SEARCH

SEARCH BY CITATION

Keywords:

  • dlk;
  • Pref-1;
  • Hematopoiesis;
  • Colony formation;
  • Stem cell factor;
  • Hes-1

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Delta-like (dlk) is a family of transmembrane proteins containing epidermal growth factor-like repeat motifs homologous to the notch/delta/serrate family. Recent studies suggest that dlk is a negative regulator of adipocyte differentiation, a promoting factor of cobblestone area colony formation, and a molecule which influences stromal cell-pre-B cell interactions and augments cellularity of developing thymocytes. However, the role of dlk in regulating the growth and differentiation of hematopoietic progenitors remains unclear. In the present study, we examined the effect of dlk on the proliferation of murine hematopoietic progenitors by hematopoietic growth factors. Soluble dlk-IgG Fc chimeric protein completely inhibited the colony formation of lineage-marker negative (Lin) bone marrow cells by GM-CSF, G-CSF, or macrophage-CSF (M-CSF) in the presence of stem cell factor (SCF). However, dlk failed to inhibit the colony formation of Lin bone marrow cells by CSF, as described above, or M-CSF plus interleukin 3. Furthermore, dlk failed to inhibit the colony formation of Hes-1-null fetal liver cells by M-CSF in the presence of SCF. These findings suggest that dlk is an important regulator of hematopoietic progenitor proliferation. Depending on the presence of SCF, dlk may act as a growth inhibitor, although dlk signaling does not mediate Hes-1 transcription factor.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Hematopoietic stem cells are defined as cells having both the capacity for self-renewal and the ability to differentiate into an all mature hematopoietic lineage [1, 2]. Recently, hematopoietic stem cells were isolated [3, 4] and it was shown that a single hematopoietic stem cell was able to reconstitute a lethally irradiated recipient mouse [3]. It is believed that self-renewal and differentiation of hematopoietic stem cells are modulated by close contact with the bone marrow microenvironment composed of stromal cells. However, the mechanisms of self-renewal and differentiation of hematopoietic stem cells are poorly understood. Recently, it was elucidated that Notch signaling is essential for the cell-cell interaction regulating cell fate decisions throughout development [5, 6]. Therefore, Notch signaling is one candidate regulator of stem cell fate. Indeed, activation of Notch1 in 32D myeloid progenitors inhibits G-CSF-induced granulocytic differentiation, but permits expansion of undifferentiated cells [7-9]. In addition, constitutively activated Notch mutations are implicated in T cell leukemia and lymphoma [10], and the Notch receptor is expressed in human CD34+ primitive hematopoietic progenitor cells [11]. Moreover, Notch1 plays a role in T cell development by CD4/CD8 cell fate decision [12] and αβ and γδ T cell choices [13]. Additionally, hairy and enhancer of split homologue-1 (Hes-1) is known as a molecular target of the Notch signal transduction pathway [6, 14], and experiments reconstituting the lymphoid system of RAG2-null mice with Hes-1-null (Hes–/–) fetal liver cells reveal that Hes-1 is essential for expansion of early T cell precursors [15]. These findings suggest that Notch ligands may play a role in self-renewal and differentiation of hematopoietic stem cells [16].

Two Notch ligand families, Delta [17, 18] and Serrate/ Jagged [19], have been identified in mammalians. All ligands are transmembrane proteins with similar overall structural features, including a large extracellular domain containing a DSL (Delta-Serrate-Lag-2) motif that is specific to Notch ligands, followed by variable numbers of iterated epidermal growth factor (EGF)-like repeats, a single transmembrane segment, and a short, less well-conserved cytoplasmic domain. Recently, it was reported that the Notch ligands are indeed expressed in stromal cells, and influence the growth and differentiation of primitive hematopoietic precursor cells [20-25].

Moore et al. identified one cDNA clone specifically expressed in stromal cell lines that is able to maintain high levels of transplantable multilineage stem cell activity [26]. This molecule is known as a delta-like (dlk) [27], preadipocyte factor-1 (Pref-1) [28, 29] and fetal antigen 1 [30]. Like Notch ligand, dlk is an EGF-like repeat protein. However, dlk lacks the DSL motif indicative of the Notch ligand family. Since the DSL motif is critical for Notch binding [31], it is thought that dlk is not a Notch ligand. The biological functions of dlk have been reported [26-30, 32, 33]. dlk is a negative regulator of adipocyte differentiation [28, 29], a stimulator of cobblestone area colony formation in stromal cell-dependent hematopoiesis [26] and a molecule that influences stromal cell-pre-B cell interactions [32]. Furthermore, dlk augments the cellularity of developing thymocytes and increases Hes-1 expression in fetal thymus organ cultures [33]. These observations suggest that dlk is one of the important components of stem cell regulation in the hematopoietic microenvironment. In this regard, the investigation of this gene is relevant and significant.

In the present study, we describe the biological functions of dlk on hematopoiesis. The present findings indicate that dlk inhibits colony formation by GM-CSF, G-CSF, or macrophage-CSF (M-CSF) in the presence of stem cell factor (SCF) as co-stimulator. However, dlk does not inhibit colony formation by GM-CSF, G-CSF, M-CSF, or M-CSF plus interleukin 3 (IL-3). Therefore, the present observations suggest that dlk inhibits SCF-induced colony formation. Furthermore, in the present study, experiments using Hes–/– fetal liver cells showed that this inhibitory effect is independent of Hes-1 expression.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Preparation of Murine dlk-Human IgG Fc Chimeric Protein

A cDNA fragment corresponding to the extracellular domains coding region (codon 140 to 1062) of dlk/Pref-1 [28, 29] was generated by reverse transcription polymerase chain reaction amplification from mRNA of the stromal cell line PA6 [34]. The primers used were 5′-CCACTCGAGATGATCGCGACCGGAGCCCT and 5′-CTTGGGATCCTCGGTGAGGAGAGGGGTACTC as sense and antisense primers, respectively. The fragment was cloned into the Eco RI site of pCRII (Invitrogen Corporation; San Diego, CA). As the result of sequence analysis, this gene was consistent with the intact extracellular domains of dlk. The XhoI-Bam HI fragment of this cloned dlk gene was inserted into the CD4:IgG1 hinge fusion gene (a gift of B. Seed, Massachusetts General Hospital; Boston, MA) [35] from which the XhoI-Bam HI fragment containing the CD4 gene was removed. To prepare murine dlk-IgG Fc chimeric protein (mDLK-Fc), this construct was transiently transfected into COS-1 cells, which were then cultured for four days with GIT medium (WAKO Pure Industries; Osaka, Japan), and the chimeric protein in the culture supernatant was purified using Prosep-A (Bioprocessing; Princeton, NJ) [36, 37]. The control chimeric protein, human CD4-IgG Fc chimeric protein (HuCD4-Fc), was also prepared in the same manner.

Hematopoietic Growth Factors

Recombinant human G-CSF (rHuG-CSF) was a gift from Dr. T. Nakahata (Kyoto University; Kyoto, Japan; http://www. kyoto-u.ac.jp/index-e.html). Other hematopoietic growth factors used in the present study were as follows: recombinant murine IL-3 (rmIL-3) from Genzyme Biochemical Ltd. (Boston, MA), rmSCF from R&D Systems Inc. (Minneapolis, MN; http://www. rndsystems.com/asp/g_home.asp), rmGM-CSF from Intergen (Purchase, NY), and rmM-CSF from Upstate Biotechnology (Lake Placid, NY). Unless otherwise indicated, all hematopoietic growth factors were used at predetermined optimal concentrations: rmIL-3, 200 U/ml; rmSCF, 50 ng/ml; rmGM-CSF, 1 ng/ml; rmM-CSF, 10 ng/ml; rHuG-CSF, 100 ng/ml.

Mice

Male BDF1 mice, seven to nine weeks old, were obtained from Japan SLC Inc. (Shizuoka, Japan) and acclimatized in our laboratory for a minimum of five days prior to use. Hes–/– mice [38] were provided by Dr. R. Kageyama (Institute for Virus Research; Kyoto University). Since Hes–/– mice were embryonic lethal, Hes–/– fetuses were obtained by crossing with HES+/– mice.

Monoclonal Antibodies (mAbs) and Immunomagnetic Beads

Rat mAbs directed against murine hematopoietic lineage antigens were used to prepare lineage-marker negative (Lin) cells. Mac-1 (M1/70) was obtained from American Type Culture Collection ([ATCC]; Rockville, MD; http://www.atcc.org). The anti-erythroid cell mAb (TER119) and the anti-B220 (RA3-6B2) mAb were provided by Dr. T. Kina (Kyoto University) and Dr. N. Minato (Kyoto University), respectively. All antibodies were purified from media conditioned by hybridoma cells. mAb to CD4 (GK1.5), CD8a (53-7.7), and Gr-1 (RB6-8C5) were purchased from PharMingen (San Diego, CA; http://www.pharmingen.com). Negative immunomagnetic selection was performed using sheep anti-rat IgG-conjugated immunomagnetic beads (M450 Dynabeads, Dynal; Oslo, Norway; http://www.dynal.no).

Bone Marrow Cells and Purification of Lin Bone Marrow Progenitors

Normal murine bone marrow cells were prepared from femura of normal BDF1 mice, and red blood cells were lysed with Tris-buffered ammonium chloride solution.

Lin bone marrow cells were purified according to a previously described protocol [39]. The bone marrow cells were suspended in NycoPrep solution (1.063 g/ml; Nycomed; Oslo, Norway; http://www.nycomed-amersham.com). This suspension was layered over higher density NycoPrep solution (1.077 g/ml; Nycomed) and centrifuged at 1,000 × g for 30 min at room temperature. The cells at the interface were harvested and washed twice with phosphate-buffered saline (PBS) containing 0.1% (wt/vol) bovine serum albumin (BSA). The cells were incubated with an antibodies cocktail (Mac-1, Gr-1, B220, CD4, CD8a, and TER119) for 40 min on ice and washed twice with PBS containing 0.1% BSA. Magnetic beads were added at a ratio of 40:1 (beads/cell) and the mixture was incubated for 30 min on ice. Labeled (Lin+) cells were removed by a magnetic particle concentrator (Miltenyi Biotec; Bergisch Gladbach, Germany; http://www.miltenyibiotec.com), and the lineage-depleted cell population was then collected from supernatant and washed with α-medium (GIBCO-BRL Laboratories; Grand Island, NY) containing 10% fetal calf serum (FCS).

Preparation of Fetal Liver Cells

Day-14 embryos were screened for Hes-1 genotype as previously described [38]. Day-14 fetal livers were dissection-free in SFO3, disaggregated into single-cell suspensions, passed through a Cell Strainer (70 μm, Falcon; Becton Dickinson; Mountain View, CA; http://www.bd.com) and then suspended in Tris-buffered ammonium chloride solution to lyse non-nucleated mature erythrocytes. Cells were washed three times in SFO3, plated in 100-mm culture dishes, and then cultured in 37°C, 5% CO2 to remove adherent cells. After 1 h, nonadherent cells were collected and used for colony assays.

Colony Assays

Methylcellulose culture was performed by using a modification of a technique described previously [40]. In serum-containing cultures, 1 × 105 of unfractionated bone marrow cells or 1 × 104 of Lin bone marrow cells, were cultured in 1 ml culture medium containing α-medium, 0.85% methylcellulose (4,450 centipoises; Shin-Etsu Chemical Co.; Nagano, Japan; http://www.shinetsu.co.jp/english/index.html), 20% prescreened heat-inactivated (56°C for 30 min) FCS (GIBCO-BRL Laboratories), 1% deionized BSA (Organon Tekunika; Boxtel, Holland), 1 × 10–4 mol/l 2-mercaptoethanol (Nacalai Tesque Inc.; Kyoto, Japan), 50 U/ml penicillin (Banyu Pharmaceutical Co.; Tokyo, Japan), 50 μg/ml streptomycin (Meiji Seika Kaisha; Tokyo, Japan) and hematopoietic factors in triplicate 35-mm petri dishes (Falcon 1008; Becton Dickinson) in a fully humidified atmosphere of 5% CO2 at 37°C. In serum-free culture, α-medium was replaced by SF-03 in which FCS was omitted. On day 7 of culture, aggregates consisting of 50 more cells were scored as a colony. Abbreviations for colony types are as follows: G = granulocyte colonies; GM = granulocyte/ macrophage colonies; M = macrophage colonies.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Effect of mDLK-Fc on IL-3 plus SCF-Induced Colony Formation

The effect of mDLK-Fc, homodimer of a soluble form of murine dlk, on the colony formation of unfractionated bone marrow cells in serum-free methylcellulose culture is shown in Table 1. In contrast to a previous study [26], colony formation of unfractionated bone marrow cells in the presence of SCF plus IL-3 was inhibited by mDLK-Fc in a dose-dependent manner. Colony formation of macrophages was significantly inhibited by mDLK-Fc, whereas colony formation of another lineage cell was essentially unaffected. When exogenous hematopoietic growth factors were not added to the culture, no colony formation was observed in the presence of mDLK-Fc or HuCD4-Fc (data not shown).

Table Table 1.. Effect of mDLK-Fc concentration on colony formation in serum-free culture
  
AdditiveConcentration μg/mlTotalMGMGBLMix
  1. a

    Cultures of 1 × 105 unfractionated bone marrow cells were stimulated by 200 U/ml of recombinant murine IL-3 plus 50 ng/ml of rmSCF with varying concentrations of chimeric proteins. After seven days of incubation, colonies were scored. Numbers indicate mean ± SD of colony numbers in triplicate cultures.

  2. b

    *The difference between factor-induced colony formation in the presence or absence of mDLK-Fc was significant (p < 0.05) using Student's t test. Abbreviations: M = macrophage; GM = granulocyte-macrophage; G = granulocyte; BL = blast.

None 62.2 ± 10.640.3 ± 3.58.3 ± 3.53.3 ± 0.610.3 ± 3.1 
mDLK-Fc5.024.5 ± 5.1*4.7 ± 3.1*5.0 ± 3.01.7 ± 1.28.7 ± 3.14.4 ± 2.4*
 2.531.0 ± 1.5*10.7 ± 4.5*6.7 ± 2.31.0 ± 1.08.3 ± 1.54.3 ± 0.6*
 1.035.6 ± 12.9*11.7 ± 5.0*8.3 ± 1.51.3 ± 0.613.3 ± 4.91.0 ± 1.7
 0.546.9 ± 8.1*19.7 ± 3.1*6.7 ± 1.23.0 ± 1.717.3 ± 4.00.6 ± 1.2
HuCD4-Fc5.059.4 ± 19.940.6 ± 14.97.3 ± 1.02.0 ± 2.010.5 ± 3.0 
 2.562.2 ± 10.840.5 ± 12.610.0 ± 1.72.0 ± 1.09.4 ± 2.30.3 ± 0.5
 1.059.4 ± 6.837.6 ± 14.57.0 ± 2.63.2 ± 1.110.7 ± 8.60.9 ± 1.0
 0.563.0 ± 22.541.5 ± 11.27.6 ± 2.63.2 ± 1.19.4 ± 11.51.3 ± 0.5

It is difficult, in such cases, to know whether the observed effects were mediated directly on primitive hematopoietic progenitors or indirectly via accessory cell function. To define the inhibitory effects of mDLK-Fc on colony formation, Lin bone marrow cells were obtained using the immunomagnetic separation method. As shown in Table 2, a similar inhibitory effect was also observed in Lin bone marrow cells. Thus, mDLK-Fc acted directly on primitive hematopoietic progenitors and inhibited macrophage colonies stimulated by SCF plus IL-3.

Table Table 2.. Effect of dlk on SCF plus IL-3-induced colony formation in serum-free culture
 
FactorsTotalMGMGBLMix
  1. a

    Triplicate cultures of 1 × 104 Lin bone marrow cells were stimulated by SCF plus IL-3 in the presence of either mDLK-Fc (5 μg/ml) or HuCD4-Fc (5 μg/ml). After seven days of incubation, colonies were scored.

  2. b

    *The difference between factor-induced colony formation in the presence or absence of mDLK-Fc was significant (p < 0.05) using Student's t test. Abbreviations: M = macrophage; GM = granulocyte-macrophage; G = granulocyte; BL = blast; rm = recombinant murine.

rmSCF + rmIL-333.4 ± 8.014.7 ± 6.15.3 ± 1.26.3 ± 1.25.7 ± 0.61.4 ± 1.2
rmSCF + rmIL-3 + mDLK-Fc32.4 ± 5.08.0 ± 1.0*4.7 ± 1.26.0 ± 4.010.7 ± 2.1*3.0 ± 2.1
rmSCF + rmIL-3 + HuCD4-Fc33.6 ± 3.315.0 ± 0.55.3 ± 1.35.8 ± 1.56.0 ± 0.91.5 ± 1.7

Effect of mDLK-Fc on SCF-Induced Colony Formation in Lin Bone Marrow Cells

The next experiment was aimed at testing the effects of mDLK-Fc on the colony formation of Lin bone marrow cells by CSFs. In serum-containing methylcellulose cultures, mDLK-Fc showed no inhibitory effect on GM-CSF, G-CSF, or M-CSF-induced colony formation (Table 3). In serum-free methylcellulose cultures, CSF alone did not stimulate colony formation as noted previously [41]. Therefore, CSFs were used in combination with SCF. As shown in Table 3, we observed the complete inhibition of colony formation by mDLK-Fc.

Table Table 3.. Effect of mDLK-Fc on CSF or CSF plus SCF-induced colony formation of Lin bone marrow cells
 
FactorsTotalMGMGBLMix
  1. a

    Triplicate cultures of 1 × 104 Lin cells were stimulated by indicated hematopoietic growth factors in the presence of either mDLK-Fc (5 μg/ml) or HuCD4-Fc (5 μg/ml). After seven days of incubation, colonies were scored.

  2. b

    *The difference between factor-induced colony formation in the presence or absence of mDLK-Fc was significant (p < 0.01) using Student's t test. Abbreviations: M = macrophage; GM = granulocyte-macrophage; G = granulocyte; BL = blast; rm = recombinant murine.

Serum-free culture      
rmGM-CSF000000
rmGM-CSF + rmSCF78.0 ± 14.161.0 ± 7.50.7 ± 1.2016.3 ± 5.90
rmGM-CSF + rmSCF + mDLK-Fc0.3 ± 0.6*0*000.3 ± 0.6*0
rmGM-CSF + rmSCF + HuCD4-Fc75.4 ± 15.656.0 ± 11.01.2 ± 1.0018.2 ± 5.10
rHuG-CSF000000
rHuG-CSF + rmSCF23.3 ± 1.22.0 ± 1.203.6 ± 0.617.7 ± 0.70
HuG-CSF + rmSCF + mDLK-Fc0*000*0*0
rHuG-CSF + rmSCF + HuCD4-Fc20.2 ± 2.32.5 ± 1.004.3 ± 1.813.4 ± 2.50
rmM-CSF000000
rmM-CSF + rmSCF67.0 ± 22.953.0 ± 20.80.3 ± 0.3013.7 ± 3.20
rmM-CSF + rmSCF + mDLK-Fc0*0*000*0
rmM-CSF + rmSCF + HuCD4-Fc67.7 ± 4.051.3 ± 4.51.0 ± 0015.3 ± 5.90
Serum-containing culture      
rmGM-CSF61.7 ± 13.246.0 ± 13.114.0 ± 1.701.7 ± 2.10
mGM-CSF + mDLK-Fc42.3 ± 3.826.0 ± 3.514.0 ± 3.502.3 ± 0.60
mGM-CSF + HuCD4-Fc51.3 ± 5.837.3 ± 5.512.3 ± 5.001.7 ± 1.50
rHuG-CSF1.3 ± 1.50.7 ± 0.60.3 ± 0.600.3 ± 0.60
rHuG-CSF + mDLK-Fc7.0 ± 4.63.6 ± 1.51.7 ± 1.50.7 ± 0.61.0 ± 1.70
HuG-CSF + HuCD4-Fc5.7 ± 3.53.7 ± 1.51.0 ± 1.00.3 ± 0.60.7 ± 1.20
rmM-CSF125.0 ± 18.5123.3 ± 18.01.7 ± 0.6000
rmM-CSF + mDLK-Fc116.7 ± 21.5115.7 ± 20.30.7 ± 1.20.3 ± 0.600
rmM-CSF + HuCD4-Fc140.3 ± 3.5140.3 ± 3.50000

To confirm whether mDLK-Fc could inhibit SCF plus CSF-induced colony formation, Lin bone marrow cells were cultured with SCF plus CSF in the presence of mDLK-Fc or HuCD4-Fc in serum-containing conditions. As observed previously, mDLK-Fc inhibited monocytic colonies stimulated by SCF plus IL-3 (Tables 1, 2, Table 2.). Therefore, we used combinations of monocyte-restricted hematopoietic growth factors, M-CSF and SCF. As shown in Table 4, the combination of M-CSF and SCF slightly increased the total number of colonies (11% increase) compared with M-CSF alone. SCF acts synergistically with M-CSF to support macrophage colony formation. Next, we tested whether mDLK-Fc inhibits the synergistic effects of SCF and M-CSF. The findings are presented in Table 4; mDLK-Fc partially inhibited colony formation by M-CSF plus SCF (37% decrease). However, anti-c-fms antibody, AFS98 [37] completely inhibited colony formation.

Table Table 4.. Effect of mDLK-Fc on M-CSF plus SCF-induced colony formation in serum-containing culture
 
FactorsTotalMGMGBLMix
  1. a

    Triplicate cultures of 1 × 104 Lin cells were stimulated by M-CSF and M-CSF plus-IL-3 in the presence of either mDLK-Fc (5 μg/ml) or HuCD4-Fc (5 μg/ml). After seven days of incubation, colonies were scored.AFS98 = anti-c-fms mAb (5 μg/ml).

  2. b

    *The difference between factor-induced colony formation in the presence or absence of mDLK-Fc was significant (p < 0.05) using Student's t test. Abbreviations: M = macrophage; GM = granulocyte-macrophage; G = granulocyte; BL = blast; rm = recombinant murine.

rmM-CSF125.0 ± 18.5123.3 ± 18.01.7 ± 0.6000
rmM-CSF + rmSCF138.7 ± 31.5134.7 ± 30.74.0 ± 1.7000
rmM-CSF + rmSCF + mDLK-Fc87.3 ± 4.9*83.3 ± 5.5*4.0 ± 3.0000
rmM-CSF + rmSCF + HuCD4-Fc132.7 ± 23.9126.7 ± 30.65.3 ± 7.600.7 ± 1.20
rmM-CSF + rmSCF + AFS980.7 ± 1.2*0.7 ± 1.2*0000

Effect of mDLK-Fc on IL-3-Induced Colony Formation

As described in the above results, in the presence of SCF, mDLK-Fc could inhibit colony formation by various hematopoietic growth factors. SCF supports the proliferation of multipotential progenitors [2]. IL-3 also supports the proliferation of multipotential progenitors [2]. We next examined whether mDLK-Fc inhibits the synergistic effects of IL-3 and M-CSF. If mDLK-Fc selectively inhibits SCF-induced colony formation, it will not inhibit colony formation by IL-3. As shown in Table 5, in serum-containing culture, the combination of M-CSF and IL-3 did not increase the total number of colonies but induced granulocyte-macrophage and granulocyte colonies compared with M-CSF alone. In contrast with the findings of M-CSF plus SCF, mDLK-Fc could not inhibit colony formation of all lineages by M-CSF plus IL-3. In serum-free culture, mDLK-Fc markedly inhibited colony formation by SCF plus M-CSF (Table 3). However, mDLK-Fc could not inhibit colony formation by IL-3 plus M-CSF in serum-free culture (Table 5). Therefore, mDLK-Fc appeared to inhibit SCF-induced colony formation.

Table Table 5.. Effect of mDLK-Fc on M-CSF plus IL-3 induced colony formation in serum-containing culture
 
FactorsTotalMGMGBLMix
  1. a

    Triplicate cultures of 1 × 104 Lin cells were stimulated by M-CSF plus SCF in the presence of either mDLK-Fc (5 μg/ml) or HuCD4-Fc (5 μg/ml). After seven days of incubation, colonies were scored.

  2. b

    *The difference between factor-induced colony formation in the presence or absence of mDLK-Fc was significant (p < 0.05) using Student's t test. Abbreviations: M = macrophage; GM = granulocyte-macrophage; G = granulocyte; BL = blast; rm = recombinant murine.

rmM-CSF125.0 ± 18.5123.3 ± 18.01.7 ± 0.5000
rmM-CSF + rmIL-3126.0 ± 20.999.7 ± 8.213.9 ± 2.44.8 ± 2.73.2 ± 1.64.2 ± 3.2
rmM-CSF + rmIL-3 + mDLK-Fc129.6 ± 15.299.2 ± 17.113.3 ± 1.010.7 ± 2.42.1 ± 1.04.3 ± 3.4
rmM-CSF + rmIL-3 + HuCD4-Fc124.8 ± 8.587.5 ± 7.226.7 ± 2.42.7 ± 1.04.3 ± 1.03.9 ± 1.9

Effect of mDLK-Fc on SCF-Induced Colony Formation in Hes–/– Fetal Liver Cells

Hes-1 is a known target of the Notch1 signaling pathway [6, 14]. To investigate whether Hes-1 is involved in the effect of dlk on SCF-induced colony formation, Hes–/– fetal liver cells were cultured with SCF plus M-CSF in the presence of mDLK-Fc or HuCD4-Fc in serum-free conditions. As shown in Figure 1, fetal liver cells from both wild-type (Hes+/+) and Hes–/– mice were capable of forming colonies in response to M-CSF plus SCF, although the numbers were reduced relative to Hes+/+ mice. Furthermore, the inhibitory effect of mDLK-Fc on colony formation by M-CSF plus SCF was observed in fetal liver cells from both Hes+/+ and Hes–/– mice. This finding suggests that dlk signaling does not influence Hes-1.

thumbnail image

Figure Figure 1.. Effect of mDLK-Fc on SCF-induced colony formation in Hes+/+and Hes–/–fetal liver cells.Fetal liver cells were obtained from either Hes+/+or Hes–/–embryos at 14 days of gestation. The cells were cultured with M-CSF plus SCF with mDLK-Fc or CD4-Fc as described inMaterials and Methods. The findings indicate the total number of colonies per 2×105fetal liver cells plated from several animals. The error bars represent one standard deviation of the mean. *p = 0.001 for total number of colonies from addition of mDLK-Fc compared with total number of colonies from addition of CD4-Fc (Student's t test).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We demonstrated that dlk actually functions in hematopoiesis using a methylcellulose culture system. In serum-free methylcellulose cultures of unfractionated bone marrow cells, IL-3 plus SCF-induced colony formation was inhibited by mDLK-Fc in a dose-dependent manner. Moreover, on Lin bone marrow cells, mDLK-Fc reduced the number of colonies induced by IL-3 plus SCF. Interestingly, colony formation of macrophages was significantly inhibited by mDLK-Fc, whereas colony formation of other lineage cells was essentially unaffected. From this finding, it is implied that mDLK-Fc inhibits macrophage colony formation.

It is known that suppressive cytokines such as macrophage inflammatory protein-1α (MIP-1α), tumor necrosis factor-α (TNF-α), interferon-γ (IFN-γ), IL-1α, and transforming growth factor-β (TGF-β) are produced by accessory cells [42-48]. These cytokines inhibit the proliferation of hematopoietic progenitors stimulated by various hematopoietic growth factors. For example, TGF-β inhibits colony formation by M-CSF [48] and IL-3 [46], but stimulates colony formation by GM-CSF [42, 48]. TNF-α also inhibits colony formation by G-CSF plus SCF, or IL-3 plus SCF [43]. Therefore, to determine whether the inhibitory effects of mDLK-Fc on colony formation mediate accessory cells, we tested the effects of mDLK-Fc on Lin bone marrow cells as accessory cells depleted cells. From this experiment, it was indicated that mDLK-Fc inhibits colony formation by IL-3 plus SCF. If accessory cells such as macrophages and lymphocytes produce suppressive cytokines by mDLK-Fc, addition of mDLK-Fc will affect the colony formation by CSF alone. Since the present findings showed that mDLK-Fc had no effect on CSF-induced colony formation (Table 3), the possibility of indirect effects through accessory cells is excluded.

SCF or IL-3, or both, promote the differentiation of multipotent stem cells as well as more lineage-restricted hematopoietic progenitors, while other hematopoietic growth factor such as GM-CSF, G-CSF, and M-CSF are known to act upon cells later in various hematopoietic lineages and exhibit relatively restricted lineage specificity [2]. Moreover, mDLK-Fc inhibited macrophage colony formation induced by IL-3 plus SCF. Based on this finding, we tested the effects of mDLK-Fc on the colony formation of Lin bone marrow cells in the presence of CSFs. In serum-containing methylcellulose cultures, mDLK-Fc did not inhibit colony formation of Lin bone marrow cells by GM-CSF, G-CSF, or M-CSF as single agents. In serum-free methylcellulose cultures, GM-CSF, G-CSF, and M-CSF as single agents failed to support colony formation from Lin bone marrow cells. Therefore, we examined the effect of mDLK-Fc on colony formation by GM-CSF, G-CSF, and M-CSF in the presence of the SCF. Interestingly, mDLK-Fc completely inhibited the colony formation of Lin bone marrow cells by the combination of SCF with GM-CSF, G-CSF, or M-CSF. This finding led to the assumption that mDLK-Fc may inhibit SCF-induced colony formation. To address this question, we tested the effect of mDLK-Fc on M-CSF plus SCF-induced colony formation of Lin bone marrow cells in serum-containing methylcellulose cultures. mDLK-Fc markedly inhibited M-CSF plus SCF-induced colony formation. When IL-3 was used as an early-acting hematopoietic growth factor instead of SCF, mDLK-Fc did not inhibit M-CSF plus IL-3-induced colony formation in serum-free or serum-containing methylcellulose cultures.

The mechanism of the inhibitory effect of dlk on SCF-induced colony formation is not known. However, it is possible that SCF induces dlk receptor but IL-3 does not. If this is the case, we can test the bindings of mDLK-Fc to SCF-precultured bone marrow cells by fluorescence activated cell sorter analyses. However, we could not detect a population binding with mDLK-Fc (data not shown). mDLK-Fc may not have a binding affinity that can detect the receptor. A second possibility is that mDLK-Fc regulates cell-surface c-kit expression on bone marrow cells. This possibility can be tested using anti-c-kit mAb such as ACK-2 [49]. mDLK-Fc did not affect c-kit expression on the SCF- and IL-3-cultured bone marrow cells (data not shown). A third possibility is that mDLK-Fc induces Hes-1 activity. Hes-1 is a DNA-binding transcription repressor in mammalian cells [50], and it represses neuronal differentiation when expressed ectopically in the mammalian central nervous system [50]. Moreover, it is shown that Notch signaling initiates expression of Hes-1 [6, 14]. Therefore, it is important to clarify whether the dlk, which is homologous to Notch ligand, induces Hes-1 expression. From the present findings of colony formation using Hes–/– fetal liver cells in the presence of mDLK-Fc, the suppressive effect of dlk did not depend on the Hes-1 pathway. These findings suggest that erythroids, myeloids, and monocytes are normally generated in Hes–/– mice, and B-lineage cell generation depends on stromal cells from Hes–/– fetal liver cells [15]. However, recent studies have reported that dlk/Pref-1-Fc stimulated increased thymocyte expression of Hes-1 transcription factor, and that Hes-1-null thymocytes did not respond to dlk/Pref-1-Fc [33]. These findings appear contradictory to the present observations that showed dlk signaling was independent of Hes-1 transcription factor. To better understand the difference it will be necessary to identify the molecular nature of the dlk receptor.

The present findings were different from those reported by Moore et al. [26]. They concluded that soluble dlk protein had no effect on colony formation in cytokine-rich methylcellulose culture. These differences might be due to the construction of the soluble dlk protein, such as monomeric (CH2) fusion or dimeric (hinge) fusion of immunogloblin-like molecules. The present soluble protein made from hinge vector is a homodimer, but the soluble dlk protein, made from CH2 vector, was a monomeric form [34]. Recently, the dlk/Pref-1-Fc dimmer increased the numbers of developing thymocytes in fetal thymus organ culture, whereas the dlk/Pref-1 monomer decreased the cell numbers [33]. Furthermore, dimmer formation of cytokines including IL-5, platelet-derived growth factor and TGF-β is essential for biological activity [51-53]. Therefore, the homodimer of the soluble dlk protein may have a higher specific activity than the monomeric form.

In conclusion, we showed novel function of dlk in hematopoiesis. dlk inhibited SCF-induced colony formation. Furthermore, this inhibitory effect was independent of Hes-1 signaling. Attempts to identify the molecular signals for dlk are currently under investigation.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank the following investigators for generously supplying materials and reagents: Dr. T. Nakahata, for G-CSF; Dr. T. Kina, for TER119; Dr. N. Minato, for RA3-6B2. We also thank Dr. T. Suda for discussion and critically reading the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References