Search

Recently , we showed that hematopoietic stem cells (HSCs) fluctuate in a dynamic trajectory for self-renewal and lineage commitment, which regulates HSC heterogeneity (Cell Rep 2018). Although the significance of heterogeneity is difficult to interpret, it might enable HSCs to maintain a dynamic equilibrium of homeostatic hematopoiesis and to adapt promptly to stress-induced hematopoiesis based on physiological demands. We hypothesized that similar to HSCs, leukemia stem cells (LSCs) are heterogeneous with active fluctuation, which contributes to treatment resistance. Hence, this study aimed to show the fluctuation of LSCs and determine the underlying molecular mechanisms. We previously reported that the endothelial-cell selective adhesion molecule (ESAM) is useful as a life-long marker for HSCs in humans and mice (Blood 2009, Exp Hematol 2013). ESAM expression levels precisely indicated the activation status of HSCs (J Immunol 2012). Furthermore, approximately two-thirds of patients with acute myelogenous leukemia (AML) were positive for ESAM expression on leukemia cells in a heterogeneous manner. (Exp Hematol 2013). We exploited human AML lines as LSC-like cells and monitored expression levels of ESAM. Flow cytometry analyses revealed wide-ranging ESAM expression levels on KG1a and CMK lines. In KG1a cells, ESAM-high cells were concentrated in the CD34+CD38- fraction with high growth capability, whereas ESAM-low cells were less proliferative. Notably, when sorted into ESAM-high or ESAM-low subgroups, both groups reconstituted the parental heterogeneous distribution of ESAM levels in culture. Furthermore, single self-renewing LSC-like clones also reproduced the same heterogeneous population regardless of their original ESAM levels. These results suggested that LSC-like clones developed the heterogeneous AML cell population by fluctuating, which could be monitored with ESAM levels. The transcriptomes of ESAM-high and ESAM-low LSC-like cells in KG1a line were compared to obtain comprehensive information of key molecules involved in fluctuation and heterogeneity. We found that expression of T-cell immunoglobulin and mucin domain 3, reported as a reliable LSC marker for human AML (Cell Stem Cell 2010), was variable in parallel with ESAM. In addition, network analysis demonstrated that "genes related to hematopoietic progenitor cells" were activated whereas "genes related to blood cell maturation" were downregulated in ESAM-high cells. These results suggested that ESAM-high LSC-like cells were more identical to the conventional concept of LSCs than ESAM-low cells. However, expression levels of BAALC, SPP1, and ITGB7 were elevated in ESAM-low cells. BAALC encodes "brain and acute leukemia cytoplasmic (BAALC)" protein, which inhibits drug-induced apoptosis in KG1a cells (Hematology 2012). SPP1 encodes "secreted phosphoprotein 1 (SPP-1), also known as osteopontin, which induces chemo-resistance in AML cells (Life sci 2017). Furthermore, a recent paper reported that the active form of integrin □7 can be a therapeutic target of myeloma stem cells (Nat Med 2018). Thus, we infer that fluctuating to the ESAM-low stage may contribute to the refractory nature of LSCs in AML. We performed an up-streaming analysis on the transcriptome data to determine the signaling pathways involved in the fluctuation. The tumor necrosis factor alpha (TNFα) and the transforming growth factor beta (TGFβ) signaling pathways were found to affect ESAM expression levels. When TNFα or TGFβ1 was added to culture medium, the proliferation of KG1a cells was suppressed in a dose-dependent manner. Furthermore, addition of TNFα or TGFβ1 significantly suppressed ESAM expression levels on a subset of KG1a cells, which enhanced their heterogeneity. When cultured in serum-free cytokine-free medium, KG1a cells became uniform in terms of ESAM expression levels, with the ESAM-low population disappearing . Intriguingly, we found that KG1a cells produced TNFα and TGFβ1 autonomously, which induced the cell heterogeneity in an autocrine/paracrine manner. Surprisingly, blocking these signaling pathways by receptor inhibitors induced apoptosis in KG1a cells. These results suggest that autocrine/paracrine cytokine signaling plays important roles in promoting the variable and heterogeneous nature of LSCs, the inhibition of which might be a strategy for therapeutic intervention in refractory AML. Disclosures Yokota: Bristol-Myers Squibb: Research Funding; Pfizer Inc.: Research Funding; CHUGAI PHARMACEUTICAL CO., LTD.: Research Funding; MSD K.K.: Research Funding; Celgene: Research Funding. Ichii:Celgene K.K.: Speakers Bureau; Kowa Pharmaceutical Co.,LTD.: Speakers Bureau; Novartis Pharma K.K.: Speakers Bureau. Shibayama:Jansen Pharmaceutical K.K: Honoraria; Celgene K.K.: Honoraria, Research Funding; Takeda Pharmaceutical Co.,LTD.: Honoraria, Research Funding; Ono Pharmaceutical Co.,LTD: Honoraria, Research Funding; Fujimoto Pharmaceutical: Honoraria, Research Funding; Bristol-Meyer Squibb K.K: Honoraria, Research Funding; Mundipharma K.K.: Honoraria, Research Funding; Novartis Pharma K.K.: Honoraria, Research Funding. Kanakura:Alexion Pharmaceuticals, Inc.: Consultancy, Honoraria, Research Funding.

Chronic myelogenous leukemia (CML) is a clonal myeloproliferative disorder caused by hematopoietic stem cells expressing the BCR-ABL fusion oncoprotein, which constitutively activates multiple signal transduction pathways such as mitogen-activated protein kinase, phosphatidylinositol 3-kinase/Akt, and Janus kinase/signal transducer and activator of transcription (JAK/STAT). Although tyrosine kinase inhibitor (TKI) therapy results in dramatic clinical success, studies have shown that TKIs are unable to eradicate leukemic stem cells (LSCs). Several key signaling molecules and pathways have been proposed to regulate the survival of CML LSCs in the presence of TKI; however, the details remain unclear. It is necessary to elucidate the mechanisms that maintain LSCs to better understand the pathogenesis of CML and develop new treatment approaches. The family of signal-transducing adaptor proteins (STAPs), which includes STAP-1 and STAP-2, has been implicated in various intracellular signaling pathways. In 2003, we cloned STAP-2 as a c-fms interacting protein and reported that STAP-2 binds to BCR-ABL and enhances activity, leading to the activation of downstream molecules such as ERK, STAT5, BCL-xL, and BCL2. STAP-1 was cloned as a c-kit interacting protein from a hematopoietic stem cell library, but it is unknown whether STAP-1 plays a role in CML. Given the structural homology between STAP-1 and STAP-2 and the hematopoietic expression of STAP-1, we hypothesized that STAP-1 might contribute to the leukemogenesis of CML. A STAP-1-deficient (KO) CML mouse model was developed. To generate this model, lineage (Lin)− Sca-1+ c-Kithigh (LSK) fraction isolated from bone marrow (BM) cells was infected with a retrovirus carrying BCR-ABL1 and GFP and subsequently transplanted into congeneric recipients. STAP-1 KO CML mice showed significantly longer survival than WT CML mice and displayed less severe splenomegaly and lung hemorrhages compared with WT mice. In recipient BM, absolute numbers of STAP-1 KO LSCs (GFP+ LSK cells) were significantly lower than WT LSCs. In the colony-forming assay, STAP-1 KO LSCs generated fewer colonies compared to WT LSCs. Using flow cytometric analysis, we found that STAP-1 KO LSCs had a higher apoptotic rate than WT LSCs. These findings suggest that the suppression of apoptosis induced by STAP-1 mediates longer survival of LSCs. To further understand the effects of STAP-1, we performed a gene expression analysis using RNA-sequence (RNA-seq) and compared WT and STAP-1 KO CML LSCs. When canonical pathways were analyzed with Ingenuity Pathway Analysis, various pathways associated with inflammatory cytokines were observed to be regulated in STAP-1 KO CML LSCs. Changes in mRNA expression, including that of SOS1, SOS2, FOXO3, FASLG, NFKB2, and BCL-xL, indicated that the PTEN signaling pathway, known to play a tumor suppressive role in CML, was significantly activated by STAP-1 KO (p=1.096E-3, activation Z-score=2.611). The pathway related to JAK/STAT signaling was also affected (p=2.04E-5, activation Z-score=-3.286). Downstream genes in the JAK/STAT signaling pathway, such as STAT5B and BCL-xL, were downregulated more than 2-fold in STAP-1 KO LSCs, suggesting that the deletion of STAP-1 inhibits the expression of STAT5-targeted anti-apoptotic protein and induced apoptosis of CML LSCs. To confirm the results of the RNA-seq experiment, an intracellular flow cytometric assay with CML Lin− cells was conducted. The frequency of cells positive for phosphorylated STAT5 was reduced for STAP-1 KO compared with that for WT. Quantitative PCR with CML LSCs confirmed the downregulation of BCL2 and BCL-xL, which are STAT5-targeted anti-apoptotic genes, in STAP-1 KO CML LSCs. In conclusion, we show that STAP-1 plays a crucial role in the maintenance of CML LSCs using a murine model of CML. STAP-1 deficiency results in the reduction of phosphorylated STAT5, downregulation of anti-apoptotic genes BCL-2 and BCL-xL, and induced apoptosis of CML LSCs. These findings suggest that STAP-1 and related signaling pathways could be potential therapeutic targets for CML LSCs. Disclosures Ichii: Celgene K.K.: Speakers Bureau; Kowa Pharmaceutical Co.,LTD.: Speakers Bureau; Novartis Pharma K.K.: Speakers Bureau. Shibayama:Fujimoto Pharmaceutical: Honoraria, Research Funding; Takeda Pharmaceutical Co.,LTD.: Honoraria, Research Funding; Celgene K.K.: Honoraria, Research Funding; Jansen Pharmaceutical K.K: Honoraria; Ono Pharmaceutical Co.,LTD: Honoraria, Research Funding; Novartis Pharma K.K.: Honoraria, Research Funding; Mundipharma K.K.: Honoraria, Research Funding; Bristol-Meyer Squibb K.K: Honoraria, Research Funding. Oritani:Novartis Pharma: Speakers Bureau. Kanakura:Alexion Pharmaceuticals, Inc.: Consultancy, Honoraria, Research Funding.

Background: Adult T-cell leukemia/lymphoma (ATLL) is a mature T-cell neoplasm, linked to the human T-cell lymphotropic virus, HTLV-1. Patients with ATLL are often at the risk of opportunistic infections. Some studies suggested that ATLL cells originate from HTLV-1-infected regulatory T cells (Tregs). It could be possible that this immunocompromised state is caused by the function of ATLL cells having similar phenotypes with Tregs. In this study, we examined the expression of immunosuppressive molecules associated with Tregs in ATLL cells, and analyzed their roles in the function of ATLL cells. Methods: The protocol of this study was approved by the Investigational Review Board of Osaka University Hospital. Peripheral blood mononuclear cells (PBMCs) were collected from 10 asymptomatic HTLV-1 carriers and 22 ATLL patients (1 with smoldering type, 5 with chronic type, 2 with lymphoma type, and 14 with acute type) after getting informed consent. PBMCs from 3 ATLL patients were separated into CD4+ CD7- CADM1+ATLL cells and adjacent CD4+CD7+ CADM1-normal T cells using Fluorescence-activated Cell Sorter (FACS), and cells in each fraction were subjected to total RNA sequencing experiments. Based on the results, we examined the expression patterns of CD39 and CD73 in HTLV-1 carriers or each type of ATLL patients, and also analyzed the immune functions of these molecules in ATLL tumor cells. Results: We compared whole transcriptome of ATLL cells and normal CD4+cells. Bioinformatic analyses showed that many genes associated with immunosuppressive functions were elevated or downregulated in ATLL cells. Among these genes we focused on CD39, CD73 and CD26, because they have recently been reported to be strongly associated with the functions of Tregs. CD39, expressed on normal Tregs, and extrinsic CD73 have immunosuppressive potential by catalyzing adenosine from extracellular ATP, and CD26 has opposite potential by resolving adenosine, which have a strong anti-inflammatory function and plays major role in Treg-mediated immunosuppression. We found that all of 4 ATLL cell lines (MJ, MT1, MT2, MT4) expressed CD39, but not CD73 just as human effector Tregs. Tumor cells from 12 acute ATLL patients (86%) and 2 chronic ATLL patients (40%) expressed CD39, but the expressions of CD73 were various. Also in asymptomatic carriers, we could detect CD39 and/or CD73 positive in CD7- CADM1+ abnormal fraction of CD4+cells. On the other hand, CD26, normally expressed on human CD4+Th cells other than effector Tregs, was negative in ATLL cell lines and primary ATLL cells except for cells in abnormal fraction of one asymptomatic carrier. CD39 negative cases in chronic/smoldering type tended to show slower disease progression after the blood collection. Next, the role of CD39 and/or CD73 in ATLL cells was assessed in vitro and in vivo. As expected, CD39+ ATLL cells converted significantly more extracellular ATP than CD39- ATLL cells, and mass spectrometry analysis of AMP/adenosine concentration identified the AMPase activity of CD73+ ATLL cells. Furthermore, we established CD39 knockout (KO) cells from ATL cell-line MJ using CRISPR/Cas9 system, and performed in vitro suppression assays for assessment of immunosuppressive function. Although wild type MJ suppressed the growth of normal CD4+ and CD8+ T cells, KO MJ did little. Next, we analyzed the role of CD39 in the progression of tumor cells in vivo. We transplanted mouse T-cell lymphoma cell-line EG7-OVA artificially expressing CD39 or mock into mice subcutaneously. The coinjection of immunoadjuvant poly(I:C) significantly suppressed the tumor growth of mock cells, but the tumor sizes of CD39 expressing cells were almost the same as those of mock cells without poly(I:C) injection (Figure). Conclusion: In this study, we reported that most of ATLL cells in acute type patients express CD39+ CD26- just as Tregs, and that CD39- KO of ATLL cell line cancelled its immunosuppressive effects, and forcibly expressed CD39 on tumor cells rejected the anti-tumor immunity in vivo. From these data, we clarified the pathological mechanism of immunosuppressive function in ATLL cells, and also showed that CD39 expression could be used as a prognostic clue and be a new therapeutic target of ATLL. Disclosures Ezoe: TAIHO Phamaceutical Co., Ltd.: Research Funding. Yokota:Celgene: Research Funding; Bristol-Myers Squibb: Research Funding; Pfizer Inc.: Research Funding; CHUGAI PHARMACEUTICAL CO., LTD.: Research Funding; MSD K.K.: Research Funding. Ichii:Novartis Pharma K.K.: Speakers Bureau; Kowa Pharmaceutical Co.,LTD.: Speakers Bureau; Celgene K.K.: Speakers Bureau. Shibayama:Novartis Pharma K.K.: Honoraria, Research Funding; Celgene K.K.: Honoraria, Research Funding; Takeda Pharmaceutical Co.,LTD.: Honoraria, Research Funding; Fujimoto Pharmaceutical: Honoraria, Research Funding; Jansen Pharmaceutical K.K: Honoraria; Ono Pharmaceutical Co.,LTD: Honoraria, Research Funding; Mundipharma K.K.: Honoraria, Research Funding; Bristol-Meyer Squibb K.K: Honoraria, Research Funding. Oritani:Novartis Pharma: Speakers Bureau. Kanakura:Alexion Pharmaceuticals, Inc.: Consultancy, Honoraria, Research Funding.

Endothelial cell-selective adhesion molecule (ESAM) is lifelong marker of hematopoietic stem cells (HSCs) (Yokota Blood 2009) and endothelial cells (ECs) (Hirarta JBC 2001). Although we previously elucidated the functional importance of ESAM on HSCs in stress-induced hematopoiesis in adults (Sudo J Immunol 2012; PLoS One 2016), how ESAM affect hematopoietic development during fetal life remained unknown. To address this issue, we analyzed the fetuses of conventional ESAM knockout mice. Approximately half of the ESAM-null fetuses exhibited severe anemia and died after mid-gestation. RNA-sequence analyses exhibited downregulation of adult-type globin genes and Alas2, which encodes the first, rate-limiting enzyme in the heme biosynthesis pathway in the ESAM-null fetal liver (FL). Culture and transplantation experiments demonstrated that these abnormalities were attributable to the malfunctioning of ESAM-null HSCs. In addition, cross-linking ESAM with an anti-ESAM antibody influenced the transcription of hemoglobin synthesis-related genes in HSCs, indicating the significance of ESAMs on HSCs for the development of definitive hematopoiesis. However, it was unclear whether ESAM expression on only HSCs was critical or if ESAM expression on ECs was also involved. To examine this, we generated ESAM-flox mice using the innovative CLICK (CRISPR with lssDNA inducing conditional knockout alleles) method that was recently developed by our group (Miyasaka BMC Genomics 2018), and analyzed the results. We first tried to establish endothelial lineage-specific ESAM-conditional knockout (ESAM-cKO) mice by crossing ESAM-flox mice with Cre-ERT2-expressing mice under the control of the VE-cadherin gene promoter (Okabe Cell 2014). After peritoneally administering tamoxifen to pregnant females on embryonic day (E) 12.5, 14.5, or 15.5, the FLs were examined on E17.5. Analyses of these fetuses showed that, as observed in conventional ESAM-null mice, severe anemia occurred in approximately half of the Cdh5-BAC-CreERT2 ESAMflox/flox fetuses. However, we noticed that ESAM expression was almost completely deleted not only on ECs but also on HSCs of this model even after a single tamoxifen injection on E15.5. Next, we made use of another ESAM-cKO model obtained by crossing ESAM-flox mice with Cre recombinase-expressing mice under the control of the Vav gene promoter (Georgiades Genesis 2002). Although this Vav-Cre model has been widely used as a hematopoietic cell-specific gene knockout system, the EC lineage is affected by mouse-to-mouse variation (Joseph Cell Stem Cell 2013). Indeed, we observed that ESAM expression on FL ECs was suppressed with varying efficiency in each mouse, whereas ESAM expression on lineage negative Sca1+c-kitHi (LSK) CD48- cells was absent. Notably, the absolute number of LSK CD48- cells in FLs was positively correlated with the remaining levels of ESAM expression on ECs. Furthermore, the anemic phenotype was less severe and fetal mortality was less evident in Vav-Cre-induced ESAM-cKO fetuses. These results indicated that ESAM on ECs also play an important role in the normal development of definitive HSCs and erythrocytes in the FL. These findings were unexpected because we previously observed that the contribution of ESAM on non-hematopoietic cells to HSC maintenance was subtle in the adult bone marrow (BM) of transplantation-induced chimeric mice (Sudo PLoS One 2016). However, this contradiction might be explained by the difference between the FL and adult BM. Various types of cells have been reported to serve as the so-called "HSC-niche" in the adult BM, whereas most HSCs are located adjacent to sinusoidal ECs in the FL. In conclusion, our data strongly suggest that ESAM expression on ECs, as well as HSCs, plays an important role in the development of definitive hematopoiesis. ESAM is known to mediate cell-cell interaction through homophilic binding. The results showed that ESAM on ECs in the developing liver compensated for ESAM deficiency on HSCs to ameliorate the lethal anemic phenotype. This indicates that ESAM might interact with undetermined molecules on developing HSCs, and these molecules might have a functional overlap with ESAM. Disclosures Yokota: CHUGAI PHARMACEUTICAL CO., LTD.: Research Funding; Pfizer Inc.: Research Funding; Bristol-Myers Squibb: Research Funding; MSD K.K.: Research Funding; Celgene: Research Funding. Ichii:Celgene K.K.: Speakers Bureau; Kowa Pharmaceutical Co.,LTD.: Speakers Bureau; Novartis Pharma K.K.: Speakers Bureau. Ezoe:TAIHO Phamaceutical Co., Ltd.: Research Funding. Shibayama:Takeda Pharmaceutical Co.,LTD.: Honoraria, Research Funding; Fujimoto Pharmaceutical: Honoraria, Research Funding; Bristol-Meyer Squibb K.K: Honoraria, Research Funding; Ono Pharmaceutical Co.,LTD: Honoraria, Research Funding; Novartis Pharma K.K.: Honoraria, Research Funding; Jansen Pharmaceutical K.K: Honoraria; Mundipharma K.K.: Honoraria, Research Funding; Celgene K.K.: Honoraria, Research Funding. Oritani:Novartis Pharma: Speakers Bureau. Kanakura:Alexion Pharmaceuticals, Inc.: Consultancy, Honoraria, Research Funding.

Graft-versus-host-disease (GVHD) after allogeneic hematopoietic stem cell transplantation (allo-HSCT) remains the most frequent complication and one of the major causes of non-relapse mortality. However, its pathogenesis has not yet been fully understood. We cloned signal-transducing adaptor protein (STAP)-2 as a c-fms binding protein from a fetal liver library in 2003. The family that contains STAP-1 and STAP-2 has a pleckstrin homology (PH) and Src-homology 2 (SH2)-like domains, suggesting that this adapter protein functions as an immune and inflammatory regulator. Indeed, STAP-2 regulates adhesion and chemotaxis in T cells (Sekine et al., J Immunol. 2009). In this study, we aimed to elucidate the roles of STAP family in GVHD. First, we examined the expression of STAP-1 and STAP-2 mRNA in various human hematopoietic subsets, including CD34+ CD38- hematopoietic stem cells (HSCs), CD34+ CD38+ hematopoietic progenitor cells (HPCs), CD19+ CD27- naïve B cells, CD19+ CD27+ memory B cells, CD3+ CD4+ helper T-cells, and CD3+ CD8+ cytotoxic T lymphocytes, using real-time PCR. As a result, STAP-1 and STAP-2 were expressed in lymphoid cells, as well as HSCs and HPCs. STAP-2 mRNA was highly expressed in T cells. Next, to investigate the role of STAPs in GVHD, we made an experimental murine model. To study the pathogenesis of immune reconstitution and tolerance after allo-HSCT, lethally irradiated BALB/c mice were injected with T and B cell-depleted bone marrow cells (5×106 cells) derived from syngeneic BALB/c or allogeneic C57BL/6 mice on day 0. Co-transplantation of splenocytes was not adapted in this model. Survival and clinical degree of GVHD were assessed by a scoring system that sums changes in 5 clinical parameters: body weight (BW) loss, posture, activity, fur texture, and skin integrity. Recipients transplanted from allogenic wild type (WT) C57BL/6 donor survived without suffering from severe GVHD symptoms, owing to development of immune tolerance against allogeneic antigens. However, compared to syngeneic transplanted mice, these recipients started to show gradual BW loss and GVHD score was increased approximately 28 days after allo-HSCT, indicating the existence of an allogeneic immune reaction. To evaluate the role of STAPs in GVHD, we generated transgenic mice (Tg) that overexpress STAP under the control of Em enhancer and Lck proximal promoter. The promoter could drive expression of the inserted cDNA in lymphoid lineage cells from the common lymphoid progenitor (CLP) stage. When STAP-2 Tg marrow was used as a donor source, we found that the overall survival of STAP-2 Tg recipients was significantly lower than that of WT recipients (22.2% and 91.7%, respectively; p In this study, we show that STAPs in reconstituted lymphocytes after allo-HSCT regulate the pathogenesis of GVHD. Our results suggest that STAP activation in lymphocytes during immune reconstitution accelerates gut and thymic GVHD. Severe thymic damage induced by STAP overexpression might contribute to impairment of immune tolerance such as a decreased number of Treg cells as well as dysfunction of thymic negative selection of host-reactive T cells after allo-HSCT, which is involved in persistence of GVHD. Future study should further elucidate the detailed molecular mechanisms involved. Disclosures Ichii: Celgene K.K.: Speakers Bureau; Kowa Pharmaceutical Co.,LTD.: Speakers Bureau; Novartis Pharma K.K.: Speakers Bureau. Shibayama:Takeda Pharmaceutical Co.,LTD.: Honoraria, Research Funding; Mundipharma K.K.: Honoraria, Research Funding; Jansen Pharmaceutical K.K: Honoraria; Ono Pharmaceutical Co.,LTD: Honoraria, Research Funding; Fujimoto Pharmaceutical: Honoraria, Research Funding; Novartis Pharma K.K.: Honoraria, Research Funding; Bristol-Meyer Squibb K.K: Honoraria, Research Funding; Celgene K.K.: Honoraria, Research Funding. Oritani:Novartis Pharma: Speakers Bureau. Kanakura:Alexion Pharmaceuticals, Inc.: Consultancy, Honoraria, Research Funding.

Murine hematopoietic stem cells (HSCs) can be isolated with high efficiency as Lineage- Sca-1+ c-kitHigh (LSK) CD34-/Low CD150+ CD48- cells. In humans, however, the same method is not useful because of critical differences between murine and human HSC phenotypes. Such discrepancy has hampered the translation of findings in mice into a human preclinical or clinical context. Therefore, the identification of common HSC antigens between the two species would be a significant advance with respect to translational studies of HSC biology. We previously identified endothelial cell-selective adhesion molecule (ESAM) as a novel maker for HSCs in mice (Blood, 2009). We also found that ESAM is functionally important for murine HSCs to reconstitute hematopoiesis after 5-FU treatment (J Immunol, 2012). In the present study, we have extended our research of ESAM to human HSCs and leukemia. We first examined whether ESAM expression showed potential as a marker of human HSCs. In addition to adult BM, the majority of CD34+ CD38- cells in cord blood (CB) and G-CSF mobilized peripheral blood expressed ESAM. The addition of anti-CD90 and CD45RA antibodies divides the adult BM CD34+ CD38- fraction into three subpopulations, namely HSCs, multipotent progenitors (MPPs), and multi-lymphoid progenitors (MLPs). We found that HSCs expressed high levels of ESAM whereas MPPs expressed lower levels and many MLPs lost ESAM expression. Functional assessment for ESAM-/Low and ESAMHigh cells in the CD34+ CD38- fractionconfirmed that high ESAM expression distinguishes progenitors that are more primitive and multipotent. We also identified a subset of CD34+ CD38- cells in adult BM and CB that expressed extremely high levels of ESAM, namely ESAMBright cells. Gene expression profiles of the CD34+ CD38- ESAMHigh and CD34+ CD38- ESAMBright populations showed that the former cells expressed HSC-related genes whereas the latter showed more endothelial-related profiles. Indeed, the CD34+ CD38- ESAMBright cells produced CD31+ endothelial cells, but not CD45+ hematopoietic cells, in co-culture with MS5 stromal cells. These results suggest that the CD34+ CD38- fraction, which is conventionally considered the human HSC fraction, also contains a substantial number of non-hematopoietic progenitors. Thus, the inclusion of ESAM provides a more accurate estimation of HSC numbers. Since some of HSC-related antigens are useful for determining leukemia lineage and have utility as prognostic indicators, we determined whether ESAM might also be a valuable addition to this antigen panel. First, we examined human leukemia cell lines. Tested myeloid leukemia lines including KG-1a, HL60, THP1, U937 and Kasumi were uniformly negative for ESAM expression. Jurkat and MOLT4, lymphoid lineage lines were also negative. On the other hand, HEL, an erythroid leukemia cell line, and CMK, a megakaryocytic leukemia cell line, exhibited high expression of ESAM. Additionally, K562 cells, which originated from CML that subsequently transformed into acute erythro-leukemia, also express ESAM. We then evaluated ESAM expression on primary acute leukemia cells, which were isolated from patients upon diagnosis. Interestingly, while all of ALL cases were virtually negative for ESAM, more than half of AML cases were ESAM-positive. Notably, the ESAM expression pattern on AML cases substantially differs even in the same FAB classification. We inferred that AML cells might change their ESAM expression levels according to cell intrinsic features and/or the surrounding environment in vivo. Therefore, we inoculated ESAM- KG-1a cells into NOD/SCID mice and harvested reconstituted KG-1a (rKG-1a) cells after the inoculation. They were then cultured in vitro and inoculated again into NOD/SCID mice. FACS analyses revealed that, although parental KG-1a cells were ESAM-negative, rKG-1a cells expressed a substantial amount of ESAM. Notably, rKG-1a cells were more aggressive and killed the recipient mice in a shorter period. This observation indicates that leukemia cells change their surface phenotype according to the environment, and that ESAM expression may be related to the acquisition of a more aggressive phenotype. In conclusion, we demonstrate that ESAM is a reliable marker of HSCs in humans as well as in mice. Additionally, ESAM is expressed on some of human acute leukemia cells and might be useful for lineage determination and as prognostic indicator. Disclosures Yokota: SHIONOGI & CO., LTD.: Research Funding. Kanakura:Alexion Pharma: Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau.

Background: Lifelong hematopoiesis is maintained by cell differentiation in which signaling pathways and transcription factors coordinately induce step-wise maturation of hematopoietic stem cells (HSCs) toward downstream effector cells. In addition, the organization of chromatin structure that creates accessible sites of target genes is also essential so as to ensure temporally and spatially adequate control of internal gene expression. Murine HSCs can be isolated with high efficiency using surface molecules including lineage-related markers, c-Kit, Sca-1, Flt3 and SLAM family proteins. However, even the highly enriched HSC fraction is still heterogeneous regarding differentiation potential, and how the HSC diversity reflects the heterogeneity of intrinsic gene-expression in HSCs is as-yet-unknown. We previously identified Special AT-rich Sequence Binding protein 1 (SATB1), a global chromatin regulator, as a lymphoid-related gene in the HSC differentiation (Satoh and Yokota et al. Immunity 2013). Indeed, SATB1 overexpression strongly enhanced both T and B lymphopoietic potential of murine HSCs whereas SATB1 deficiency caused malfunctions of HSCs in the lymphopoietic activity. Furthermore, another report showed that SATB1-deficient HSCs were less quiescent in transplanted recipients and more prone to differentiate preferentially to myeloid-erythroid lineages (Will et al. Nat Immunol 2013). These results suggested that SATB1 is likely indispensable not only for the lymphopoietic potential but also for the integrity of HSCs. Here, to better understand the mechanism how SATB1 influences homeostatic HSC functions in adult bone marrow (BM), we have developed a new mouse model in which SATB1 expression can be precisely monitored along the HSC differentiation. Methods: The Tomato gene, coding a red fluorescent protein, was knock-in to the coding region of endogenous Satb1 gene. The heterozygous SATB1/Tomato knock-in mice in which one Satb1 allele was replaced with the Tomato were used to sort HSCs in adult BM. The sorted cells were evaluated for the differentiation potential with methylcellulose colony assays and co-cultures with MS5 stromal cells. Further, the long-term reconstitution ability was evaluated by transplantation to lethally irradiated mice. To obtain transcriptome information, total RNA was isolated from SATB1/Tomato- and SATB1/Tomato+ HSCs, and then next-generation sequencing was performed. The data were analyzed with the Ingenuity Pathway Analysis software. Results: We defined Lin- Sca1+ c-KitHi (LSK) CD150+ Flt3- cells as HSCs, especially adopting FLT3- to exclude FLT3+ lymphoid-primed multipotent progenitors from our functional analyses. We found that the LSK CD150+ Flt3- fraction contains substantial number of SATB1/Tomato+ cells. While both SATB1/Tomato- and SATB1/Tomato+ HSCs produced numerous CFU-Mix and CFU-GM/G/M colonies, the latter were less potent to produce BFU-E. In co-culture with MS5 stromal cells that support B and myeloid lineages, the output of B lineage cells from SATB1+ HSCs was more robust than that of SATB1- HSCs. Upon transplantation, enhanced B-lineage engraftment was observed in the SATB1+ HSC-transplanted recipients. Interestingly, while the two types of HSCs showed obvious difference in the differentiation potential toward lymphoid or myeloid lineage, both HSCs reconstituted the LSK CD150+ Flt3- fraction that similarly contained SATB1/Tomato- and SATB1/Tomato+ cells. With the RNA-sequencing data of SATB1- and SATB1+ HSCs, biological pathway analyses revealed that the "Hematological System Development and Function" pathway was remarkably up-regulated in the SATB1+ HSCs. Among subcategories of the "Hematological System Development and Function" pathway, the "quantity of lymphocytes" pathway was increased whereas "quantity of myeloid cells" and "quantity of granulocytes" pathways were decreased. Conclusion: We have developed a new mouse system that can be used to identify and isolate viable lymphoid-biased HSCs in the most primitive hematopoietic cell fraction of adult BM. While the SATB1- and SATB1+ HSCs differ genetically and functionally, both subtypes have displayed a self-renewal activity with mutual interconversion in transplanted recipients. These findings suggest that functional heterogeneity and variability within the HSC population is, at least in part, a manifestation of SATB1 expression. Disclosures Yokota: SHIONOGI & CO., LTD.: Research Funding.

Understanding lineage specific markers contributes to investigation into lineage commitment processes in hematopoiesis. Particularly in the human study, information about hematopoietic lineage divergence is essential to refine hematopoietic lineage tree. Lineage markers are also potentially useful for therapeutic target, such as CD20 in B-cell lymphoma, and CD33 in acute myeloid leukemia. We have recently reported that special AT-rich sequence-binding protein 1 (SATB1), a global chromatin organizer, promotes lymphocyte production from hematopoietic stem cells (HSCs) (Immunity 38;1105, 2013). Expression level of SATB1 increases with early lymphoid differentiation, whereas it is shut off in committed myeloid progenitors. To search a novel cell surface molecule that marks the point of branching lineage along early myeloid and lymphoid differentiation, we performed microarray analyses comparing SATB1-overexpressed HSCs with mock-transduced HSCs. The results drew our attention to membrane-spanning 4-domains, subfamily A, member 3 (MS4A3). MS4A3, also called hematopoietic cell-specific transmembrane 4 (HTm4), is a member of the MS4A family. CD20, encoded by MS4A1 gene, belongs to the same family. We observed that expression level of MS4A3 in SATB1-overexpressed HSCs was decreased almost one tenth of that of mock HSCs. To confirm the relationship of SATB1 and MS4A3 in human hematopoietic cells, we first used chronic myeloid leukemia cell line K562, which was found to clearly express MS4A3 on their cell surface. While SATB1 expression was undetectable in original K562 cells, the exogenous expression of SATB1 significantly reduced their MS4A3 expression level, suggesting that SATB1 negatively regulates MS4A3 expression in human cells. Next, we analyzed MS4A3 expression pattern in primary human hematopoietic stem/progenitor cells. Bone marrow (BM) cells were obtained from healthy donors or patients with acute myeloid leukemia. The Institutional Review Board of Osaka University School of Medicine approved all of protocols, and written informed consents were obtained from all participants. Mononuclear cells were separated from the BM samples by density gradient centrifugation, and subsequently applied to cell sorting for Lineage marker-negative (Lin-) CD34+ CD38- HSCs, Lin- CD34+ CD38+ IL-3 receptor α (IL-3Rα)+ CD45RA- common myeloid progenitors (CMPs), Lin- CD34+ CD38+ IL3-Rα+ CD45RA+ granulocyte-macrophage progenitors (GMPs) and Lin- CD34+ CD38+ IL-3Rα- CD45RA-megakaryocyte-erythroid progenitors (MEPs). MS4A3 expression levels of the sorted cells were analyzed with real-time RT-PCR. We detected more than 10-fold amount of MS4A3 transcripts in CMPs than HSCs. Furthermore, its expression level continuously increased along myeloid lineage differentiation to GMP. On the other hand, megakaryocyte-erythroid lineage differentiation was not accompanied by MS4A3 expression and the amount of MS4A3 transcripts in MEPs was minimum as in HSCs. Flow cytometry analyses confirmed that HSCs and MEPs do not express MS4A3 on their cell surface whereas the MS4A3 expression on CMPs and GMPs is detectable. Further, the Lin- CD34+ CD38+ CD33+ cells could be fractionated according to the intensity of cell surface MS4A3 expression. To investigate the significance of cell surface MS4A3 expression for functional analyses of myeloid progenitor cells, we performed methylcellulose colony-forming assays. We found that MS4A3+ cells in Lin- CD34+ CD38+ CD33+ fraction only produced granulocyte/macrophage colonies, losing erythroid colony- and mixed colony-forming capacity. These results suggest that cell surface expression of MS4A3 is useful to distinguish granulocyte/macrophage lineage-committed progenitors from other lineage-related ones in early human hematopoiesis. We also analyzed MS4A3 expression in BM cells obtained from patients with acute leukemia. Flow cytometry analyses revealed that leukemia cells of some patients expressed substantial amount of cell surface MS4A3. In conclusion, MS4A3 is useful to monitor early stage of myeloid differentiation in human hematopoiesis. In addition, our findings of MS4A3 expression on myeloid leukemia cells, while no expression on normal HSCs, imply that MS4A3 might be a therapeutic target molecule in myelogenous leukemia. Further studies would clarify the application of MS4A3 to anti-leukemia therapy. Disclosures No relevant conflicts of interest to declare.

Introduction Anamorsin (AM) is an anti-apoptotic molecule we originally isolated from hematopoietic cell line, Ba/F3. That is dependent on interleukin-3 (IL-3) stimulation. We showed that AM conferred resistance to apoptosis induced by the lack of IL-3 stimulation. AM has no structural homology to any known anti-apoptotic molecules such as Bcl-2 and IAP family members. AM-deficient mice are embryonic lethal at late gestation due to the defect of definitive hematopoiesis. Its embryos are anemic and very small in size. Thus AM is considered to be crucial in erythropoiesis but the mechanisms how AM works remain unclear. In this study, we tried to clarify the roles of AM in erythropoiesis. Methods and results At first, in order to examine the expression levels of AM in each differential stage of erythroid cells, we sorted those cells from eight-week-old mice bone marrow (n=4) using flow cytometry. The expression levels of AM were evaluated by Immunoblotting. The results showed that the proerythroblast population (CD71high Ter119low) expressed AM higher than any other erythroid cells. The expression of AM in CD71high Ter119 high cells and CD71low Ter119high cells, which correspond to basophilic erythroblasts and orthchromatophilic erythroblasts respectively, were markedly decreased. Next, we analyzed whether AM expression is induced by erythropoietin (EPO). Recombinant human EPO was injected intravenously (0.1mg/kg) with a single treatment (day0) to eight-week-old mice (n=3). The bone marrow and spleen cells were collected seven days after injection and the samples were subjected to immunoblotting. The results showed that EPO treatment led to increased expression of AM in both bone marrow and spleen. At 14 days after EPO injection, the expression of AM returned to the basal. Furthermore, we examined the AM expression in the acute anemia model. Blood of 0.2ml was exsanguinated from each mouse every day until hematocrit reached to the value less than 25 % (day 0).In this phlebotomized mice, the AM expression was strongly induced specially in proerythroblast population on day 1 and returned to the basal after the recovery from anemia. To explore the AM expression under the other stress, 5-FU (150 mg/kg), a chemotherapeutic drug, was administered to mice (n=3) with a single treatment (day 0). The increased expression of AM was also detected when the mice were most severely anemic (day7). To investigate the significance of AM expression induced by anemia, we transplanted 106 14.5 dpc wild-type (AM+/+) and AM-deficient (AM–/–) fetal liver cells into lethally irradiated congenic recipient mice. At 8 weeks after transplantation, we analyzed the response to anemia induced by phlebotomy. The results showed that donor-derived cells from AM+/+ fetal liver cells were detected the smooth recovery on day 3 after phlebotomy. On the other hand, donor-derived cells from AM–/– fetal liver cells showed an approximately 5-fold reduction compared to those from AM+/+fetal liver cells. Conclusion and discussion The proerythroblast population expressed AM at the highest level among various erythroid lineage populations. The AM expression level was increased by acute anemia induced by phlebotomy and AM-deficiency could show inefficient recovery of erythropoiesis from the induced acute anemia. Previously we reported that AM is essential for Fe/S cluster biogenesis and AM deficiency impaired the iron homeostasis (ASH abs. #610 (2012)). We also reported that the proerythroblast population in AM-deficient fetal liver cell showed more increased labile iron pool and significantly induced apoptosis. From our previous and present data, it is possible that AM may control erythropoiesis by preventing apoptosis when erythrocyte production is needed. Disclosures No relevant conflicts of interest to declare.

Immune cells show quick responses to infection. Many studies concerning cellular and humoral immunity have focused on the regulation of mature lymphocyte function. However, recent studies revealed that the early stage of hematopoiesis plays an important role in the immune system. In bone marrow, hematopoietic stem/ progenitors (HSPC) are targets of pathogen products and danger signals. After the exposure to Gram-negative lipopolysaccharide, the ligand of Toll-like receptor 4, hematopoietic stem cells enter cell-cycle and differentiate into myeloid lineage cells while B lymphopoiesis almost arrests. Little has been known how early T lymphopoiesis in thymus changes. Previously, we cloned signal-transducing adaptor protein-2 (STAP-2) as a c-fms/M-CSFR interacting protein, and found STAP-2 in T lymphocytes or macrophages is crucial for immune responses. The function of STAP-2 is generally recognized under inflammatory condition, interacting with a variety of signaling or transcriptional molecules. We reported that STAP-2 binds to STAT5 for regulation of T cell proliferation, and in macrophages, STAP-2 combines with MyD88 and IκB kinase to activate NF-κB and enhances the production of IL-6 and TNFα. In this study, the effects of STAP-2 on early T progenitors were evaluated using gene-modified mice. All experimental procedures were conducted under specific pathogen-free conditions, according to protocols approved by Institutional Animal Care and Use Committees of Osaka University. We first evaluated the expression level of STAP-2 in murine thymus with quantitative PCR. STAP-2 mRNA was ubiquitously observed all through T cell development, including the CD4- CD8- double-negative (DN) stage. To test the influences on T lymphopoiesis, we generated knock-out and transgenic mice (Tg) that are modified STAP-2 gene expression. In Tg mice, STAP-2 was overexpressed under the control of the Lck proximal promoter. The promoter could drive expression of the inserted cDNA in T lineage cells from the late of DN 2 stage (CD44+ CD25+). We found that thymus was significantly enlarged in Tg mice (46.7 ± 11.15 mg in WT vs 88.1 ± 25.2 mg in Tg), while the number of T lymphocytes in periphery was comparable to wild-type mice (WT). Results from flow cytometric analysis showed STAP-2 enhanced the percentages of DN2 and DN3 (CD44- CD25+) T progenitors, and the actual numbers of DN2, DN3, DN4 (CD44- CD25-) and CD4+ CD8+ double-positive subpopulations. There were no differences between control and knock-out mice in thymus and peripheral bloods. When Lin- Sca1+ cKitHigh HSPC derived from Tg mice were cultured with Delta-like 1-transduced OP9 stromal cells (OP9-DL1) under T cell generation condition, the development of DN4 cells was accelerated (26.9 ± 6.2 % in WT vs 35.2 ± 4.1 % in Tg). Co-cultures from Lin- CD44+ CD25- cKitHigh early T cell progenitors showed the same tendency. These results indicate that STAP-2 regulates the proliferation and differentiation of T progenitors during DN3 to DN4 stage. To elucidate the signaling regulated by STAP-2, microarray experiment with DN3 T progenitors was conducted. The bioinformative approach with Ingenuity Pathway Analysis showed the canonical pathways related with IL-12 signaling, 4-1BB (CD137) signaling and helper T cell differentiation were significantly influenced. Interestingly, we found that STAP-2 affected the distribution of functional T lymphocytes. The ratio of helper CD4+ cells to suppressor CD8+cells in peripheral bloods was lower in Tg mice than that in WT. In summary, we found that STAP-2 regulates the early T lymphopoiesis in thymus. DN2 to DN4 stages of T progenitors increased in STAP-2 transgenic mice, and STAP-2 promoted the differentiation in vitro. Moreover, STAP-2 affected the cell decision in development to helper CD4+ cells or suppressor CD8+ cells. Our study indicates the up-regulation of STAP-2 under inflammatory condition might be crucial for immune response at the early stage of T lymphopoiesis. Further study would clarify the precise molecular mechanisms of the enhancement of T lymphopoiesis by STAP-2. Disclosures No relevant conflicts of interest to declare.

Introduction: Anamorsin (AM, also called CIAPIN1) was originally isolated as a molecule that conferred resistance to apoptosis caused by growth factor deprivation. AM deficient (AM KO) mice die during late gestation; AM KO embryos are anemic and small compared to wild type (WT) embryos. It suggests that AM is indispensable for embryo growth and hematopoiesis. To determine which signaling pathways AM utilizes for these functions, we analyzed murine embryonic fibroblast (MEF) cells generated from E-14.5 AM KO or WT embryos. Proliferation of AM KO MEF cells was markedly retarded, and PKCθ, PKCδ, and p38MAPK were more highly phosphorylated in AM KO MEF cells. Expression of cyclinD1, the target molecule of p38MAPK, was down-regulated in AM KO MEF cells. P38MAPK inhibitor as well as PKC inhibitor restored expression of cyclinD1 and cell growth in AM KO MEF cells. These data suggested that PKCθ, PKCδ, and p38MAPK activation lead to cell cycle retardation in AM KO MEF cells. However, functions of AM still remain not fully understood. In order to elucidate functions of AM, we generated AM transgenic (Tg) mice under control of CAG promoter. Since our previous study showed AM overexpression in approximately 30% of B cell type malignant lymphoma (DLBCL and FL) cases, we focused on AM overexpressed B cells from AM Tg mice in this study. Methods and Results: At first, we compared the number of B cells in the peripheral blood and spleen size between AM Tg mice and WT mice and found no difference. Next, we analyzed the lipopolysaccharide (LPS) stimulated B cells. B cells were selected from spleen by using anti-B220 antibody. The purified B cells were exposed to LPS for three days and measured. Unexpectedly, LPS-stimulated proliferation of B cells from AM Tg mice was decreased compared to WT mice. Since we initially confirmed that AM overexpression did not affect TLR4 expression on B cells, we examined TLR4 signaling pathway activation status by detecting phosphorylation of the signal transduction molecules using Western-blotting, and found that phosphorylation of Erk1/2 and IKBα were decreased in LPS-stimulated B cells from AM Tg mice. Next, we performed cDNA microarray analysis to reveal the mechanisms of inhibition of the LPS signaling pathways by comparing the differential gene expression profiles in B cells of AM Tg and WT mice with or without LPS stimulation. We extracted 2375 genes from the data sets that met the following criteria: genes with > 2.0 fold change between LPS- vs LPS+ in AM Tg or WT mice, and secondly with >1.5 fold difference in the rate of gene expression change with LPS stimulation between AM Tg and WT mice, and performed upstream regulator analysis to predict the upstream regulators on the cascade of LPS signaling pathway with Ingenuity Pathway Analysis software. The data suggested that Ras activation was decreased in LPS-stimulated AM Tg B cells compared to WT B cells. Then, we confirmed the impairment of LPS induced Ras activation in AM Tg B cells using Ras activation assay kit. Furthermore, we found that AM mRNA level was increased (1.8-4.0 fold) in WT B cells at 4 hours after LPS stimulation. From these data, it was shown that AM overexpression in AM Tg B cells inhibited Ras signaling pathways and retarded the cell proliferation of LPS-stimulated B cells. Conclusion and Discussion: In our previous study, we showed that AM negatively regulates novel PKCs and p38MAPK by using MEFcells generated from AM KO mice, while in this study it was shown that overexpressed AM negatively regulates Ras signaling pathways by using AM Tg B cells. Although there are no reports other than ours that show the relationships between the signaling molecules and AM, our present data showed the novel role of AM that regulates the signaling molecules in the different manner. LPS increased AM expression via Ras activation in B cells, and AM overexpression inhibited Ras, which suggested AM might induce a negative feedback loop that attenuates Ras activation in LPS stimulated B cells. Our previous data showed that low international prognostic index (IPI) DLBCL patients with AM overexpression had a poor prognosis, especially in the patients who received chemotherapy without rituximab. In those patients, AM overexpression might contribute at least in part to chemotherapy resistance possibly through negative regulation of Ras. In summary, we showed that overexpressed AM is one of negative regulators of LPS induced signaling pathways in B cells. Disclosures No relevant conflicts of interest to declare.

B lymphopoiesis is a complex and multistep process originated from hematopoietic stem cells (HSC). Recent studies showed that the microenvironment surrounding progenitor cells affects the development of B cells. Soluble factors that bone marrow (BM) stromal cells produce are crucial to the process. For example, SDF1α regulates the survival and homing of HSC and the lineage progenitor cells. IL7 is produced by osteoblasts and promotes the differentiation and proliferation of pre B cells. Although various molecules have been reported for the important roles, there are many proteins, which are secreted from stromal cells and function as a direct regulator of progenitors, remained unknown. In this study, we aimed to find novel secreted or membrane proteins, which modulate B cell differentiation via the microenvironment. We used MS-5 stromal cells to identify the novel secreted regulators because the cell line is known to have the potential to support HSC and B lymphopoiesis. The secreted proteins have signal sequence to pass into endoplasmic reticulum, followed by cell surface expression. Thus, we applied the modified signal sequence trap method (Tashiro et al, Science, 1993). Briefly, the cDNA library from MS-5 was inserted to the HPC4-TF/pEFBOS vector to produce fusion proteins, composed of proteins from cDNA, a HPC4-epitope, and a tissue factor transmembrane. With the screening of each plasmid based on the capacity to express HPC4-epitope on cell surface, we successfully identified 21 secreted or transmembrane proteins, which MS-5 cells produce. Among these proteins, pleiotrophin, proliferin-2 and osteoblast stimulating factor-5 (OSF-5) were selected because of the limited mRNA expression within stromal cell lines. For the functional analysis without the effects on the process of fetus development, we generated transgenic chimera mice (Tg), which produce the indicated protein under the control of kappa chain promoter. As a result, OSF-5, but not pleiotrophin or proliferin-2, Tg showed impaired B lymphocyte development. In OSF-5 chimera mice, the number of B lineage cells was decreased. B220+ cells in the spleen as well as pre B cells and immature B cells in the BM were significantly decreased (B220low CD43low IgM- pre B cells: 2.2 ± 0.2 x 105 cells in OSF-5 Tg vs. 6.7 ± 1.9 x 105 cells in control, B220+ CD43- IgM+ immature B cells: 2.3 ± 0.6 x 105 cells in OSF-5 Tg vs. 5.3 ± 0.8 x 105 cells in control). OSF-5 is widely expressed in mice BM, spleen, thymus, liver, kidney and lung, and the effects on hematopoiesis have never been examined. OSF-5 includes two splicing variants. Variant 1 is a secreted protein and known as aortic carboxypeptidase like protein. Variant 2 is a non-secreted, intracellular protein, known as adipocyte enhancer binding protein. In mice BM, OSF-5 variant 1 is secreted only from stromal cells, while OSF-5 variant 2 is expressed in hematopoietic cells. First, to exclude the cell-intrinsic effects of OSF-5, we ectopically expressed variant 2 in lineage- Sca-1+ c-Kit+ Flt3- HSC and co-cultured them with MS-5. As a result, the generated number of CD19+ B cells was not changed. In contrast, when we knocked down the secreted type of OSF-5 in OP9 stromal cell line to mimic the BM environment, the modified OP9 cells could support the proliferation of pre B cell line, 2E8 more efficiently, compared to control (7.3 ± 1.1 x 105 cells in KD vs. 4.4 ± 0.7 x 105 cells in control). In addition, the knock-down (KD) of variant 1 protein increased the recovered number of CD19+ cells in co-cultures of BM mononuclear cells (6.6 ± 2.8 x 105 cells in KD vs. 4.7 ± 2.8 x 105 cells in control). Finally, we found that colony-forming unit of pre B cells was decreased in the existence of OSF-5 variant 1 (17 ± 8 colonies in variant 1 vs. 198 ± 22 colonies in control). This result indicated that OSF-5 produced by BM stromal cells had a direct effect to inhibit the proliferation of pre B cells. In conclusion, we identified OSF-5 as a BM stromal cell-derived secreted protein, which has an ability to inhibit B lymphopoiesis via regulating the pre B cell proliferation. Our findings could help us to understand molecular regulatory mechanisms of normal B lymphopoiesis as well as causes of B lymphocyte dysregulation, such as change during aging. Disclosures No relevant conflicts of interest to declare.

A large body of research has demonstrated that the maternal immune system is elaborately regulated during pregnancy to establish immunological tolerance to the fetus. Although our previous works have revealed that female sex hormones, particularly estrogen, play pivotal roles in suppressing maternal B-lymphopoiesis, the precise molecular mechanisms that mediate their functions are largely unknown. Because T and B lymphocytes function coordinately in the adaptive immune system, the inhibition of B-lymphopoiesis during pregnancy should be involved, at least in part, in “maternal-fetal immune tolerance.” Understanding the molecular mechanisms of tolerance would contribute to the development of new methods to inhibit immune responses after organ transplantation, such as rejection by the host or graft-versus-host diseases. The goal of our present study is to identify the molecular pathways through which estrogen exerts its suppressive effect on B-lymphopoiesis. We performed global analyses of estrogen-inducible genes in bone marrow (BM) stromal cells and identified the secreted frizzled-related protein (sFRP) family. A sFRP1-immunoglobulin G (Ig) fusion protein inhibited early differentiation of B-cells originating from BM-derived hematopoietic stem/progenitor cells (HSPC) in culture (Yokota T. et al. Journal of Immunol, 2008). Conversely, sFRP1 deficiency in vivo caused dysregulation of HSPC homeostasis in BM and aberrant increase of peripheral B lymphocytes (Renström J. et al. Cell Stem Cell, 2009). Therefore, in the present study we generated sFRP1 transgenic chimera (TC) mice that produced high levels of circulating sFRP1 after birth to examine the influence of sFRP1 on adult lymphopoiesis in vivo. Further, we generated sFRP5 TC mice using the same procedure to determine whether there were functional differences or redundancies between sFRP1 and sFRP5. The two are most closely related isoforms among the sFRP family and are known to play redundant roles during embryonic development; however, their physiological function in the immune system is largely unknown. Unexpectedly, while only subtle change was detected in the lymphoid lineage of sFRP1 TC mice, we found that the number of B cells was significantly reduced in the sFRP5 TC mice. The frequency of B cells, which normally account for approximately 50% of peripheral leukocytes of wild-type (WT) mice, was reduced to less than 20% in the sFRP5 TC mice. The suppression was likely specific to the B lineage, because overexpression of sFRP5 did not affect myeloid, T, or NK cells. Compared with WT littermates, the body size of sFRP5 TC mice was slightly, but significantly smaller. Thymocyte counts were not affected. In contrast, the number of splenocytes, particularly those of the B lineage, significantly decreased. In BM of sFRP5 TC mice, early B-cell differentiation was inhibited, resulting in the accumulation of cells whose phenotype corresponds to those of common lymphoid progenitors (CLPs). Gene array analyses of the accumulated CLPs indicated that sFRP5 affects the expression of adaptive immune system-related genes. Further, the sFRP5 overexpression was found to induce the expression of Wnt and Notch-related molecules that regulate the integrity of HSPCs. To determine the physiological involvement of sFRP5 in the inhibition of early B-cell differentiation, we exploited mice lacking sFRP5. It is noteworthy that, although the level of sFRP5 expression was minimal in steady-state BM, it was markedly induced after estrogen treatment. We injected water-soluble β-estradiol into WT or sFRP5-null mice for 4 days and evaluated their lympho-hematopoiesis 12 h after the last injection. While the highly HSPC-enriched Lineage- Sca-1+ c-kitHi Flt3- fraction of WT mice was resistant to the treatment, the same fraction of sFRP5-null mice showed a declining trend. Further, although the CLP fraction was significantly reduced in both strains, CLPs of sFRP5-null mice were more sensitive to estrogen than those of WT. We also performed gene expression analyses of WT and sFRP5-null mice after the estrogen treatment. We found that estrogen induced the expression of Hes1 in HSPCs of WT but not sFRP5-null mice. Thus, we conclude that estrogen-inducible sFRP5 blocks the differentiation of HSPCs in BM to B-lymphocytes in the presence of high levels of estrogen, at least in part by activation of the Notch pathway. Disclosures: No relevant conflicts of interest to declare.

Identification of novel markers associated with hematopoietic stem cells (HSCs) is important to progress basic and clinical research regarding the HSC biology. We previously reported that endothelial cell-selective adhesion molecule (ESAM) marks HSCs throughout life in mice (Yokota et al. Blood, 2009). We also demonstrated that ESAM can be a useful indicator of activated HSCs after bone marrow (BM) injury and that ESAM is functionally important for recovering hematopoiesis by using ESAM knockout mice (Sudo et al. J Immunol, 2012). However, the discrepancy between species has been a long-standing obstacle to apply findings in mice to human. For example, established murine HSC markers such as Sca-1 or CD150 are not expressed on human HSCs. Thus, it is important to know if ESAM marks HSCs beyond species and serves as a functional molecule for the HSC property, but information regarding ESAM expression in human HSCs has been quite limited. In this study, we have examined the ESAM expression pattern on human HSCs derived from diverse sources. In addition, we have performed functional assessment of the ESAM-expressing cells. Cord blood (CB), aspirated BM, and granulocyte-colony stimulating factor-mobilized peripheral blood (GMPB) were obtained from healthy donors. BM was also obtained from head of femora of patients who received the hip replacement surgery. All of the protocols were approved by the Institutional Review Board of Osaka University School of Medicine, and we obtained the written agreement form with informed consent from all participants. Mononuclear cells were separated using Ficoll centrifugation from CB, aspirated BM and GMPB. For preparation of BM cells adjacent to bone tissues, trabecular tissues of femora were treated with 2 mg/ml collagenase IV and DNase and gently agitated for 1 hour at 37 °C. Collected cells were analyzed using flow cytometry for cell surface expression of ESAM and other markers. Further, the CD34+ CD38−cells were fractionated according to the intensity of ESAM expression and evaluated in vivo and in vitro functional assays. Flow cytometry analyses revealed that the majority of CB CD34+ CD38− cells expressed ESAM. According to the expression level, CB CD34+ CD38− cells could be subdivided into three populations, namely ESAM−/Low, ESAMHigh, and ESAMBright. While all CB contained a robust ESAMHigh population in CD34+ CD38− cells, the percentage of ESAMBright cells varied widely among CB samples. The ESAMHigh CD34+ CD38− cells also expressed CD90 and CD133, which are known as HSC markers. Methylcellulose colony-forming assays and limiting dilution assays revealed that ESAMHigh fraction enriches primitive hematopoietic progenitors. Further, ESAMHigh cells also reconstituted the long-term human hematopoiesis in NOD/Shi-scid, IL-2Rγnull (NOG) mice. Therefore, as in mice, ESAMHighmarks authentic HSCs in human. On the other hand, ESAMBright CD34+ CD38− cells showed low colony-forming activities and no reconstitution of human hematopoiesis in NOG mice. These ESAMBright CD34+ CD38− cells expressed CD118/leukemia inhibitor factor receptor and endothelial markers such as VE-Cadherin, Flk-1, and CD146, but not CD45. These results suggested that ESAMBright cells in the CB CD34+ CD38− fraction are non-hematopoietic cells. With respect to the other HSC sources such as aspirated BM and GMPB, almost all CD34+ CD38− cells were ESAMHigh and ESAMBright cells were not found in this fraction. Interestingly, however, ESAMBright cells were found in the CD34+ CD38− fraction isolated from collagenase-treated femora. These BM-derived ESAMBright CD34+ CD38− cells expressed endothelial markers as did the CB-derived cells. They could generate CD31+endothelial cells, but not hematopoietic cells in coculture with MS5 stromal cells with vascular endothelial growth factor, stromal-cell-derived factor, and interleukin 16. In conclusion, ESAM expression serves as a marker to enrich HSCs in human regardless of the HSC sources. In addition, the very high intensity of this marker might be useful to isolate non-hematopoietic progenitors from CD34+ CD38− cells, which has been conventionally used as human HSCs. The common feature of ESAM expression of murine and human HSCs suggests a possibility that functional significance of ESAM expression obtained from mouse studies could be applicable to human. Disclosures: No relevant conflicts of interest to declare.

Abstract 610 Introduction: Iron has crucial roles in many cellular biological processes. Cellular iron uptake and export must be tightly regulated. Insufficient iron concentrations impair the function of numerous iron proteins, whereas excess free iron can oxidize and damage the contents of cells. Anamorsin (AM, also called CIAPIN-1) is an anti-apoptotic factor, which we originally isolated as a molecule that confers factor-independent survival of hematopoietic cells. AM-deficient mice are embryonic lethal at late gestation due to the defect of definitive hematopoiesis. It is thought that AM plays a crucial role in hematopoiesis, however its precise biological mechanisms remain unclear. Recently, it was reported that the yeast AM homolog, Dre2, was implicated in cytosolic iron-sulfur (Fe/S) cluster assembly (Zhang Y., et al. Mol.Cell.Biol. 28:5569–5582, 2008). The AM carries conserved cysteine motifs (CX2CXC and twin CX2C) at its C termini, which may form iron binding sites. In this study, we have focused on the possibility that AM may be involved in the maturation of Fe/S cluster and the cellular iron homeostasis, especially, the regulation of labile iron pool (LIP) and that AM may affect the accumulation of reactive oxygen species (ROS), leading to impaired erythropoiesis. Methods and Results: To analyze the function of Fe/S protein, we established wild-type cell lines (AMWT) and AM-deficient cell lines (AMKO) from wild-type and AM-deficient fetal liver (14.5dpc) respectively by using SV40 large T antigen. Iron regulatory protein 1 (IRP1) is a well-known Fe/S protein with dual functions. In the presence of Fe/S cluster, IRP1 functions as a cytosolic aconitase. While, in the absence of Fe/S cluster, IRP1 stabilizes the transferrin receptor (TfR) mRNA by binding to the iron responsive element (IRE). We compared the aconitase activity and the IRE binding activity of IRP1 between AMWT and AMKO. The results showed that the cytosolic aconitase activity in AMKO decreased approximately 30% compared to AMWT and the IRE binding activity of IRP1 in AMKO increased 3-fold compared to AMWT. Furthermore, we compared the iron homeostasis. In the presence of iron chelator, desferrioxamine, the expression of TfR in AMWT was markedly elevated, while it was hardly elevated in AMKO. The LIP is a pool of chelatable and redox-active iron, which serves as a crossroad of cell iron metabolism. The measurement of LIP with the metal-sensitive sensor calcein acetoxymethyl ester showed that AMKO had 5-fold higher cellular LIP than AMWT. Moreover we evaluated the accumulation of ROS and the induction of apoptosis by extracellular iron uptake between AMWT and AMKO. The results showed the accumulation of ROS and the induction of apoptosis in AMKO were enhanced about twice as much as in AMWT. These enhancements could be restored by transduction of AM expressing retrovirus vector to AMKO. We also evaluated the effects of AM-deficiency on erythroid differentiation. Fetal liver cells from wild-type or AM-deficient embryos (14.5dpc) were divided into primitive and more matured erythroid populations based on their expression of CD71 and Ter119 by FACS analysis. AM-deficient fetal liver cells had a significant increase in the CD71low TER119low population, containing primitive erythroid progenitors, compared to wild-type (9.4±2.1% vs. 5.2±1.1%, P Moreover we studied LIP in wild-type or AM-deficient embryo fetal liver cells. In accordance with the cell lines, the LIP in AM-deficient fetal liver cells increased 3 to 5-fold more than in wild-type fetal liver cells. The accumulation of ROS and the number of apoptotic cells also increased 2 to 5-fold in AM- deficient fetal liver cells compared to wild-type fetal liver cells. Thus, it was showed that AM deficiency impaired the iron homeostasis and conferred low sensitivity for iron concentration, resulting in the increase of LIP, the accumulation of ROS and the induction of apoptosis. Furthermore, dysregulation of cellular iron homeostasis was thought to be the cause of the embryonic lethal due to AM deficiency. Conclusion: Our current findings indicate that AM functions in cytosolic Fe/S cluster biogenesis and iron homeostasis and is essential for erythropoiesis. Disclosures: Kanakura: Shire: Consultancy.

It has long been known that lymphopoiesis is transiently suppressed during pregnancy, and that this can be experimentally simulated by treatment with estrogen. Indeed, sensitivity to estrogen made it possible to identify very primitive lymphoid cells in the Lin- c-kitHi Sca-1+ fraction that is enriched with hematopoietic stem cells (HSC) and multipotent progenitors of mouse bone marrow. However, analyses showing how the earliest events in lymphopoiesis is suppressed in those circumstances have not been performed. Hematopoietic and environmental cells are both potential hormone targets and, because of this complexity, very little has been known regarding mechanisms. In the present study, we performed high resolution analysis using RAG-1/GFP reporter mice and confirmed that early lymphoid progenitors (ELP) in the Lin-c-kitHi Sca-1+ fraction are particularly depressed in pregnancy or after estrogen injection. The Lin- c-kitLo fraction containing common lymphoid progenitors (CLP)/pro-lymphocytes was very sensitive to estrogen in stroma-free serum-free culture, but earlier stages of lymphopoiesis seemed to be blocked in vivo. Thus, we focused on identifying mechanisms involving bone marrow environment. Two independent strategies with macroarrays or differential display-PCR were used to isolate stromal cell genes that were both estrogen regulated and associated with inability to support B lymphopoiesis. We have identified secreted frizzled related protein 1 (sFRP1) as an estrogen inducible gene, and its expression corresponded to inability to support lymphopoiesis. This member of the complex Wnt family has been previously described as both an agonist and an antagonist of canonical, β-catenin dependent signaling. We found that sFRP1, like recombinant Wnt3a, stimulated the canonical pathway in the Lin- c-kitHi Sca-1+ fraction, and blocked progression in the B lymphocyte lineage even when exogenous Wnt ligands were not provided. The suppressive effects of sFRP1 and Wnt3a on B lymphopoiesis were canceled out when both were present simultaneously. To better understand the vulnerability of early progenitors to the Wnt signaling, we examined the expression pattern of Frizzled receptors. Interestingly, the HSC-enriched fraction (Lin- c-kitHi Sca-1+ IL-7Rα-RAG-1/GFP-) highly expressed 7 out of 9 Frizzled receptors, which markedly declined as they differentiated to ELP (Lin- c-kitHi Sca-1+ IL-7Rα- RAG-1/GFP+) and CLP (Lin- c-kitLo Sca-1Lo/- IL-7Rα+). Myelo-erythroid progenitors were less affected by sFRP1 in culture, suggesting that it is similar to estrogen with respect to lineage specificity. Bone-lining stromal cells express sFRP1 and the transcripts were elevated in bone marrow by pregnancy or estrogen injection. In summary, these observations implicate sFRP1 as a possible mediator in hormone regulation of the earliest events in lymphopoiesis.

Constitutively activated tyrosine kinases (TKs) cause various types of leukemia. For example, BCR/ABL and TEL/PDGFRβ (T/PRβ) are the causative genes of CML and CMML, respectively. However, these leukemogenic TKs (LTKs) activate common sets of downstream molecules such as Ras/MAPK, PI3-K/Akt, and STATs. So, it remains unknown how disease phenotypes (i.e., the lineage and differentiation stage of leukemic cells and the clinical course of the patient) are determined by LTKs. One possibility is that LTK by itself has a potential to promote the development of leukemic cells specific for each LTK. Alternatively, these phenotypes might originate from the lineage and/or differentiation stage of leukemic stem cells, in which each LTK was initially generated. To characterize these mechanisms, in this study, we utilized FIP1L1/PDGFRα (F/PRα), a causative gene of hyper eosinophilic syndrome (HES)/chronic eosiinophilic leukemia (CEL). First, we transduced F/PRα and T/PRβ into murine c-KithighSca-1+Lin− (KSL) cells using the retrovirus system. As a result, both LTKs enabled KSLs to grow under factor-deprived conditions. Next, we examined the effects of these LTKs on the development of Gr1+IL-5 receptor(R)+ eosinophil progenitors (EoPs) from KSL cells. After 6-day cultures with SCF, TPO, IL-6, and FLT3L, 52% of F/PRα-transfected KSL cells became to be Gr1+IL-5R+, while only 4% and 15% of the cultured cells were Gr1+IL-5R+ in mock- and T/PRβ-transfected KSL cells, respectively. Similarly, F/PRα preferentially developed EoPs from common myeloid progenitor (CMP)s. Furthermore, when expressed in megakaryocyte/erythrocyte progenitor (MEP)s and common lymphoid progenitor (CLP)s, F/PRα inhibited the development of erythroid cells, megakaryocytes and B lymphocytes, respectively. Importantly, F/PRα but not T/PRβ aberrantly developed EoPs from MEPs and CLPs. However, F/PRα was not able to immortalize CMP, granulocyte/monocyte progenitor (GMP), MEP, or CLP. These results suggest that, although F/PRα can reprogram these progenitors towards Eo lineage, it must be generated in hematopoietic stem cells to cause HES/CEL. Next, we expressed chimeric LTKs (F/PRβ and T/PRα) and two activated forms of PRα each harboring point mutation detected in gastrointestinal stromal tumors (GIST) (PRαV561D and PRαD842V) in KSL cells. As a result, all these TKs enabled KSL cells to proliferate under factor-deprived conditions. However, only T/PRα promoted the development of EoP from KSL cells, indicating that specific activity derived from chimeric PRα is necessary to cause HES/CEL. Finally, we examined the effects of F/PRα on the expression and function of lineage specific transcription factors by RT-PCR analysis and luciferase assays. When expressed in KSL cells, F/PRα augmented the expression of C/EBPα and GATA-2 and decreased PU.1 expression in KSL cells as early as 24h prior to the apparent development of EoPs as compared with mock and T/PRβ. Furthermore, we found that F/PRα and its downstream Ras but not STAT5 or PI3-K inhibited the activity of PU.1, but not of GATA-2 or C/EBPα. Together, these results suggest that F/PRα selectively induces EoP from hematopoietic stem/progenitor cells by modifying the expression and function of lineage specific transcription factors.