The transcription factor CCAAT/enhancer-binding protein β (C/EBPβ) regulates the differentiation of a variety of cell types. Here, the role of C/EBPβ expressed by bone marrow mesenchymal stromal cells (BMMSCs) in B-cell lymphopoiesis was examined. The size of the precursor B-cell population in bone marrow was reduced in C/EBPβ-knockout (KO) mice. When bone marrow cells from C/EBPβ-KO mice were transplanted into lethally irradiated wild-type (WT) mice, which provide a normal bone marrow microenvironment, the size of the precursor B-cell population was restored to a level equivalent to that generated by WT bone marrow cells. In coculture experiments, BMMSCs from C/EBPβ-KO mice did not support the differentiation of WT c-Kit+ Sca-1+ Lineage− hematopoietic stem cells (KSL cells) into precursor B cells, whereas BMMSCs from WT mice did. The impaired differentiation of KSL cells correlated with the reduced production of CXCL12/stromal cell-derived factor-1 by the cocultured C/EBPβ-deficient BMMSCs. The ability of C/EBPβ-deficient BMMSCs to undergo osteogenic and adipogenic differentiation was also defective. The survival of leukemic precursor B cells was poorer when they were cocultured with C/EBPβ-deficient BMMSCs than when they were cocultured with WT BMMSCs. These results indicate that C/EBPβ expressed by BMMSCs plays a crucial role in early B-cell lymphopoiesis. Stem Cells2014;32:730–740
Early B-cell lymphopoiesis occurs in bone marrow, during which hematopoietic cells in the bone marrow interact with nonhematopoietic cells in the bone marrow microenvironment . The molecular mechanisms in hematopoietic cells that underlie early B-cell lymphopoiesis have been well investigated. Several transcription factors in hematopoietic cells are essential for early B-cell lymphopoiesis such as PU.1, Ikaros, E2A, early B-cell factor, and paired box protein five . By contrast, the transcription factors in nonhematopoietic cells that are important for early B-cell lymphopoiesis remain unclear.
The transcription factor CCAAT/enhancer-binding protein β (C/EBPβ) is crucial for the differentiation of a variety of cell types, including adipocytes , hepatocytes, keratinocytes, and mammary epithelial cells [4, 5]. C/EBPβ regulates myelopoiesis and granulopoiesis in hematopoiesis [6-8]. However, the role(s) of C/EBPβ expressed by bone marrow mesenchymal stromal cells (BMMSCs), one of the major hematopoiesis-supporting cellular constituents in the bone marrow microenvironment [9-11], in B-cell lymphopoiesis is unknown. In this study, detailed analysis of B-cell subpopulations in C/EBPβ-knockout (KO) mice revealed that the level of precursor B cells was decreased in the bone marrow of C/EBPβ-KO mice. The bone marrow microenvironment, rather than hematopoietic cells, contributed to the defective generation of precursor B cells in C/EBPβ-KO mice. C/EBPβ-deficient BMMSCs did not support the differentiation of precursor B cells from hematopoietic stem cells (HSCs), whereas BMMSCs from wild-type (WT) mice did. In addition, the proliferation and survival of leukemic precursor B-cells cocultured with C/EBPβ-deficient BMMSCs was examined.
Materials and Methods
C57BL/6 (Ly5.2) and Severe combined immunodeficiency (SCID) (C.B-17/lcr-scid/scidJcl) mice were purchased from CLEA Japan (Tokyo, Japan, http://www.clea-japan.com). C/EBPβ-KO (Ly5.2) mice  were back-crossed to C57BL/6 (Ly5.2) mice at least eight times. C/EBPβ-KO (Ly5.2) mice and WT (Ly5.2) mice were obtained by intercrossing heterozygous (C/EBPβ+/−) mice. Transgenic green fluorescent protein (GFP)-expressing (GFP+) mice were kindly provided by Dr. Masaru Okabe (Osaka University) . To obtain GFP+ C/EBPβ-KO and GFP+ WT mice, GFP+ mice were crossed with C/EBPβ+/− mice. C57BL/6 (Ly5.1) mice were kindly provided by Dr. Shigekazu Nagata (Kyoto University). All mice used in the experiments were 7–12-weeks-old. All mice were maintained under specific pathogen-free conditions at the Institute of Laboratory Animals, Kyoto University. All animal experiments were approved by the Committee on Animal Research of the Kyoto University Faculty of Medicine.
Bone Marrow Transplantation
Bone marrow cells (1 × 106 cells per mouse) were administrated intravenously through the tail vein into WT mice that received 10 Gy of total body irradiation prior to transplantation. For competitive transplantation experiments, lethally irradiated (10 Gy) WT mice received bone marrow cells from both GFP− WT mice (1 × 106 cells per mouse) and either GFP+ C/EBPβ-KO or GFP+ WT mice (1 × 106 cells per mouse). B-cell lymphopoiesis was evaluated by flow cytometric analysis at 14–20 weeks after transplantation.
BMMSC Culture and In Vivo Bone Formation Assays
Murine BMMSCs were isolated as previously described [14-16]. Briefly, murine bone marrow cells (1.5 × 107) from long bones was seeded into 10 cm culture dishes, incubated for 3 hours at 37°C to allow adherent cells to attach, and then washed twice with phosphate-buffered saline (PBS) to remove nonadherent cells. BMMSCs formed adherent colonies, and adherent cells were collected after 12–15 days of culture. Primary cultures were passaged to disperse the colony-forming cells (passage 1). Cells at passage 1 were used in experiments. The culture medium consisted of alpha-minimum essential medium supplemented with 20% fetal bovine serum (FBS), 2 mM l-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin, and 55 µM 2-mercaptoethanol (all from Gibco-BRL, Gaithersburg, MD, http://www.gibcobrl.com). For osteogenic induction in vitro, 2 mM β-glycerophosphate (Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com), 100 µM l-ascorbic acid 2-phosphate (Wako Pure Chemical Industries Ltd., Osaka, Japan, http://www.wako-chem.co.jp/english), and 10 nM dexamethasone (Sigma Aldrich) were added to the culture media. Calcium deposition was evaluated by staining with 1% Alizarin Red S after 4 weeks of osteogenesis-inducing culture. For adipogenic induction in vitro, 0.5 mM isobutylmethylxanthine, 60 µM indomethacin, 0.5 µM hydrocortisone, and 10 µg/mL insulin were added to the culture media. Oil red O staining was used to assess lipid-laden fat cells after 1–2 weeks of adipogenesis-inducing culture. The area of the mineralized areas and the number of Oil red O+ cells were quantitated using Image J software. In some experiments, the differentiated BMMSCs were used in further experiments, including immunoblot analysis or quantitative real-time polymerase chain reaction (PCR) to analyze the expression of osteogenesis- or adipogenesis-associated molecules. For in vivo bone formation assays, cultured BMMSCs (1–4 × 106 cells) were mixed with 40 mg of hydroxyapatite/tricalcium phosphate (HA/TCP) ceramic powder (Zimmer, Warsaw, IN, http://www.zimmer.com). The mixture was implanted subcutaneously into the dorsal surface of 7–9-weeks-old SCID mice, and the implants were harvested 8 weeks later. Histological analysis and quantification of bone formation in the harvested implants were performed as previously described [17, 18]. Human bone marrow samples were obtained with informed consent and the approval of the ethical committee of Kyoto University Hospital. Human BMMSCs were isolated from bone marrow samples as previously described [19, 20] and cultured in α-MEM containing 15% FBS, 100 µM l-ascorbic acid 2-phosphate, 2 mM l-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin. To exclude the possibility that hematopoietic cells in BMMSC cultures affect the results, CD45+ cells were isolated from primary cultures of BMMSCs using anti-CD45 immunomagnetic microbeads (Miltenyi Biotech, Bergisch Gladbach, Germany, http://www.miltenyibiotec.com). These cells did not have multidifferentiation capabilities in vitro and did not express C-X-C motif chemokine 12 (CXCL12)/stromal cell-derived factor-1 (SDF-1), stem cell factor (SCF), and interleukin-7 (IL-7), whereas CD45− cells did (Supporting Information Fig. S1). In some experiments, human bone marrow cells and BMMSCs were purchased from AllCells (Emeryville, CA, http://www.allcells.com) or Lonza (Basel, Switzerland, http://www.lonza.com). The clinical features of the precursor B-cell acute lymphoblastic leukemia (B-ALL) samples are shown in Supporting Information Table S2.
Sorting of KSL Cells and Coculture of KSL Cells with BMMSCs
Single-cell suspensions of bone marrow cells from WT (Ly5.1) mice were labeled with allophycocyanin (APC)-conjugated anti-mouse c-Kit (2B8), fluorescein isothiocyanate (FITC)-conjugated anti-mouse Ly6A/E [Sca-1] (D7), and PerCP-Cy5.5-conjugated anti-mouse lineage markers, including CD3e (145-2C11), CD4 (RM4–5), CD8 (53-6.7), CD19 (1D3), B220 (RA3–6B2), CD11b (M1/70), Gr-1 (RB6–8C5), and Ter119 (TER119). All antibodies were purchased from eBioscience (San Diego, CA, http://www.ebioscience.com). Sorting of c-Kit+ Sca-1+ Lineage− (KSL) HSCs was performed with a FACSAria cell sorter (Becton Dickinson, Franklin Lakes, NJ, http://www.bd.com). KSL cells (1 × 104) were cocultured with BMMSCs (2 × 106) in 3 mL of Roswell Park Memorial Institute (RPMI) 1640 supplemented with 10% FBS, 2 mM l-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin, 10 ng/mL SCF, 10 ng/mL fms-like tyrosine kinase 3-ligand (Flt3-L), and 10 ng/mL IL-7 in a 6-cm dish. The medium was replenished after 6 days, and B-cell lymphopoiesis was analyzed after 10 days of coculture. For CXCL12/SDF-1 rescue experiments, recombinant CXCL12/SDF-1 (R&D Systems, Minneapolis, MN, http://www.rndsystems.com) was added to the medium to be a final concentration of 10 ng/mL.
Coculture of Murine Precursor B-ALL Cells with BMMSCs
Murine precursor B-ALL cells, BaF3 that were transfected with Bcr/Ablp185 (BaF3/Bcr-Abl), were generated as previously described . BaF3/Bcr-Abl cells were cocultured with BMMSCs derived from either C/EBPβ-KO or WT mice using a coculture system as previously reported . BMMSCs (1 × 106) were attached to the reverse side of the membrane of a six-well cell culture insert (Corning, Corning, NY, http://www.corning.com) and cultured for 3 days. Then, BaF3/Bcr-Abl cells (5 × 104) were seeded onto the upper side of the insert membrane. In coculture experiments, cells were cultured in RPMI 1640 containing 10% FBS, 2 mM l-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin. The number of BaF3/Bcr-Abl cells on the upper side of the insert was counted after 3 days of coculture. BaF3/Bcr-Abl cells were stained with propidium iodide (PI) after 2 days of coculture to label DNA for cell cycle analysis. After coculture with BMMSCs for 3 days, the percentage of apoptotic BaF3/Bcr-Abl cells was determined by performing APC-conjugated Annexin V and PI (BD Bioscience, San Diego, CA, http://www.bdbiosciences.com) costaining according to the manufacturer's instructions.
Bromodeoxyuridine (BrdU) Incorporation Assay
BaF3/Bcr-Abl cells (2 × 104 cells per well in a 96-well plate) were cultured in RPMI 1640 supplemented with 0, 0.01, 0.1, 1, or 10 ng/mL CXCL12/SDF-1. After 24 hours of culture, BrdU incorporation was analyzed by using the colorimetric immunoassay Cell Proliferation ELISA, BrdU (Roche Applied Science, Basel, Switzerland, http://www.roche-applied-science.com) according to the manufacturer's instructions.
Flow Cytometric Analysis
Single-cell suspensions of bone marrow cells were stained with fluorescence-conjugated antibodies and analyzed with a FACSCanto II (Becton Dickinson). B-cell populations were identified based on the Hardy Fraction . The antibodies used were FITC-conjugated anti-mouse CD43 (R2/60), FITC-conjugated anti-mouse CD45.1/Ly5.1 (A20), phycoerythrin (PE)-conjugated anti-mouse BP-1 (6C3), PE-conjugated anti-mouse IgD (11–26c), APC-conjugated anti-mouse B220 (RA3-6B2), APC-conjugated anti-mouse CD11b (M1/70), PE-Cy7-conjugated anti-mouse CD24 (M1/69), and PE-Cy7-conjugated anti-mouse IgM (II/41). All antibodies were purchased from eBioscience. Dead cells were excluded by staining with PI. Data were analyzed using FlowJo software (Tree Star, Ashland, OR, http://www.treestar.com).
Quantitative Real-Time PCR
Total RNA was extracted using the QIAamp RNA Blood Mini Kit (Qiagen, Hilden, Germany, http://www1.qiagen.com). cDNA was prepared using the PrimeScript RT reagent kit (Perfect Real-Time) (Takara, Otsu, Japan, http://www.takara.co.jp). Real-time PCR was performed using the StepOnePlus real-time PCR system (Applied Biosystems, Foster City, CA, http://www.appliedbiosys tems.com) and a Universal ProbeLibrary (Roche Applied Science). The primer sets and universal probes used are listed in Supporting Information Table S1. Gene expression levels were normalized to the mRNA level of glyceraldehyde-3-phosphate dehydrogenase. All samples were analyzed in duplicate.
Cell lysates were boiled in SDS sample buffer, separated by SDS-PAGE, and transferred to PVDF membranes. Primary antibodies against runt-related transcription factor 2 (Runx2, Abcam, Cambridge, U.K., http://www.abcam.com), alkaline phosphatase (ALP, Abcam), lipoprotein lipase (Lpl, Abcam), fatty acid-binding protein 4 (FABP4, Abcam), peroxisome proliferator-activated receptor γ (PPARγ, Cell Signaling Technology, Beverly, MA, http://www.cellsignal.com), and β-actin (Sigma-Aldrich) were used. Immunoreactive proteins were detected using horseradish peroxidase-conjugated anti-mouse (for anti-Runx2, anti-Lpl, and anti-β-actin) or anti-rabbit (for anti-ALP, anti-PPARγ, and anti-FABP4) immunoglobulin G (GE Healthware Japan, Tokyo, Japan, http://www.gehealthware.com) and visualized using enhanced chemiluminescence or enhanced chemiluminescence prime kits (GE Healthcare). Relative protein expression levels were measured using Quantity One software (Bio-Rad, Hercules, CA, http://www.bio-rad.com).
Histological analysis and quantification of bone formation were performed as previously described . Briefly, bone specimens from the lower vertebrae of mice were stained with hematoxylin and eosin, and trabecular bone areas were measured at ×100 magnification using Image J software. Images were acquired using a DP71 digital camera and DP Controller software in combination with a CX41 microscope and a PlanCN objective lens (×10, 0.25 numerical aperture) (Olympus, Tokyo, Japan, http://www.olympus.co.jp).
Culture supernatant was collected after culturing murine BMMSCs (2 × 106 cells in 3 mL per dish) for 3 days or after coculturing WT KSL cells with the same number of BMMSCs for 10 days in 6 cm dishes. The level of CXCL12/SDF-1 protein in the culture supernatant was measured using the Quantikine mouse CXCL12/SDF-1α ELISA kit (R&D Systems) according to the manufacturer's instructions.
The unpaired Student's t test was used for analysis. Data in bar graphs indicate the mean ± SD, and statistical significance is expressed as follows: *, p < .05; **, p < .01; n.s., not significant.
The Level of Precursor B Cells Is Decreased in the Bone Marrow of C/EBPβ-KO Mice
The levels of B-cell subsets in C/EBPβ-KO mice were analyzed by flow cytometric analysis based on the expression patterns of B220, CD43, BP-1, CD24, surface IgM (sIgM), and surface IgD (sIgD) . The percentage of B220+ B cells in the bone marrow was significantly lower in C/EBPβ-KO mice than in WT mice (Fig. 1A: WT, 26.5% ± 7.3%; C/EBPβ-KO, 19.1% ± 7.1%; *, p < .05). The percentage of B220+ B cells in the spleen was similar in WT and C/EBPβ-KO mice (Supporting Information Fig. S2A–S2C), which is consistent with a previous study . The percentage of B220+CD43+ precursor B cells in the bone marrow was significantly lower in C/EBPβ-KO mice than in WT mice (Fig. 1B). In detail, the percentages of pre-pro-B cells (B220+CD43+BP-1−CD24−; Fraction A), pro-B cells (B220+CD43+BP-1−CD24+ and B220+CD43+BP-1+CD 24low; Fraction B/C), and pre-BI cells (B220+CD43+BP-1+CD24high; Fraction C′) were all significantly lower in C/EBPβ-KO mice than in WT mice (Fig. 1C, 1D). The percentages of pre-BII cells (B220+CD43−sIgM−sIgD−; Fraction D), immature B cells (B220+CD43−sIgM+sIgDlow; Fraction E), and mature B cells (B220+CD43+sIgM+sIgDhigh; Fraction F) in bone marrow were not significantly different between C/EBPβ-KO mice and WT mice (Supporting Information Fig. S3A, S3B).
The Bone Marrow Microenvironment Contributes to the Impairment of B-Cell Lymphopoiesis in C/EBPβ-KO Mice
The bone marrow microenvironment is crucial for early B-cell development ; therefore, we hypothesized that the bone marrow microenvironment contributes to the decreased level of precursor B cells in C/EBPβ-KO mice. We conducted bone marrow transplantation experiments to test this hypothesis (Fig. 2A). Bone marrow cells from C/EBPβ-KO mice were transplanted into lethally irradiated (10 Gy) WT mice (KO → WT). As a control, bone marrow cells from WT mice were transplanted into WT mice (WT → WT). The recipient WT mice should provide a normal bone marrow microenvironment. Fourteen weeks after transplantation, the percentage of B220+ B cells in bone marrow was similar in recipients that received C/EBPβ-deficient bone marrow cells and those who received WT bone marrow cells (Fig. 2B). Detailed analysis of B-cell subsets showed that the levels of pre-pro B cells (Fraction A), pro-B cells (Fraction B/C), and pre-BI cells (Fraction C′) were similar in the two groups (Fig. 2C). The size of the B-cell population in the spleen was similar in both groups (Supporting Information Fig. S4A, S4B).
Next, we performed competitive bone marrow transplantation experiments. The same number of GFP− WT bone marrow cells and either GFP+ WT bone marrow cells (Transplantation-A) or GFP+ C/EBPβ-deficient bone marrow cells (Transplantation-B) were cotransplanted into lethally irradiated WT mice (Fig. 2D). GFP+ C/EBPβ-deficient hematopoietic cells engrafted into recipient bone marrow with approximately 50% cellularity (Fig. 2E, BM, Transplantation-B), which was equivalent to the engraftment rate of GFP+ WT hematopoietic cells (Fig. 2E, BM, Transplantation-A). The percentage of GFP+ cells in the spleens of recipients was also similar in both transplantations (Fig. 2E, SP). GFP+ C/EBPβ-deficient bone marrow cells and GFP− WT bone marrow cells gave rise to a similar percentage of B220+ B cells in the recipient bone marrow when they were cotransplanted (Fig. 2F, Transplantation-B, open bar vs. closed bar). GFP+ WT bone marrow cells and GFP− WT bone marrow cells also generated a similar percentage of B220+ B cells (Fig. 2F, Transplantation-A, dotted bar vs. closed bar). These findings demonstrate that C/EBPβ-deficient bone marrow hematopoietic cells are able to generate B cells as efficiently as normal WT bone marrow hematopoietic cells when transplanted into a normal bone marrow microenvironment. Taken together, these results indicate that the impairment of B-cell lymphopoiesis in C/EBPβ-KO mice is due, at least in part, to the bone marrow microenvironment.
Reverse transplantation from WT mice to C/EBPβ-KO mice was also performed (Supporting Information Fig. S5A). However, donor-derived (Ly5.1) B-cell reconstitution after transplantation could not be evaluated in most C/EBPβ-KO recipient mice because of early death, although an irradiation dose was reduced. When an irradiation dose of 5 Gy was used, donor cells did not engraft (Supporting Information Fig. S5B, S5C). When an irradiation dose of 7 Gy was used, the level of donor-derived (Ly5.1) B220+ cells in peripheral blood and the bone marrow was lower in surviving C/EBPβ-KO recipients than in WT recipients (Supporting Information Fig. S5D, S5E). Perhaps the C/EBPβ-deficient bone marrow microenvironment cannot support hematopoietic and/or immune recover after transplantation.
C/EBPβ-Deficient BMMSCs Have an Impaired Ability to Support the Differentiation of HSCs into Precursor B Cells
BMMSCs are important for the regulation of B-cell lymphopoiesis in the bone marrow microenvironment [1, 11, 25]; therefore, we explored whether C/EBPβ-deficient BMMSCs have an impaired ability to support B-cell lymphopoiesis. KSL HSCs from the bone marrow of WT (Ly5.1) mice (WT-KSL cells) were cocultured with BMMSCs from C/EBPβ-KO (Ly5.2) mice in the presence of SCF, Flt3-L, and IL-7 (Fig. 3A). BMMSCs from WT (Ly5.2) mice were used as a control in the coculture experiments (Fig. 3A). The generation of hematopoietic cells from WT-KSL cells was significantly lower when cells were cocultured with C/EBPβ-deficient BMMSCs than when they were cocultured with WT BMMSCs (Fig. 3B). The generation of B220+ B cells from WT-KSL cells was also significantly lower when cells were cocultured with C/EBPβ-deficient BMMSCs than when they were cocultured with WT BMMSCs (Fig. 3C, 3D; Supporting Information Fig. S6A). Detailed analysis of B-cell subsets showed that differentiation of WT-KSL cells into precursor B-cells was reduced (Fig. 3E, left panels) and differentiation from pre-pro-B cells (Fraction A) to pro-B cells/pre-BI cells (Fraction B/C/C′) was suppressed when cells were cocultured with C/EBPβ-deficient BMMSCs compared to when they were cocultured with WT BMMSCs (Fig. 3E, right panels, 3F). Therefore, C/EBPβ-deficient BMMSCs have an impaired ability to support the differentiation of normal HSCs into precursor B cells.
Reduced Production of CXCL12/SDF-1 by C/EBPβ-Deficient BMMSCs Partially Contributes to Impaired Differentiation of HSCs into B Cells
Next, the expression of B-cell lymphopoiesis-associated humoral factors in BMMSCs was examined. Levels of CXCL12/SDF-1 (Fig. 3G) and Flt3-L (Supporting Information Fig. S6C) mRNA were significantly lower in C/EBPβ-deficient BMMSCs than in WT BMMSCs. Levels of IL-7 (Supporting Information Fig. S6D) and SCF (Supporting Information Fig. S6E) mRNA tended to be lower in C/EBPβ-deficient BMMSCs than in WT BMMSCs, although the difference was not statistically significant. CXCL12/SDF-1 is essential for hematopoiesis, particularly B-cell lymphopoiesis [26, 27]. In addition to mRNA expression, the protein concentration of CXCL12/SDF-1 was significantly lower in the culture supernatant of C/EBPβ-deficient BMMSCs than that of WT BMMSCs (WT, n = 5,9.90 ± 1.93 ng/mL; C/EBPβ-deficient, n = 5,4.47 ± 1.16 ng/mL; **, p < .01) (Fig. 3H). The concentration of CXCL12/SDF-1 in the supernatant of BMMSC cocultures correlated with the number of B220+ B cells that differentiated from WT-KSL cells (Fig. 3I). The addition of exogenous CXCL12/SDF-1 to the coculture of C/EBPβ-deficient BMMSCs and WT-KSL cells slightly increased the total number of cells and the number of B cells (Fig. 3B, 3D). The frequencies of pre-pro-B cells, pro-B, and pre-BI cells (Fractions A, B and C+C′) that differentiated from KSL cells were not apparently affected by the addition of CXCL12/SDF-1 to the culture medium (Supporting Information Fig. S6A, S6B). Thus, reduced production of CXCL12/SDF-1 by C/EBPβ-deficient BMMSCs is partially associated, and other functional abnormalities of C/EBPβ-deficient BMMSCs may be associated with the impaired differentiation of HSCs into B cells in the coculture.
C/EBPβ-Deficient BMMSCs Have an Impaired Multidifferentiation Capability
We sought to identify other differentiation characteristics in which C/EBPβ-deficient BMMSCs were defective. The ability to differentiate into multiple cell types is a fundamental property of bone marrow mesenchymal stem cells [28, 29]; therefore, C/EBPβ-deficient BMMSCs were evaluated in osteogenic and adipogenic differentiation assays. When BMMSCs were cultured under osteogenesis-inducing conditions in vitro, calcium accumulation was significantly lower in C/EBPβ-deficient BMMSCs than in WT BMMSCs, as assessed by Alizarin Red S staining (Fig. 4A, 4B). The expression of the osteogenic master molecule Runx2 and of another crucial osteogenic marker ALP was downregulated in C/EBPβ-deficient BMMSCs in this assay (Fig. 4C, 4D). Moreover, when BMMSCs were subcutaneously implanted with HA/TCP into SCID mice, C/EBPβ-deficient BMMSCs induced less bone formation than WT BMMSCs (Fig. 4E, 4F). Therefore, the osteogenic differentiation capability of C/EBPβ-deficient BMMSCs was defective compared to that of WT BMMSCs. Findings of skeletal examinations of C/EBPβ-KO mice were in agreement with this. Male and female C/EBPβ-KO mice had a shorter crown-rump length and total length than sex- and age-matched WT mice (Supporting Information Fig. S7A, S7B). C/EBPβ-KO mice had less trabecular bone (TB) and fewer bone-lining cells on the surface of TB that were positive for osteocalcin, a marker of osteoblasts, than age-matched WT mice (Supporting Information Fig. S7C–S7F). Next, in vitro adipogenic differentiation assays were performed. When BMMSCs were cultured under adipogenesis-inducing conditions in vitro, lipid deposition was significantly lower in C/EBPβ-deficient BMMSCs than in WT BMMSCs (Fig. 5A–5C). Furthermore, expression of the adipogenic markers PPARγ, Lpl, and Fabp4 was downregulated in C/EBPβ-deficient BMMSCs in this assay (Fig. 5D, 5E). Therefore, the adipogenic differentiation capability of C/EBPβ-deficient BMMSCs was defective compared to that of WT BMMSCs. Taken together, these results demonstrate that C/EBPβ-deficient BMMSCs have an impaired multidifferentiation capability.
Survival of Leukemic Precursor B Cells Is Suppressed when Cocultured with C/EBPβ-Deficient BMMSCs
The results described so far demonstrate that C/EBPβ expressed by BMMSCs plays a crucial role in supporting physiological early B-cell lymphopoiesis; therefore, we explored whether C/EBPβ expressed by BMMSCs is involved in the proliferation and survival of leukemic precursor B cells. The murine precursor B-ALL cell line BaF3/Bcr-Abl was cocultured with BMMSCs from C/EBPβ-KO or WT mice (Fig. 6A). The number of BaF3/Bcr-Abl cells was higher when they were cocultured with WT BMMSCs than when they were cultured alone (Fig. 6B). However, the number of BaF3/Bcr-Abl cells was similar when they were cocultured with C/EBPβ-deficient BMMSCs and when they were cultured alone (Fig. 6B). This difference was not associated with cell cycle (Fig. 6C, 6D). The proportion of apoptotic BaF3/Bcr-Abl cells was significantly higher when they were cocultured with C/EBPβ-deficient BMMSCs than when they were cocultured with WT BMMSCs (Fig. 6E, 6F). Production of CXCL12/SDF-1 was reduced in C/EBPβ-deficient BMMSCs (Fig. 3G, 3H); therefore, the response of BaF3/Bcr-Abl cells to CXCL12/SDF-1 was examined. Stimulation with CXCL12/SDF-1 increased the BrdU incorporation of BaF3/Bcr-Abl cells in a dose-dependent manner (Fig. 6G). These results suggest that reduced production of CXCL12/SDF-1 and impaired antiapoptotic activity in C/EBPβ-deficient BMMSCs contribute to the reduced survival of BaF3/Bcr-Abl cells. In human BMMSCs derived from precursor B-ALL bone marrow samples, the mRNA levels of C/EBPβ and CXCL12/SDF-1 were increased in some cases (Supporting Information Fig. S8A, S8B). Thus, the expression level of C/EBPβ in BMMSCs might modulate the survival of leukemic precursor B cells.
In this study, we showed that the bone marrow microenvironment contributed to the reduced level of precursor B cells in the bone marrow of C/EBPβ-KO mice, and that C/EBPβ-deficient BMMSCs had an impaired ability to support differentiation of HSCs into precursor B cells and an impaired multidifferentiation capability. We and others have previously shown that myelopoiesis and the number of hematopoietic stem and progenitor cells are comparable between C/EBPβ-KO and WT mice [6, 8]. Therefore, impaired B-cell lymphopoiesis in C/EBPβ-KO mice is not due to a reduced level of HSCs and is not associated with increased myelopoiesis.
It was previously reported that reduced production of IL-7 by bone marrow stromal cells and weak response of B220+IgM− B cells to IL-7 contribute to the impairment of B-cell lymphopoiesis in C/EBPβ-KO mice . In this study, the expression of IL-7 tended to be lower in C/EBPβ-deficient BMMSCs than in WT BMMSCs; however, this difference was not statistically significant. Rather, C/EBPβ-deficient bone marrow hematopoietic cells were able to generate B cells as efficiently as normal WT bone marrow hematopoietic cells when they were transplanted into the normal bone marrow microenvironment of WT mice. This finding demonstrates that the bone marrow microenvironment, rather than hematopoietic cells, is responsible for the impairment of B-cell lymphopoiesis in C/EBPβ-KO mice, at least at steady-state. The contributions of hematopoietic cells and the bone marrow microenvironment to the impairment of B-cell lymphopoiesis in C/EBPβ-KO mice reported here differ from those previously reported ; however, the reason(s) for this discrepancy is unclear. The two reports used different methods to evaluate B-cell generation; in vitro liquid culture was performed in the previous report , and in vivo reconstitution following hematopoietic cell transplantation was performed in our analysis. IL-7 is an essential cytokine for early B-cell lymphopoiesis that is expressed in the bone marrow microenvironment , and C/EBPβ-deficient hematopoietic cells exhibit impaired responses to various cytokines in addition to IL-7 . Thus, both the bone marrow microenvironment and hematopoietic cells might contribute to the impairment of B-cell lymphopoiesis in C/EBPβ-KO mice, and the relative contribution of each might vary in different states.
Transcription factors expressed by hematopoietic cells that are essential for early B-cell lymphopoiesis have been well studied . Whereas WT BMMSCs supported the differentiation of normal purified HSCs (WT-KSL cells) into precursor B cells in coculture experiments, C/EBPβ-deficient BMMSCs did not. Therefore, the ability of BMMSCs to support early B-cell lymphopoiesis is dependent upon C/EBPβ. Production of CXCL12/SDF-1 was reduced in C/EBPβ-deficient BMMSCs, and the level of differentiation of normal HSCs into precursor B cells correlated with the concentration of CXCL12/SDF-1 in the supernatant of BMMSC and HSC cocultures. Several lines of evidence indicate that CXCL12/SDF-1 has essential roles in early B-cell lymphopoiesis [26, 27, 31]. This study did not investigate transcriptional regulation of CXCL12/SDF-1 by C/EBPβ directly; however, previous reports demonstrated that C/EBPβ is one of the major regulatory elements driving transcription of CXCL12/SDF-1 [32, 33]. The addition of exogenous CXCL12/SDF-1 to the coculture of C/EBPβ-deficient BMMSCs and WT-KSL cells slightly increased the total number of B cells. Hematopoiesis is regulated by the complicated interaction between hematopoietic cells and the bone marrow microenvironment via adhesion molecules and humoral factors including Flt3-L, IL-7, SCF [1, 10, 11]. Reduced production of CXCL12/SDF-1 by C/EBPβ-deficient BMMSCs is partially associated, but other functional abnormalities of C/EBPβ-deficient BMMSCs may be associated with the impaired differentiation of HSCs into B cells in the coculture. Further studies are needed to elucidate the contribution of C/EBPβ in BMMSCs to B-cell lymphopoiesis.
BMMSCs are multipotent nonhematopoietic cells capable of differentiating into a variety of cell types, including osteoblasts, adipocytes, and chondrocytes [34-37]. Given that C/EBPβ-deficient BMMSCs had an impaired ability to support B-cell lymphopoiesis, we hypothesized that their multidifferentiation capability was also impaired. C/EBPβ-deficient BMMSCs had an impaired osteogenic differentiation capability compared to WT BMMSCs, which was observed concomitant with reduced expression of the osteogenic master molecule Runx2 in vitro, and reduced formation of TB bone and a reduced number of osteoblasts in vivo. Moreover, C/EBPβ-deficient BMMSCs had an impaired adipogenic differentiation capability compared to WT BMMSCs. In summary, C/EBPβ-deficient BMMSCs have an impaired ability to support B-cell lymphopoiesis and reduced osteogenic/adipogenic differentiation capabilities; thus, C/EBPβ is presumably an important regulatory transcription factor needed for BMMSCs to exert their biological effects. Several links have been reported between B-cell lymphopoiesis and osteogenic/adipogenic cells. Osteoblasts support B-cell differentiation and commitment from HSCs . Furthermore, CXCL12 abundant reticular cells, which have characteristics of adipo-osteogenic progenitors, are essential for the proliferation and survival of precursor B cells [25, 39]. The impaired osteogenic and adipogenic differentiation capabilities of C/EBPβ-deficient BMMSCs seem to be associated with their impaired ability to support B-cell lymphopoiesis.
Recently, the contribution of the bone marrow microenvironment to leukemogenesis and chemoresistance, called the “leukemic niche,” has been clarified [40-43]. The association of the leukemic niche with pathogenesis and its potential as a therapeutic target in precursor B-ALL have also been described [44-46]. In this study, the survival of precursor B-ALL cells was suppressed when they were cocultured with C/EBPβ-deficient BMMSCs which produced CXCL12/SDF-1 less than WT BMMSCs. The involvement of CXCL12/SDF-1 in the suppressive effect of C/EBPβ-deficient BMMSCs on BaF3/Bcr-Abl cell survival is in agreement with several studies reporting that the CXCL12 (SDF-1)/CXCR4 axis contributes to the pathogenesis of precursor B-ALL [47-49]. In addition, our cell cycle analyses are consistent with a previous study reporting that the promotion of cell survival through CXCL12 (SDF-1)/CXCR4 axis is independent of cell cycle progression . Several CXCL12/CXCR4 antagonists have been investigated as leukemic niche-targeting therapies [47, 49]. C/EBPβ is a regulator of CXCL12/SDF-1 [32, 33]; therefore, C/EBPβ might also be a therapeutic target in some cases in which C/EBPβ is highly expressed in BMMSCs. The antiapoptotic activity of mesenchymal stem cells in some conditions has been reported , although the detail mechanism underlying this remains unknown. Further studies are needed to elucidate the antiapoptotic effect of BMMSCs on precursor B-ALL cells. Among the various adult hematological malignancies, precursor B-ALL remains difficult to treat with conventional chemotherapy. Better understanding of the contribution of altered C/EBPβ expression in BMMSCs to the pathogenesis of precursor B-ALL may help identify a novel therapeutic target(s) for this disease.
C/EBPβ expressed by BMMSCs regulates B-cell lymphopoiesis, particularly precursor B-cell differentiation. In coculture experiments, survival of leukemic precursor B cells is associated with the expression level of C/EBPβ in BMMSCs. Further studies are needed to elucidate the contribution of C/EBPβ in BMMSCs to the regulation of physiological and pathological B-cell lymphopoiesis in the bone marrow microenvironment.
We thank Yoko Nakagawa and Yoshiko Manabe for their excellent technical assistance. This work was supported in part by a Grant-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology in Japan (to Y.M., Y.H., T.I., H.H., and T.M.), a Grant-in-Aid from the Japan Science and Technology Agency (to Y.M.), and a Grant-in-Aid from the Ministry of Health, Labor and Welfare in Japan (to T.I. and T.M). This work was also supported in part by the Japan Leukemia Research Fund (to Y.M.), the Kyoto University Translational Research Center (to Y.M.), the Ichiro Kanehara Foundation (to Y.M. and S.Y.), the National Cancer Center Research and Development Fund (to T.M., 23-A-23), the Kobayashi Foundation for Cancer Research (to T.M.), the Cell Science Research Foundation (to Y.M.), and the Senshin Medical Research Foundation (to T.M.).
S.Y. and H.Y.: conception and design, collection of data, data analysis and interpretation, and manuscript writing; Y.M.: conception and design, financial support, collection of data, data analysis and interpretation, and manuscript writing; S.S.: collection of data and data analysis and interpretation; Y.H. and A.T.: data analysis and interpretation; T.H.: conception and design, collection of data, and data analysis and interpretation; T.I. and T.M.: conception and design, financial support, data analysis and interpretation, and manuscript writing; H.H.: conception and design, financial support, collection of data, data analysis and interpretation, and manuscript writing; A.T.-K.: conception and design, data analysis and interpretation, and manuscript writing; All authors listed approve this manuscript.
Disclosure of Potential Conflicts of Interest
The authors indicate no potential conflicts of interest.