Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL

Osteoclasts, the multinucleated cells that resorb bone, develop from hematopoietic cells of monocyte/macrophage lineage. Osteoclast-like cells (OCLs) are formed by coculturing spleen cells with osteoblasts or bone marrow stromal cells in the presence of bone-resorbing factors. The cell-to-cell interaction between osteoblasts/stromal cells and osteoclast progenitors is essential for OCL formation. Recently, we purified and molecularly cloned osteoclastogenesis-inhibitory factor (OCIF), which was identical to osteoprotegerin (OPG). OPG/OCIF is a secreted member of the tumor necrosis factor receptor family and inhibits osteoclastogenesis by interrupting the cell-to-cell interaction. Here we report the expression cloning of a ligand for OPG/OCIF from a complementary DNA library of mouse stromal cells. The protein was found to be a member of the membrane-associated tumor necrosis factor ligand family and induced OCL formation from osteoclast progenitors. A genetically engineered soluble form containing the extracellular domain of the protein induced OCL formation from spleen cells in the absence of osteoblasts/stromal cells. OPG/OCIF abolished the OCL formation induced by the protein. Expression of its gene in osteoblasts/stromal cells was up-regulated by bone-resorbing factors. We conclude that the membrane-bound protein is osteoclast differentiation factor (ODF), a long-sought ligand mediating an essential signal to osteoclast progenitors for their differentiation into osteoclasts. ODF was found to be identical to TRANCE/RANKL, which enhances T-cell growth and dendritic-cell function. ODF seems to be an important regulator in not only osteoclastogenesis but also immune system.     

Effects of parathyroid hormone (PTH)-related protein and PTH on osteoclasts and osteoclast precursors in vivo

Increased production of PTH-related protein (PTHrP) and PTH is frequently responsible for hypercalcemia and its associated morbidity. However, it is unclear whether these peptides produce identical effects on cells in the osteoclast lineage in vivo. To examine the effects of continuous in vivo exposure to these factors on both the osteoclast precursors and mature osteoclasts, we inoculated Chinese hamster ovarian cells expressing PTH-(1-84), PTHrP-(1-141), or nontransfected Chinese hamster ovarian cells into nude mice. The effects of these tumors on blood ionized calcium, plasma PTH and PTHrP concentrations, and osteoclast formation were then determined. PTH and PTHrP tumor-bearing mice became hypercalcemic (1.90 +/- 0.04 and 1.97 +/- 0.16 mmol/liter, respectively) compared with control mice (1.29 +/- 0.015 mmol/liter). After 4 days of hypercalcemia, mice were killed, and bone marrow cells were harvested to examine cells at three discrete stages of osteoclast development: multipotent osteoclast precursors, the granulocyte/macrophage colony-forming unit; more committed marrow mononuclear osteoclast precursors; and mature osteoclasts. Neither PTH nor PTHrP had an effect on granulocyte/macrophage colony-forming unit, but similarly increased the number of more committed mononuclear osteoclast progenitors as well as mature osteoclasts in the calvaria. No differences were detected between the effects of PTH and PTHrP on cells in the osteoclast lineage in vivo. Thus, PTH and PTHrP appear to affect only more differentiated cells in the osteoclast lineage, and the differences in osteoclastic bone resorption between primary hyperparathyroidism and humoral hypercalcemia of malignancy are probably due to mechanisms other than effects on osteoclast precursor cells in vivo.     

Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia. European Organization for Research and Treatment of Cancer--Childhood Leukemia Cooperative Group

Osteoclasts (OCL) resorb bone. They are essential for the development of normal bones and the repair of impaired bones. The function of OCL is presumed to be supported by cytokines and other biological mediators, including tumor necrosis factor (TNF)-alpha and nitric oxide (NO). Bacterial lipopolysaccharide (LPS) is a potent inducer of TNF-alpha and inducible nitric oxide synthase (iNOS), which is the specific enzyme for synthesizing NO from L-arginine. To obtain direct evidence on LPS-induced TNF-alpha production and iNOS expression by OCL, OCL-enriched cultures were prepared by 7-day cocultures of bone marrow cells of adult BALB/c mice and osteoblastic cells (OBs) derived from calvaria of newborn BALB/c mice, and the generation of TNF-alpha and iNOS in OCL stimulated with LPS was examined immunocytochemically. When the cultured cells were stimulated with 100 ng/ml of LPS, OCL clearly showed TNF-alpha and iNOS expression. Without LPS-stimulation, no expression was observed. TNF activity in the culture supernatants of the OCL-enriched cultures in the presence of LPS was also detected by cytotoxic assay that used TNF-sensitive L929 cells. The dentin resorption activity of OCL was estimated by area and number of pits formed on dentin slices, which were covered by the OCL fraction and cultured in the presence or absence of LPS, sodium nitroprusside (SNP; a NO generating compound), N(G)-monomethyl L-arginine acetate (L-NMMA; a competitive inhibitor of NO synthase (NOS)), or LPS plus L-NMMA. Pit formation was obviously inhibited in the presence of SNP and slightly inhibited in the presence of L-NMMA, but it was not affected in the presence of LPS or LPS plus L-NMMA. These findings indicate that OCL produces TNF and expresses iNOS in response to LPS, but the LPS-activation of OCL scarcely affects pit formation by them.     

Cadherin-6 mediates the heterotypic interactions between the hemopoietic osteoclast cell lineage and stromal cells in a murine model of osteoclast differentiation

Osteoclasts are multinucleated cells of hemopoietic origin that are responsible for bone resorption during physiological bone remodeling and in a variety of bone diseases. Osteoclast development requires direct heterotypic cell-cell interactions of the hemopoietic osteoclast precursors with the neighboring osteoblast/stromal cells. However, the molecular mechanisms underlying these heterotypic interactions are poorly understood. We isolated cadherin-6 isoform, denoted cadherin-6/2 from a cDNA library of human osteoclast-like cells. The isolated cadherin-6/2 is 3,423 bp in size consisting of an open reading frame of 2,115 bp, which encodes 705 amino acids. This isoform lacks 85 amino acids between positions 333 and 418 and contains 9 different amino acids in the extracellular domain compared with the previously described cadherin-6. The human osteoclast-like cells also expressed another isoform denoted cadherin-6/1 together with the cadherin-6. Introduction of cadherin-6/2 into L-cells that showed no cell-cell contact caused evident morphological changes accompanied with tight cell-cell association, indicating the cadherin-6/2 we isolated here is functional. Moreover, expression of dominant-negative or antisense cadherin-6/2 construct in bone marrow-derived mouse stromal ST2 cells, which express only cadherin-6/2, markedly impaired their ability to support osteoclast formation in a mouse coculture model of osteoclastogenesis. Our results suggest that cadherin-6 may be a contributory molecule to the heterotypic interactions between the hemopoietic osteoclast cell lineage and osteoblast/bone marrow stromal cells required for the osteoclast differentiation. Since both osteoclasts and osteoblasts/bone marrow stromal cells are the primary cells controlling physiological bone remodeling, expression of cadherin-6 isoforms in these two cell types of different origin suggests a critical role of these molecules in the relationship of osteoclast precursors and cells of osteoblastic lineage within the bone microenvironment.     

Rat osteoclast precursors in vivo express a vitronectin receptor and a chloride-bicarbonate exchanger

In vivo osteoclast precursors, which are mononuclear, were previously found to express TRAP (tartrate-resistant acid phosphatase) and CTR (calcitonin receptor), like multinucleated osteoclasts. In vitro, they were found to express, in addition, VNR (vitronectin receptor) and CBE (chloride-bicarbonate exchanger). In order to ascertain that osteoclast precursors in vivo express VNR and CBE like their in vitro counterparts, we used immunohistochemistry to localize these molecules in developing long bones of neonatal rats. Frozen sections of metatarsals and phalanges of 1-2 day-old rats were stained for TRAP and mineralization using histochemistry or were reacted with polyclonal antibodies specific for either the beta3 chain of the VNR or synthetic sequences of the CBE. Both mature, multinucleated osteoclasts within the forming marrow cavity of metatarsals (as shown previously) and mononuclear osteoclast precursors located outside the bony collar of the phalangeal calcified rudiment (as shown here for the first time) expressed both TRAP, VNR and CBE. These findings suggest that mononuclear osteoclast precursors express many of the phenotypical markers of multinucleated osteoclasts prior to their fusion and multinucleation which may allow them to resorb bone, as suggested by in vitro observations of pit formation by preosteoclasts cultured on resorbable substances.     

Antigen expressed on tumor cells fails to elicit an immune response, even in the presence of increased numbers of tumor-specific cytotoxic T lymphocyte precursors

We have used T-cell receptor (TCR) transgenic mice to analyze the interaction of tumors with the immune system. We show that the tumor cell line Lewis lung-lymphocytic choriomeningitis virus (LL-LCMV), genetically manipulated to express an H-2 Db-restricted epitope of the lymphocytic choriomeningitis virus glycoprotein (LCMV33-41), can grow progressively in TCR transgenic mice, where approximately 50% of CD8+ T cells are specific for LCMV33-41. TCR transgenic T cells were not deleted in tumor-bearing mice, and their surface phenotype and cytokine secretion patterns remained typical of naive T cells. Also, TCR transgenic T cells from tumor-bearing mice had undiminished capacity to proliferate to antigen in vitro. Tumor-protective immune responses could be elicited in TCR transgenic mice by immunization with LCMV33-41 peptide-loaded dendritic cells. Tumor resistance correlated with the switch of TCR transgenic T cells from a CD44low to a CD44high phenotype and increased capacity to produce IFNgamma in vitro. Results similar to those obtained in TCR transgenic mice were also obtained using an adoptive transfer system, where small numbers of TCR transgenic T cells were injected into normal C57BL/6 hosts. These data indicate that even large tumors may not induce specific immunization, tolerance, or anergy to tumor antigens, and that high numbers of tumor-specific CTL precursors are not sufficient to provide tumor resistance.     

Reliability and validity of three physical activity questionnaires in Flemish males

Many B cell precursors die while differentiating in mouse bone marrow. To ascertain the mechanisms involved in this process, populations of B lineage cells and their tissue localization were analyzed in bone marrow of transgenic mice overexpressing the apoptosis inhibitor, Bcl-2. Immunofluorescence labeling and mitotic arrest were used to quantitate the number and proliferative activity of mu- pro-B cells (terminal deoxynucleotidyl transferase [TdT]+B220-, TdT+B220+, and TdT-B220+); pre-B cells (cmu+); and B cells (smu+). Mature B cells (IgM+IgD+) were increased 16- to 20-fold. In addition, immature B lymphocytes (IgM+IgD-/low), representing newly formed cells, were increased three- to sixfold, whereas pre-B cells and late pro-B cells were increased 30 to 60% in production rate. Earlier pro-B cells expressing TdT were unaffected. In spleen, both mature and immature B cells were greatly increased, but cells of precursor phenotype were few and TdT+ cells were absent. The in vivo location of B cells was examined by autoradiography using light and electron microscopy after intravenous injection of 125I-labeled antibodies. B lineage cells (B220+) were increased throughout bone marrow, often within dilated venous sinusoids, particularly in subosteal regions. Many intravascular and perisinusoidal cells were IgDhigh mature B lymphocytes. In contrast, many other IgM+ and IgDlow immature B lymphocytes clustered extravascularly around the central venous sinus. Plasma cells with distended endoplasmic reticulum were numerous. These findings provide evidence that, in addition to expanding the recirculating pool of B cells entering bone marrow from the blood stream, high levels of Bcl-2 can inhibit some of the apoptosis occurring during B cell differentiation, thereby expanding populations of B lymphopoietic precursor cells within the bone marrow parenchyma.     

Mice lacking osteopontin show normal development and bone structure but display altered osteoclast formation in vitro

We have used homologous recombination in embryonic stem cells to generate mice with a targeted disruption of the osteopontin (Opn, or Spp1, for secreted phosphoprotein 1) gene. Mice homozygous for this disruption fail to express osteopontin (OPN) as assessed at both the mRNA and protein level, although an N-terminal fragment of OPN is detectable at extremely low levels in the bones of -/- animals. The Opn -/- mice are fertile, their litter size is normal, and they develop normally. The bones and teeth of animals not expressing OPN are morphologically normal at the level of light and electron microscopy, and the skeletal structure of young animals is normal as assessed by radiography. Ultrastructurally, proteinaceous structures normally rich in OPN, such as cement lines, persist in the bones of the Opn-/- animals. Osteoclastogenesis was assessed in vitro in cocultures with a feeder layer of calvarial osteoblast cells from wild-type mice. Spleen cells from Opn-/- mice cells formed osteoclasts 3- to 13-fold more frequently than did control Opn+/+ cells, while the extent of osteoclast development from Opn -/- bone marrow cells was about 2- to 4-fold more than from the corresponding wild-type cells. Osteoclast development occurred when Opn-/- spleen cells were differentiated in the presence of Opn-/-osteoblasts, indicating that endogenous OPN is not required for this process. These results suggest that OPN is not essential for normal mouse development and osteogenesis, but can modulate osteoclast differentiation.     

Head injuries in Papua New Guinea

Human calcitonin (hCT) has been reported to have a less hypocalcemizing effect on rats and to have a lower binding affinity for the receptor of mouse osteoclasts than salmon CT(sCT). In this study we comparatively examined the effect of hCT and sCT on osteoclastic bone-resorbing activity of unfractionated cells obtained from human giant cell tumor of bone and from rabbit and mouse long bones. We found that hCT had the same inhibitory effect as sCT on the bone-resorbing activity of human and rabbit osteoclastic cells, but a different one on that of mouse cells. These results indicate that the activity of drugs should be assayed using human cells if possible.     

Big and little forms of osteoclast activating factor

We have further characterized osteoclast activating factor (OAF) using a bioassay for bone resorption which utilizes the release of previously incorporated (45)Ca from fetal rat long bones in organ culture. When supernatant media from activated leukocyte cultures were concentrated on Amicon PM10 membranes (assigned molecular weight cutoff 10,000 daltons) and chromatographed on Sephadex G-50 columns, the bone-resorbing activity eluted between the molecular weight markers chymotrypsinogen (25,000 daltons) and cytochrome c (12,500 daltons). This peak of biological activity has been called big OAF. When filtrates from the PM10 membranes were concentrated on Amicon UM2 membranes (assigned molecular weight cutoff 1,000 daltons) and chromatographed on Sephadex G-50 columns, some of the biological activity eluted between the molecular weight markers chymotrypsinogen and cytochrome c (big OAF), but there was a separate peak of biological activity which eluted with [(3)H]proline (140 daltons). This second peak has been called little OAF. Little OAF was eluted from Bio-Gel P6 columns between the molecular weight markers calcitonin (approximately 3,500 daltons) and vitamin B(12) (1,330 daltons), but was retained by Spectrapor dialysis tubing (nominal molecular weight cutoff 3,500 daltons). Big OAF was converted to little OAF by equilibration in 1 M NaCl or 2 M urea. Little OAF was self-associated back to big OAF by equilibration in buffers of low ionic strength (Tris-HCl 10-50 mM). Little OAF was extracted into the organic phase in ethyl acetate after acidification of the sample to pH 3.5. The biological activity remained in the aqueous phase after ethyl acetate extraction at pH 7.5-8.4. Little OAF has been purified more than 6,000-fold compared with the original material so that bone-resorbing activity is maximal in a sample with a protein concentration of 80 ng/ml.     

A role for TGFbeta1 in langerhans cell biology. Further characterization of the epidermal Langerhans cell defect in TGFbeta1 null mice

Previous studies of TGFbeta1 null (-/-) mice indicated that the epidermis was devoid of Langerhans cells (LC) and that the LC deficiency was not secondary to the inflammation that is the dominant feature of the -/- phenotype (Borkowski, T.A., J.J. Letterio, A.G. Farr, and M.C. Udey. 1996. J. Exp. Med. 184:2417-2422). Herein, we demonstrate that dendritic cells could be expanded from the bone marrow of -/- mice and littermate controls. Bone marrow from -/- mice also gave rise to LC after transfer into lethally irradiated recipients. Thus, the LC defect in TGFbeta1 null mice does not result from an absolute deficiency in bone marrow precursors, and paracrine TGFbeta1 production is sufficient for LC development. Several approaches were used to assess the suitability of -/- skin for LC localization. A survey revealed that although a number of cytokine mRNAs were expressed de novo, mRNAs encoding proinflammatory cytokines known to mobilize LC from epidermis (IL-1 and TNFalpha) were not strikingly overrepresented in -/- skin. In addition, bone marrow-derived LC populated full-thickness TGFbeta1 null skin after engraftment onto BALB/c nu/nu recipients. Finally, the skin of transgenic mice expressing a truncated loricrin promoter-driven dominant-negative TGFbeta type II receptor contained normal numbers of LC. Because TGFbeta1 signaling in these mice is disrupted only in keratinocytes and the keratinocyte hyperproliferative component of the TGFbeta1 -/- phenotype is reproduced, these results strongly suggest that the LC defect in TGFbeta1 null mice is not due to an epidermal abnormality but reflects a requirement of murine LC (or their precursors) for TGFbeta1.     

Ultrastructural characteristics of Fc epsilon R-positive basophils in the spleen and bone marrow of mice immunized with goat anti-mouse IgD antibody

BACKGROUND: Previous work showed that injection of mice with goat anti-mouse IgD antibodies results in increased numbers of Fc epsilon R-positive, non-B, non-T cells in the spleen and Fc epsilon R-positive cells in the bone marrow, and that some of these cells had ultrastructural features of basophils. Fc epsilon R-positive, non-B, non-T cells express virtually all of the capacity of mouse splenic "non-B, non-T cells" to produce interleukin-4 in response to stimulation by cross-linking of Fc epsilon R or Fc gamma R, or by the calcium ionophore, ionomycin. EXPERIMENTAL DESIGN: The present study is a detailed ultrastructural analysis of Fc epsilon R-positive bone marrow cells or Fc epsilon R-positive splenic non-B, non-T cells sorted from mice injected with goat anti-mouse IgD antibody and of Fc epsilon R-positive bone marrow cells or spleen cells pooled from normal mice not injected with goat anti-IgD. RESULTS: Basophils represented the majority (90%) of the granulated cells present in the Fc epsilon R-positive splenic non-B, non-T cells or Fc epsilon R-positive bone marrow cells of goat anti-IgD-injected mice. In contrast, the cytoplasmic granule-containing Fc epsilon R-negative cells sorted from spleen or bone marrow of goat anti-IgD-injected animals contained predominantly a mixture of neutrophils, eosinophils, monocytes and their precursors. Both the Fc epsilon R-positive and -negative preparations contained rare (< 5%) cells with ultrastructural features of very immature mast cells. Basophils were also identified in Fc epsilon R-positive cells sorted from total bone marrow cells or spleen cells of normal mice not injected with goat anti-IgD. CONCLUSIONS: Taken together with data concerning the numbers of Fc epsilon R-positive, non-B, non-T cells in the spleen, and Fc epsilon R-positive B220-negative cells in the bone marrow, these ultrastructural findings indicate that injection of mice with goat anti-IgD results in increased numbers of basophils, particularly in the spleen, that exhibit an 8-fold increase in basophils as a result of injection of goat anti-IgD.     

The cytoplasmic domain of CD4 promotes the development of CD4 lineage T cells

Thymocytes must bind major histocompatibility complex (MHC) proteins on thymic epithelial cells in order to mature into either CD8+ cytotoxic T cells or CD4+ helper T cells. Thymic precursors express both CD8 and CD4, and it has been suggested that the intracellular signals generated by CD8 or CD4 binding to class I or II MHC, respectively, might influence the fate of uncommitted cells. Here we test the notion that intracellular signaling by CD4 directs the development of thymocytes to a CD4 lineage. A hybrid protein consisting of the CD8 extracellular and transmembrane domains and the cytoplasmic domain of CD4 (CD884) should bind class I MHC but deliver a CD4 intracellular signal. We find that expression of a hybrid CD884 protein in thymocytes of transgenic mice leads to the development of large numbers of class I MHC-specific, CD4 lineage T cells. We discuss these results in terms of current models for CD4 and CD8 lineage commitment.     

Human osteoclast-like cells are formed from peripheral blood mononuclear cells in a coculture with SaOS-2 cells transfected with the parathyroid hormone (PTH)/PTH-related protein receptor gene

Subclones of the human osteosarcoma cell line SaOS-2 were established by transfecting with an expression vector containing the human PTH/PTH-related protein (PTHrP) receptor, and their abilities to support osteoclast-like multinucleated cell (OCL) formation were examined in coculture with mouse or human hemopoietic cells. Of four subclones examined, SaOS-2/4 and SaOS-4/3 bound high levels of [125I]-PTH and produced a significant amount of cAMP in response to PTH. OCLs were formed in response to PTH in the cocultures of mouse bone marrow cells with either SaOS-2/4 cells or SaOS-4/3 cells. Human OCLs were also formed in response to PTH in the coculture of SaOS-4/3 cells and human peripheral blood mononuclear cells. Adding dexamethasone together with PTH greatly enhanced PTH-induced human OCL formation. Like mouse OCLs, human OCLs formed in response to PTH were tartrate-resistant acid phosphatase positive, expressed abundant calcitonin receptors and vitronectin receptors, and formed resorption pits on dentine slices. Other osteotropic factors such as 1alpha,25-dihydroxyvitamin D3, prostaglandin E2, and interleukin 6 plus soluble interleukin 6 receptors failed to induce mouse and human OCLs in cocultures with SaOS-4/3 cells. Both mouse and human OCL formation supported by SaOS-4/3 cells were inhibited by either adding an antibody against macrophage-colony stimulating factor or adding granulocyte/macrophage-colony stimulating factor. Thus, it is likely that human and mouse OCL formation supported by SaOS-4/3 cells are similarly regulated. These results indicate that the target cells of PTH for inducing osteoclast formation are osteoblast/stromal cells but not osteoclast progenitor cells in the coculture. This coculture model will be useful for investigating the abnormalities ofosteoclast differentiation and function in human metabolic bone diseases.     

Heterogeneity of glycans at each N-glycosylation site of horseradish peroxidase

Development of hematopoietic stem cells is regulated by stromal cells of the bone marrow. Many stromal cell lines have been established from temperature-sensitive SV40 large T-antigen gene transgenic mice and used to examine regulation of the purified stem cells. When the sorted stem cells were cocultured on the stromal cell layers, cobblestone formation was induced by the stromal cells. The cobblestones were formed by finite cell division (8 divisions on average) of sorted Lin- c-Kit+ Sca1+ stem cells committed to myeloid or lymphoid lineages. These stromal cell lines showed variable activities supporting the stem cell development. In one stromal cell line, TBR59, two waves of cobblestone formation committed to either myeloid lineage or lymphoid lineage were induced. TBR31-1, another bone marrow stromal cell line, induced only the cobblestone formation committed to lymphoid lineage. These results indicate that the bone marrow stromal cells selectively induce lineage-specific commitment of the stem cells. Both cobblestone formations require c-Kit function as well as adhesive interaction through VLA4 and VCAM1.     

Chimeric homeobox gene E2A-PBX1 induces proliferation, apoptosis, and malignant lymphomas in transgenic mice

Expression of the homeobox fusion gene E2A-PBX1 under control of the immunoglobulin heavy chain enhancer efficiently induced malignancies in transgenic mice. All animals died before 5 months of age with lymphomas that demonstrated phenotypes consistent with transitional intermediate thymocytes (CD4+/CD8+/CD3med). E2A-PBX1 also markedly altered lymphoid development in pretumorous animals, reducing the number of thymocytes and bone marrow B lineage progenitors to 20% of normal levels. In spite of the observed reductions in lymphoid cells, premalignant animals contained significantly increased numbers of cycling thymocytes, but a higher proportion was also undergoing apoptosis, suggesting that increased cell death resulted in the marked lymphopenias. These data indicate that the chimeric homeodomain protein E2A-PBX1 paradoxically induces both proliferation and apoptosis in lymphoid cells, suggesting an in vivo association between nuclear oncogene-induced cell cycle progression and programed cell death.     

Interactions between c-kit and stem cell factor are not required for B-cell development in vivo

The receptor-type tyrosine kinase, c-kit is expressed in hematopoietic stem cells (HSC), myeloid, and lymphoid precursors. In c-kit ligand-deficient mice, absolute numbers of HSC are mildly reduced suggesting that c-kit is not essential for HSC development. However, c-kit- HSC cannot form spleen colonies or reconstitute hematopoietic functions in lethally irradiated recipient mice. Based on in in vitro experiments, a critical role of c-kit in B-cell development was suggested. Here we have investigated the B-cell development of c-kit-null mutant (W/W) mice in vivo. Furthermore, day 13 fetal liver cells from wild type or W/W mice were transferred into immunodeficient RAG-2-/- mice. Surprisingly, transferred c-kit- cells gave rise to all stages of immature B cells in the bone marrow and subsequently to mature conventional B2, as well as B1, type B cells in the recipients to the same extent as transferred wild type cells. Hence, in contrast to important roles of c-kit in the expansion of HSC and the generation of erythroid and myeloid lineages and T-cell precursors, c-kit- HSC can colonize the recipient bone marrow and differentiate into B cells in the absence of c-kit.