38+CD138- cells contained within the MM cell line H929 and determined that a fraction of these cells (29.9%±1.4%) expresses CD166. CD166 is a member of the immunoglobulin superfamily capable of mediating both homophilic and heterophilic (CD6) interactions and has been shown to enhance metastasis and invasion in several tumors including breast cancer and melanoma. Studies from our laboratory suggest that CD38+CD138-CD166+ MM cells possess many functional properties commonly associated with MMSC including cell cycle quiescence, maintenance and propagation of daughter cells on a stromal substrate and gene expression profile. We hypothesized that CD166 promotes MM cell trafficking to the BM and is critical for disease progression. To test this hypothesis, H929-GFP myeloma cells were injected intravenously into NSG mice and GFP cells were recovered from the BM 14hr later. While only 3.3%±1.5% of total H929-GFP cells express the CD38+CD138- phenotype, the frequency of CD38+CD138- cells contained in BM-homed H929-GFP cells was significantly higher (53.4%±3.7%, n=3, p Disclosures: No relevant conflicts of interest to declare.

Bone marrow sinusoidal endothelial cells (ECs) are an integral component of the hematopoietic stromal microenvironment and have been implicated in hematopoietic stem and progenitor cell (HSPC) homeostasis, though their role in regulation of hematopoietic trafficking remains uncertain. In this study, we demonstrate that bone marrow (BM) sinusoidal ECs actively regulate HSPC retention and enforced egress out of BM quite surprisingly by serving as “gatekeepers” of HSPC within the marrow niche, a non-paradigmatic mechanism independent of traditional adhesive/chemotactic mechanisms. Treatment with G-CSF (100 µg/kg, 4 days) resulted in down-regulated expression of VE-cadherin (50±3.3% reduction) and CD31 (35±2.1%) along the gap junctions connecting ECs (CD45neg Ter119- VEGFR3+) forming BM sinuosoids, reducing cell-cell contact and allowing for increased HSPC trafficking to the peripheral blood. To validate the functional contribution of ECs in HSPC trafficking, mouse (C166 cells) or human (HUVEC) endothelial monolayers were treated with G-CSF for 24 hours in vitro. Treatment of EC monolayers with G-CSF increased dextran-FITC permeability across EC monolayers and increased migration of HSPCs across endothelium, suggesting that a component of G-CSF mobilization activity is the “opening” of EC gap junctions to allow successful HSPC migration through the EC layer. To confirm this increase in endothelial monolayer permeability in response to G-CSF treatment, we measured EC cell-cell contacts both in vitro and in vivo. Immunofluorescence analysis of HUVECs demonstrated significant VE-cadherin+ cell-cell contact in our cultured HUVEC monolayers with a significant loss of cell-cell contact after G-CSF exposure. Histological examination of femur sections from G-CSF treated mice showed a similar disruption of cell-cell contact of VEGFR+and Dil-Ac-LDL sinusoids. Previous reports have demonstrated that pharmacologic antagonism of the dipeptidyl peptidase CD26, or its gene deletion, results in reduced HSPC egress in response to G-CSF. As CD26 is highly expressed on BM ECs, and expression dynamically changes in response to G-CSF treatment, we hypothesized that sinusoidal ECs expressing CD26 may regulate sinusoidal integrity and HSPC egress. Blockade of CD26 activity, or gene deletion of CD26 in ECs, completely reversed the G-CSF-mediated reduction in EC cell-cell junctions accompanied by reduced trans-endothelial migration of HSPC. Given that CD26 is an enzyme that exerts physiologic effects by N-terminal cleavage of effector proteins, we performed a systematic protein sequence search for proteins containing putative CD26 recognition sites. Intriguingly, the neurotransmitter neuropeptide Y (NPY) can be cleaved by CD26, and the full length and cleaved versions have differing functions on NPY receptors expressed on EC. While HSPC mobilization in response to G-CSF was reduced in the absence of CD26 activity, mobilization was restored by treating mice with the cleaved NPY (NPY3-36) and was inhibited by antagonizing NPY2 and Y5 receptors on ECs, suggesting that reduced mobilization in CD26 knockout mice is caused by a lack of cleaved NPY, preventing an opening of EC cell-cell junctions and allowing for HSPC egress. If ECs are acting as active “gatekeepers” of HSPC within the marrow space, regulated by CD26 and NPY, we hypothesized HSPC egress in response to agents other than G-CSF would also depend on EC cell-cell junction opening. Interestingly, LPS administration, a mimic of bacterial infection, and AMD3100, a CXCR4 antagonist, also reduced VE-cadherin and CD31 expression along BM sinusoidal EC junctions coincident with enhanced HSPC mobilization. Our results identify an unappreciated active role for bone marrow niche ECs in maintaining HSPC within the marrow space, and describe a heretofore unknown CD26-mediated EC-neuropeptide Y axis regulating optimal HSPC trafficking. These results further refine our knowledge of hematopoietic trafficking mechanisms and identify potential new pharmaceutic targets to regulate trafficking in homeostatic and stress conditions. Disclosures: Hoggatt: Fate Therapeutics: Consultancy. Pelus:Fate Therapeutics: Consultancy.

Abstract 830 Myelopoiesis influences multiple physiological processes including skeletal remodeling. Osteoclasts (OCs) are specialized myeloid cells responsible for bone resorption. Dysregulated myeloid differentiation and OC bone lytic activity can lead not only to hematopoietic deficits, but also to skeletal diseases such as osteoporosis and osteopetrosis. Haploinsufficient loss of the Nf1 tumor suppressor gene leads to Ras mediated cytokine hypersensitivity in multiple myeloid lineages, including marrow-derived myeloid progenitor cells, mast cells, and OCs. Aside from neoplasia, patients with the genetic disorder neurofibromatosis type 1 (NF1) exhibit a spectrum of osseous defects including osteoporosis and pseudarthrosis – severe non-union fractures that respond poorly to treatment and often require amputation. Detailed investigation of the cellular and molecular mechanisms underlying NF1 pathological bone repair is needed to identify targeted therapies. In a series of adoptive bone marrow transfer experiments, we recently established that Nf1 haploinsufficiency in the hematopoietic system plays a critical role in the pathogenesis of tibial fracture non-union in the NF1 murine model. Since Nf1+/− OCs exhibit multiple gain-in-fuctions, we reasoned that haploinsufficient loss of Nf1 in myeloid OC progenitors is required for the genesis of osseous phenotypes. To test this hypothesis, we generated Nf1f/+;LysMCre+ mice harboring conditional inactivation of a single Nf1 allele in myeloid lineages. Here, we report that Nf1f/+;LysMCre+ mice exhibit increased frequency of osteoclast progenitors (CFU-M), enhanced osteoclastogenesis, and accelerated bone resorption following ovariectomy induced resorptive stress in vivo. Interestingly, recombination of the floxed Nf1 allele under CtskCre, restricted to mature osteoclasts, did not recapitulate these phenotypes in analogous experiments. These data provide direct genetic evidence that haploinsufficiency of Nf1 in myeloid progenitors, but not terminally differentiated OCs alone, is required to expand the pool of OC progenitors, thereby inducing increased bone resorption in vivo. To further investigate the contribution of Nf1 hapolinsufficient myeloid cells in a NF1 psuedarthrosis mouse model, we transplanted bone marrow (BM) cells from WT and Nf1f/+;LysMCre+ mice and monitored tibial fracture repair after stable hemapoietic reconstitution. Compared to WT BM cells, mice reconstituted with Nf1f/+;LysMCre BM cells exhibited deficient cortical bridging and significantly reduced callus bone volume fraction (BV/TV) as determined by micro-computed tomography (μCT), indicating substantial deficits in fracture repair. Histological analysis revealed fibrous infiltrates and excessive osteoclast numbers within the fracture callus. To determine the molecular mechanisms by which Nf1 status affects the proliferation and commitment of OC progenitors, we considered transcription factors that are important for early myeloid differentiation. Here, we show that Nf1+/− OC gain-in functions are associated with increased nuclear levels of phosphorylated Pu.1. These data implicate Ras as a molecular switch regulating macrophage and OC development by controlling both the expression and post-translational modification of the critical myeloid transcription factor Pu.1. In sum, our genetic and bone marrow transplantation studies further demonstrate that Nf1+/− OCs and their progenitors are the culprit hematopoietic cell lineage responsible for the pathogenesis of NF1 non-union fracture. Thus, therapies targeting the myeloid lineage, especially OCs and their progenitors, may be required for the treatment of NF1 skeletal deficits. Disclosures: No relevant conflicts of interest to declare.

Abstract 724 Hematopoeitic stem (HSC) and progenitor cells (HPC) are localized in niches contained within the bone marrow (BM) microenvironment. To facilitate acquisition of HSC/HPC for hematopoietic transplantation, donors can be treated with granulocyte-colony stimulating factor (G-CSF) to “mobilize” HSC/HPC from the BM niche to the peripheral system. Our laboratory has recently discovered that short term administration of non-steroidal anti-inflammatory drugs (NSAIDs) also mobilizes HSC/HPC and works in synergy with G-CSF. To understand the mechanisms underlying this novel therapeutic utility of NSAIDs we analyzed the BM niche post-NSAID and G-CSF treatment. Femurs of mice were analyzed after a 4 day regimen of the NSAID Meloxicam (3mg/kg, bid) and compared to control or G-CSF treated mice. Gross histological analysis showed a remarkable osteoblast (OB) “flattening” similar to that seen with G-CSF. Goldners trichrome staining revealed significantly reduced osteoid bone surfaces, with ∼3-fold increase in quiescent surfaces. Similarly, dynamic bone formation analysis using calcein/tetracycline labeling demonstrated reductions in mineral apposition rate and bone formation rate. There were no significant alterations in trabecular bone as determined by MicroCT. TRAP+ osteoclasts (OCs) were slightly elevated in both Meloxicam and G-CSF treated groups. To further assess the role of OCs, mice were mobilized with or without zoledronic acid (ZA) treatment, which inhibits OC activity. Similar to a recent report (Winkler, Blood, 2010), ZA resulted in an increase in HSC/HPC mobilization by both Meloxicam and G-CSF, suggesting that increased OC activity is not a mitigating mechanism for NSAID-mediated mobilization. Immunohistochemical (IHC) staining showed marked reductions in osteopontin (OPN), stromal-derived factor-1 (SDF-1) and N-cadherin expression by Meloxicam and G-CSF treated mice. In a separate set of experiments, OBs and mesenchymal stem cells (MSCs) were sorted by flow cytometry and gene expression assessed by Taq-Man assay. Similar to IHC analysis, Meloxicam treatment resulted in reduced SDF-1, OPN, Jagged-1, Runx2 and VCAM-1 gene expression. While the interaction of SDF-1 with its hematopoietically expressed receptor CXCR4 is a well known mediator of niche retention, both HSC and HPC were significantly mobilized by Meloxicam in CXCR4 knockout (KO) mice, suggesting that while reduced SDF-1 expression may concurrently play a role in NSAID mediated mobilization, it is not the definitive mechanism. In contrast, when OPN KO mice were mobilized with Meloxicam or G-CSF, Meloxicam unexpectedly increased mobilization of HPC only, and not HSC, while both HPC and HSC were mobilized by G-CSF. This surprising result indicates that NSAID-mediated OPN reduction is specifically responsible for the observed HSC mobilization, while HPC mobilization appears to be mediated by another mechanism(s). Recently, it has been reported that G-CSF reduces resident F4/80+ monocytes/macrophages (MOs), which normally support niche OBs and MSCs, and this reduction is at least partially responsible for niche attenuation and hematopoietic mobilization. However, in contrast to G-CSF, IHC analysis showed no reduction in F4/80+ cells after Meloxicam treatment, nor was there a reduction in CD169+ BM macrophages as assessed by flow cytometry. Therefore, while NSAIDs and G-CSF attenuate the niche microenvironment similarly, these agents function through independent mechanisms, perhaps explaining the synergistic mobilization seen by the two. In conclusion, NSAID treatment results in significant attenuation of the BM niche, including the niche components SDF-1, OPN, and N-cadherin. NSAID-mediated mobilization is independent of reduced SDF-1/CXCR4 signaling, as CXCR4 KO mice can be mobilized by NSAID. OPN KO specifically blocks the NSAID mobilization of HSC, but not HPC, suggesting differing mechanisms, or possibly multiple niche locations for these two populations; an intriguing finding suggesting that other mobilization strategies may be able to specifically target HSC or HPC. While G-CSF treatment results in reduced MOs, NSAIDs do not alter BM levels of these supportive cells. These results define not only a novel strategy for mobilization of HSC/HPC, but also suggest a possible strategy for targeted niche attenuation for other therapeutic applications, such as reduced conditioning regimens. Disclosures: Hoggatt: Fate Therapeutics: Consultancy. Pelus:Fate Therapeutics: Consultancy.