Apoptosis and C/EPBε in Mice Myeloid Cells

C/EBPε −/− mice: increased rate of myeloid proliferation and apoptosis

Article metrics


CCAAT/enhancer binding protein epsilon (C/EBPε) is essential for terminal granulocytic differentiation. Its expression begins at the transition between the proliferative and non-proliferative compartments of myelopoiesis. We studied the effect of targeted disruption of the C/EBPε gene on murine myeloid proliferation and apoptosis. Bone marrow cellularity of C/EBPε −/− and wild-type mice was 95% and 65%, respectively. The C/EBPε −/− mice had an expansion in the number of their CFU-GM/femur. The number of myeloid committed progenitor cells in the peripheral blood and the spleen of these mice was also increased. Bromodeoxyuridine (BrDU) pulse labeling studies demonstrated that the fraction of actively proliferating cells was two-fold higher in the bone marrow of C/EBPε −/− mice. However, the number of myeloid colonies arising from purified Sca-1+/lin− early hematopoietic progenitor cells and from bone marrow mononuclear cells grown in different cytokine combinations was not significantly different between wild-type and knock-out mice. Also, long-term marrow growth, and CFU were not different between the wild-type and C/EBPε −/− mice. The sensitivity to induction of apoptosis in the committed progenitor cell compartment after either withdrawal of growth factor or brief exposure to etoposide was normal. However, Gr-1 antigen-positive C/EBPε −/− granulocytic cells showed an increased rate of apoptosis in comparison to their wild-type counterparts. In summary, the myeloid compartment appears to be expanded in mice lacking C/EBPε. However, this is not the consequence of an intrinsic myeloproliferation but due to an indirect, possibly cytokine-mediated stimulation of myelopoiesis in vivo. C/EBPε may have a role in the inhibition of apoptosis in maturing granulocytic cells.


Transcription factor CCAAT enhancer binding protein epsilon (C/EBPε) plays a crucial role in terminal granulocytic differentiation.12 Mice with a targeted disruption of the C/EBPε locus lack mature granulocytes. The most differentiated C/EBPε −/− granulocytic cells are dysplastic and not fully functional. They have a significantly reduced capacity to produce superoxide and an impaired migratory function.1 Expression of C/EBPε is nearly limited to the granulocytic pathway, levels of the protein become prominent at the promyelocyte stage345 and continue to be expressed in the most mature stages of granulocytopoiesis. In mature neutrophils, C/EBPε is involved in the regulation of genes important for antimicrobial defense such as the cathelin-related protein (murine homolog of hCAP18) and lactoferrin.678 Due to these defects in host defense, the animals die at the age of 3–5 months of gram-negative sepsis. Upregulation of C/EBPε in terminal granulocytic differentiation is associated with a loss of proliferation. Interestingly, C/EBPα is known to be crucial for the loss of proliferation in differentiating hepatocytes and the upregulation of the cyclin-dependent kinase inhibitor, p21/waf1.910 The overexpression of factors associated with proliferation, such as c-myc, antagonizes myeloid differentiation. C/EBPε restores the ability of 32D cells, which overexpress c-myc, to differentiate.11 Hence, C/EBPε may play an active role in the pathways integrating proliferation and differentiation in granulocytopoiesis. Alternatively, C/EBPε induces the expression of target genes such as lactoferrin in mature granulocytes which may be involved in the negative (feedback) regulation of myelopoiesis.1213 The bone marrow morphology of C/EBPε −/− mice suggested the coexistence of increased proliferation and apoptosis, with nearly 100% cellularity on biopsy specimen of the marrow and prominent pseudo-Gaucher cells throughhout the marrow. We, therefore, investigated whether myeloid cells in C/EBPε −/− mice show abnormalities in their proliferative behavior in vivo and in vitro as well as in their response to apoptotic stimuli.

Materials and methods


Homozygous C/EBPε −/− mice were established in a mixed 129/SvEv × NIH Black Swiss background by Yamanaka et al.1 C/EBPε −/− and control C/EBPε +/+ 129/SvEv × NIH Black Swiss mice were kindly provided by Dr K Xanthopoulos and bred under sterile conditions in the animal housing facility at Cedars-Sinai Medical Center. For all experiments, age matched littermates were used. Mice were sacrificed by cervical neck dislocation. Bone marrow was flushed from isolated femurs and tibiae with IMDM + 20% FBS using a No. 26 gauge needle into plastic culture dishes. Cells were incubated for 1 h at 37°C in an humidified atmoshere with 5% CO2 in air. Non-adherent cells were either used directly for colony-forming assays or after Ficoll–Hypaque density gradient centrifugation to generate mononuclear cells. Neutrophils were harvested 12 h after intraperitoneal injection with 2 ml 4% thioglycollate by peritoneal lavage with sterile PBS.

Cell cycle analysis

Mononuclear bone marrow cells were prepared and resuspended in 1 ml PBS. Two ml ice-cold methanol was added dropwise for fixation. Prior to cell cycle analysis, cells were treated with RNAse H and resuspended in PBS containing 5 μg/ml propidiumiodide.

BrdU incorporation

Mononuclear bone marrow cells of C/EBPε −/− mice and controls were incubated in IMDM + 20% FBS supplemented with 10 μM bromodeoxyuridine for 1 h. Cells were washed twice in cold PBS and resuspended in 100 μl PBS. Cells were added dropwise to 5 ml ice-cold 70% ethanol and incubated for 30 min. After fixation, cells were incubated in 1 ml of 2 N HCl/Triton X-100 for 30 min followed by neutralization in 1 ml of 0.1 M Na2B4O7. Cells were then centrifuged and 106 cells were resuspended in 50 μl Tween 20/BSA/PBS. For direct immunofluorescence, 20 μl of anti-BrdU FITC antibody (Becton Dickinson, Heidelberg, Germany) was used. After 30 min incubation at room temperature, cells were centrifuged and resuspended in 1 ml of PBS containing 5 μg/ml propidium iodide (PI).

Cell sorting

A FITC-conjugated antibody to Sca1, and phycoerythrin-conjugated antibodies to Gr-1, CD11b, B220 and CD5 were purchased from Pharmingen, San Diego, CA,USA. 5 × 107 bone marrow mononuclear cells were isolated by Ficoll–Hypaque density gradient centrifugation. These cells were preincubated with FcBlock (Pharmingen) for 15 min on ice. Titrated amounts of Sca-1, Gr-1, CD11b, B220, CD5 antibodies were added and cells were incubated for an additional 15 min on ice. Then, cells were washed twice in cold PBS and resuspended in RPMI + 10% FBS for cell sorting. Sca+(FITC+)/lin-(phycoerythrin−) cells were sorted under sterile conditions using a FACS Star cell sorter into sterile 2 ml Falcon tubes prefilled with 1 ml warm IMDM plus 10% FBS.

Colony-forming assay

Bone marrow was prepared as described above. A total of 104 bone marrow cells (either non-fractionated, non-adherent cells or mononuclear cells) or 2 × 103 Sca1+/lin− sorted cells were added 1:10 to 2 ml Methocult M3234 (Stem Cell Technology, Vancouver, Canada). The final concentrations of methylcellulose and the supplements were: 1% methycellulose, 15% FBS, 1% BSA, 10 μg/ml bovine pancreatic insulin, 200 μg/ml iron saturated transferrin, 10−4 M 2-mercaptoethanol, and 2 mM L-glutamine. Recombinant cytokines were added as indicated in individual experiments. Each sample was plated in triplicate in 0.5 ml aliquots into a 24-well plate. Culture plates were incubated at 37°C in an humidified atmosphere with 5% CO2 in air. Colonies were counted on day 10. Numbers are given as means ± standard deviations.

Long-term bone marrow culture

The murine fibroblast cell line M2–10B4 was propagated in RPMI 1640 + 10% FBS. Confluent adherent cells were trypsinized, irradiated with 70 Gy and seeded at 5 × 104/cm2 into 25 cm2 Falcon culture flasks. Long-term cultures were initiated with 8 × 106 bone marrow mononuclear cells in 8 ml IMDM + 10% FBS, 10% pretested horse serum, 2 mM L-glutamine and 10−4 M hydrocortisone. The cultures were incubated at 33°C with 5% CO2 in air. Fifty percent of the medium was carefully removed once a week and replaced with fresh long-term culture medium. Colony assays were performed with cells in the supernatant.

Annexin V staining

Granulocytes/monocytes isolated by peritoneal lavage at 12 h after intraperitoneal injection of 4% thioglycollate were stained immediately with a phycoerythrin-conjugated antibody to the granulocytic differentiation antigen Gr-1, and a FITC-conjugated antibody to Annexin V and 7-AAD (Viaprobe (Pharmingen)). The percentage of apoptotic (annexin V-positive/7-AAD-negative) and dead (annexin V-positive/7-AAD-positive) cells was measured by FACS analysis. The percentage was determined in the Gr-1-positive population.

Cytokine ELISA (murine TNFα, GM-CSF)

Murine TNF and GM-CSF were measured by two-antibody ELISA using biotin–streptavidin–peroxidase detection. Polystyrene plates (Maxisorb; Nunc, Rochester, NY, USA) were coated with capture antibody in PBS overnight at 25°C. The plates were washed four times with 50 nM Tris, 0.2% Tween 20, pH 7.0–7.5 and then blocked for 90 min at 25 °C with assay buffer (PBS containing 4% BSA (Sigma, St Louis, MO, USA) and 0.01% Thimerosal, pH 7.2–7.4). The plates were washed four times, and 50 μl assay buffer was added to each well along with 50 μl of sample or standard prepared in assay buffer and incubated at 37°C for 2 h. The plates were washed four times and 100 μl of biotinylated detecting antibody in assay buffer was added and incubated at 25°C for 30 min. After washing the plate four times, streptavidin–peroxidase polymer in casein buffer (RDI) was added and incubated at 25°C for 30 min. The plate was washed four times and 100 μl commercially available substrate (TMB, DAKO, Carpinteria, CA, USA) was added and incubated at 25°C for approximately 10–30 min. The reaction was stopped with 100 μl 2 N HCl and the A 450 (minus A 650) was read on a microplate reader (Molecular Dynamics, Piscataway, NJ, USA). A curve was fitted to the standards using a computer program (Soft Pro, Molecular Dynamics) and the cytokine concentration in each sample was calculated from the standard curve equation. The ranges of the assays were 15.6 pg/ml to 1000 pg/ml for TNF and 12.5 pg/ml to 800 pg/ml for GM-CSF.


Descriptive statistics (mean, standard deviation (s.d.)) and statistical comparisons between groups with a two-tailed Student's t-test were performed using the Instat 2 software (Graphpad, San Diego, CA, USA). Cell cycle data were evaluated with a univariate analysis considering age and group (C/EBPε −/− vs C/EBPε +/+) as influencing factors.



The visual estimate of bone marrow cellularity of C/EBPε −/− and wild-type mice was 95% and 65%, respectively. The composition of myeloid bone marrow cells differed between C/EBPε −/− and wild-type mice. C/EBPε −/− mice had a left shift with an increased percentage of myeloblasts (1.5% vs 1% in normals), promyelocytes (2% vs 1%), myelocytes (15% vs 8%), metamyelocytes (31% vs 9%) and band forms (19% vs 18%) (Ref. 1 and Figure 1). Fully mature granulocytes were missing from the bone marrow (0 vs 22%) and peripheral blood of C/EBPε −/− mice (Figures 1a–c). Granulocytic cells in the peripheral blood showed a dysplastic morphology. The percentage of nucleated red cells was decreased (18 vs 25%) in the knock-out mice. The percentage of nucleated red cells was very much age-dependent and decreased in both wild-type and knock-out mice with age. The myeloid:erythroid (M:E) ratio increased from 1:1.5 at the age of 3 days to 3:1 at the age of 3 months in the wild-type and up to 4:1–5:1 in the knock-out mice. The percentage of mature Gr-1 positive cells in the bone marrow was 15–20% lower in the knock-out animals than in the wild-type animals (Table 1). The mean white cell count in the peripheral blood of young C/EBPε −/− mice (3 days to 3 weeks) was 1.7-fold higher than age-matched controls; the mean WBC of older knock-out animals (4 weeks to 4 months) was 2.4-fold higher than wild-type controls (Table 1). This increase was due to a higher absolute number of granulocytic lineage cells in the peripheral blood. The mean percentage of granulocytes was 12% in wild-type animals vs 24–39% dysplastic granulocytic cells in knock-out animals.

Figure 1

 (a) Myeloperoxidase (MPO) immunohistochemistry of C/EBPε −/− and wild-type bone marrow (femur of 2-month-old mice). C/EBPε −/− bone marrow is hypercellular which is mainly due to myeloid hyperplasia. (b) C/EBPε −/− bone marrow cells contain a higher percentage of mitotic cells (1%) than wild-type controls (0.3%) (Wright-Giemsa stain). (c) Comparison of granulocyte morphology in the peripheral blood (Wright-Giemsa stain).

Table 1  Comparison between wild-type and C/EBPε −/− age-dependent cell cycle distribution and immunophenotype of total nucleated bone marrow cells and peripheral blood white blood cell counts

Cell cycle analyses of bone marrow cells from C/EBPε −/− and wild-type mice were performed. To reflect their in vivo proliferative status, bone marrow cells were fixed in methanol immediately after their harvest. Figure 2a shows the cell cycle distribution of marrow cells from C/EBPε −/− and wild-type mice, both 3 weeks old. Furthermore, Table 1 summarizes a comparison of cell cycle data of marrow cells from wild-type and C/EBPε −/− bone marrows from mice of different ages (n = 3 for each group). All samples were also stained with antibodies directed against the granulocytic antigen Gr-1 and the myeloid marker CD11b to estimate the number of myeloid cells and their degree of maturity. In both wild-type and C/EBPε −/− mice, the percentage of cells in S phase decreased with age along with an increase of myeloid cells. This result represents the age-dependent decrease of erythroid precursor cells in wild-type and knock-out bone marrow and their decreasing contribution to S-phase. However, consistent with our morphological analysis, we found an increased number of hematopoietic cells in S-phase in the knock-out mice at all ages that were examined. Knock-out animals of all ages showed a 50–100% increase of cells in S-phase compared to age-matched controls. These results were confirmed by pulse-labeling studies with bromodeoxyuridine (BrDU) which also showed a higher percentage of cells undergoing DNA replication in the C/EBPε −/− mice (Figure 2b). To exclude an intrinsic proliferative capacity of more mature myeloid cells, we performed a cell cycle analysis of nucleated peripheral blood cells; proliferating cells were not observed (data not shown).

Figure 2

 (a) Cell cycle analysis of total nucleated bone marrow cells from C/EBPε +/+ and C/EBPε −/− mice. The comparison of the cell cycle distributions shows that C/EBPε −/− mice have a higher percentage of cells in the S-phase of the cell cycle (32% vs 23%). (b) Comparison of bromodeoxyuridine (BrDU) incorporation during a 1 h labeling period in C/EBPε +/+ and C/EBPε −/− bone marrow cells. The BrDU incorporation was almost two-fold higher in C/EBPε −/− bone marrow cells.

The number of myeloid colonies derived from total nucleated bone marrow cells was 1.5 to two-fold higher in the knock-out as compared to the wild-type mice (Table 2). Since the total number of nucleated bone marrow cells isolated from a femur of the same size and age mouse was approximately 1.5-fold higher in the knock-out compared to the wild-type mice (2.4 ± 0.38 × 106 vs 1.5 ± 0.35 × 106, respectively), reflecting the hypercellular bone marrow, the absolute number of myeloid progenitors was estimated to be two- to three-fold higher in the bone marrow of C/EBPε −/− mice. Interestingly, the number of CFU-GM derived from the peripheral blood and the spleen was also significantly higher in C/EBPε −/− mice, an additional sign of an expanded pool of myeloid progenitors (Table 2).

Table 2  Wild-type vs C/EBPε −/− mice: myeloid progenitor cell growth

Also, we tested the in vivo response to injections of G-CSF. We administered i.p. 250 μg/kg G-CSF/day into three wild-type and three C/EBPε −/− mice for 2 weeks; the dose was increased to 750 μg/kg/day for the following 2 weeks. White blood cell counts (WBC) were determined on days 8, 10, 12, 15, 17, 20, 22, 24, 27, 30. (Figure 3). As described above, C/EBPε −/− mice had a slightly higher WBC than wild-type mice in their steady state. However, in the experiment depicted here, the difference between the groups at baseline did not reach statistical significance. Both wild-type and knock-out mice had an increase in white blood cell counts with G-CSF treatment. After 7 days G-CSF treatment, the C/EBPε −/− cohort showed significantly higher WBC than wild-type animals (P < 0.01; Figure 3). These results are consistent with an increased pool of committed, G-CSF responsive, progenitor cells in C/EBPε −/− mice. At later time points, variation within the cohorts was high and no significant differences were observed.

Figure 3

 Peripheral white blood cell counts (WBC) of C/EBPε +/+ and C/EBPε −/− mice (n = 3) before and during injection of human recombinant G-CSF at 250 μg/kg for days 1–14 and 750 μg/kg for days 15–30. Means and standard deviations are plotted.

Suprisingly, we did not observe a significant difference in myeloid colony formation between wild-type and knock-out using bone marrow mononuclear cells after Ficoll–Hypaque density gradient centrifugation for colony-forming assays (Table 2). A comparison of colony growth in various growth factor combinations including G-CSF, M-CSF, GM-CSF alone or in combination with stem cell factor, IL-3 and IL-6 did not show any significant differences in colony numbers between wild-type and knock-out bone marrow mononuclear cells (Figure 4). To better understand the influence of Ficoll separation on myeloid colony growth, we studied the composition of wild-type and C/EBPε −/− bone marrow cells before and after the procedure. Cytospin preparations demonstrated that in wild-type bone marrow, immature myeloid cells were enriched by Ficoll separation since granulocytes and bands were removed: 16 ± 5% residual granulocytes, bands and metamyelocytes after Ficoll vs 49% without Ficoll. Dysplastic C/EBPε −/− metamyelocytes/bands were very inefficiently removed by Ficoll separation. In comparison to a mean of 50% bands and metamyelocytes without Ficoll we counted 47 ± 8% residual dysplastic metamyelocytes and bands after Ficoll. The discrepancy between our results with total bone marrow nucleated cells and bone marrow mononuclear cells was therefore, at least in part, due to a difference in the composition of C/EBPε −/− and wild-type bone marrow cells.

Figure 4

 Myeloid colony formation of 104 C/EBPε +/+ and C/EBPε −/− bone marrow mononuclear (MNC) cells with different cytokine combinations: 100 ng/ml recombinant, murine GM-CSF; 100 ng/ml recombinant human G-CSF; 100 ng/ml recombinant murine M-CSF; 50 ng/ml recombinant murine stem-cell factor; 10 ng/ml interleukin 3; 10 ng/ml interleukin 6.

Also, no difference was observed in colony numbers between C/EBPε −/− and wild-type mice when sorted Sca1+/lin− early hematopietic stem cells were used in the colony-forming assays (Table 2).

Cell survival and apoptosis

The composition of bone marrow cells is determined not only by proliferation and differentiation but also by the survival of different myeloid cell populations. We determined the rate of apoptosis in response to apoptotic stimuli and the percentage of apoptotic cells in the mature myeloid compartment.

To study progenitor cell survival, we initiated long-term cultures with bone marrow of wild-type and knock-out mice. The median decline in colony numbers during the first 3 weeks of culture was identical (Figure 5). We also tested the sensitivity of bone marrow cells to exposure to etoposide, a known apoptosis inducing agent. Wild-type and C/EBPε −/− bone marrow cells (106) were incubated for 4 h in various concentrations of etoposide, washed thoroughly and plated in methylcellulose with SCF, IL-3, IL-6 and GM-CSF. Colony formation was determined on day 10. No difference was found between wild-type and knock-out cells in respect of their sensitivity towards etoposide (Figure 6).

Figure 5

 Parallel decline of C/EBPε +/+ and C/EBPε −/− committed progenitor cells (CFU) in the long-term bone marrow culture (LTC) system. Long-term cultures were initiated with 8 × 106 bone marrow mononuclear cells in 8 ml IMDM + 10% FBS, 10% pretested horse serum, L-glutamine and 10−4 M hydrocortisone. The cultures were incubated at 33°C with 5% CO2 in air. Fifty percent of the medium was carefully removed once a week and replaced with fresh long-term culture medium. Colony assays were performed with all cells in the supernatant in 1% methylcellulose, 15% FBS, 1% BSA, 10 μg/ml bovine pancreatic insulin, 200 μg/ml iron saturated transferrin, 10−4 M 2-mercaptoethanol, 2 mM L-glutamine, 50 ng/ml recombinant murine stem cell factor; 10 ng/ml interleukin 3; 10 ng/ml interleukin 6 as described above. Colony numbers/ml supernatant were calculated. The figure summarizes data from four independent experiments, eg bone marrow samples from four wild-type and four knock-out mice. The natural logarithm of the mean colony number/ml was plotted vs the time after initiation of the LTC.

Figure 6

 Total bone marrow cells isolated from the femur of C/EBPε +/+ and C/EBPε −/− mice were incubated for 4 h in different concentrations of etoposide, washed thoroughly and plated in methylcellulose with 50 ng/ml murine recombinant stem cell factor, 100 ng/ml interleukin 3, and 10 ng/ml interleukin 6. Myeloid colony growth was scored on day 10.

To study the survival of more mature myeloid cells, we collected Gr-1-positive C/EBPε −/− and wild-type cells from the peritoneal cavity 12 h after the i.p. injection of 4% thioglycollate. Immediately after isolation, the percentage of apoptotic and dead cells was determined by triple immunofluorescence staining with Annexin V FITC, 7-AAD and the phycoerythrin-conjugated surface marker Gr-1. The samples were immediately analyzed in a FACScan. As described previously, C/EBPε −/− granulocytic cells showed a markedly reduced expression of the Gr-1 antigen.1 Although the rate of apoptosis is normally high in a population of short-lived mature granulocytes in wild-type mice (here: mean, three mice: 42% ± 7 apoptotic cells), the percentage of apoptotic cells was significantly higher in the Gr-1-positive cell compartment from the C/EBPε −/− mice (mean, three mice: 69% ± 12 apoptotic cells, P < 0.01) (Figure 7a). The morphological examination of the cells supported the notion of a very high rate of apoptosis with nuclear condensation and karyorhexis in the population of granulocytic cells in the C/EBPε −/− mice (Figure 7b).

Figure 7

 Comparison of the rate of apoptotic cell death between wild-type (C/EBPε +/+) and C/EBPε −/− granulocytes. (a) C/EBPε +/+ and C/EBPε −/− phagocytes were recruited to the peritoneal cavity by i.p. injection of 4% thioglycollate. Cells were harvested after 12 h and immediately stained with Gr-1PE, Annexin V FITC and 7-AAD. The analysis of the Gr-1-positive population showed an increased percentage of apoptotic cells (annexin-V-positive, 7-AAD-negative cells) for the C/EBPε −/− mice: 69% C/EBPε −/− vs 42% wild-type (SSC: side scatter). (b) Wright-Giemsa stain of peritoneal phagocytes demonstrated the high percentage of apoptotic, dysplastic granulocytes in C/EBPε −/− mice.

Cytokine levels

An increased myeloid proliferation may be mediated indirectly by a reactive cytokine cascade. We examined levels of serum GM-CSF and TNFα in the serum of six wild-type and 10 C/EBPε −/− mice of different ages (7 days to 4½ months). None of the mice had a detectable serum level of GM-CSF (lower detection limit: 12.5 pg/ml). None of the controls showed detectable levels of TNFα. Only one C/EBPε −/− mouse (age 4½ months) had an increased serum TNFα level: 51 pg/ml (lower detection limit: 15 pg/ml).


Our findings indicate a high turnover in the myeloid cell compartment in C/EBPε −/− mice. An increased proliferation rate is combined with an increased rate of apoptosis of the more mature granulocytic cells. Our in vitro data do not support the notion of an intrinsic myeloproliferation. Purified early hematopoietic progenitor cells from C/EBPε −/− mice produced the same number of myeloid colonies in the presence of IL-3, IL-6, SCF ± GM-CSF as the wild-type progenitors. Also, no apparent difference in colony size was noted. C/EBPε −/− progenitor cells have the same survival as the wild-type progenitors after either growth factor withdrawal or short-term incubation with etoposide. Furthermore, C/EBPε −/− granulocytic progenitor cells showed a normal proliferative response to G-CSF in vivo and in vitro. This is consistent with the expression of the receptor for G-CSF on granulocytic progenitor cells from C/EBPε −/− cells.1

Interestingly, we have found an increased absolute number of both nucleated cells and myeloid progenitors in the bone marrow, spleen and the peripheral blood, reflecting an expanded myeloid progenitor cell pool. Since immature myeloid progenitors do not express C/EBPε, an intrinsic alteration of cell adhesion molecules and a subsequent redistribution of progenitors is an unlikely cause for this finding.2

Various possibilities could explain a reactive in vivo expansion of myelopoiesis in C/EBPε −/− mice. One possibility might be stimulation from chronic endotoxin exposure as a consequence of the antimicrobial defect of the neutrophils.1415 We studied the possibility that bacterial infections (LPS) may trigger TNFα secretion in C/EBPε −/− monocytes which could lead to increased serum G-CSF and/or GM-CSF levels.1617 Only one (4½ month old) out of 10 mice had increased TNFα serum levels. None had increased GM-CSF levels. In addition, very young C/EBPε −/− mice raised in a germ-free environment did not appear to get sick, but they already had an increased number of proliferating bone marrow cells. Infections appear to be a late complication in the knock-out mice (approximately at 4 months), not the driving force of myeloproliferation.

Granulocyte colony-stimulating factor (G-CSF) is known to be the most important factor for the homeostasis of in vivo granulocytopoiesis.18 Endogenous G-CSF levels, which have not been studied here, could be increased due to LPS-mediated stimulation of G-CSF secretion by monocytes or potentially due to decreased G-CSF plasma clearance, possibly resulting from the absence of mature granulocytes.1920

An interesting speculation is that C/EBPε −/− mice lack an inhibitory factor normally expressed by mature granulocytes. As recently reported, C/EBPε −/− granulocytes lack expression of secondary granule products including lactoferrin because transcriptional expression of these genes requires the transactivation by C/EBPε.56 A provocative series of experiments by Broxmeyer et al21222324 showed that lactoferrin was an inhibitory factor for myelopoiesis in vitro and in vivo. Lactoferrin was shown to inhibit myeloid colony-forming cells and suppress GM-CSF and IL-1 production by a subpopulation of monocytes/macrophages. IL-1 can induce GM-CSF expression in fibroblasts.2526 However, GM-CSF serum levels were not increased in C/EBPε −/− mice which conflicts with this speculation.

An alternative possibility relies on the fact that 19 chemokines have been reported to inhibit myelopoiesis,2728 including monocyte–chemotactic peptide-1 (MCP-1), macrophage inflammatory protein (MIP)-1 alpha, MIP-related protein 2 (MRP2) and interleukin 8. Some of these have a direct inhibitory effect on the growth of myeloid progenitors.2930313233 We recently showed that the targeted disruption of C/EBPε also affects monocyte function.6 MCP and MIP-1 alpha have been implicated as target genes of C/EBPε in monocytes.34 A reduced or absent expression of these myelosuppressive cytokines or chemokines by C/EBPε −/− monocytes/macrophages could be an additional or alternative mechanism for the enhanced myelopoiesis in C/EBPε −/− mice.

Mature C/EBPε −/− granulocytic cells rapidly underwent apoptosis when recruited to the peritoneal cavity. The higher absolute number of granulocytic cells in the peripheral blood of C/EBPε −/− mice, however, demonstrates that the net output of immature granulocytes is increased. This situation contrasts to myelodysplasia which is associated with intramedullary apoptosis of CD34-positive progenitor cells resulting in cytopenias of the peripheral blood. We hypothesize that C/EBPε −/− granulocytic cells undergo apoptosis after activation by inflammatory stimuli such as thioglycollate in the peritoneal cavity or repeated exposure to gram-negative bacteria. Further studies in progress using subtractive libraries, representational difference analysis (RDA), and chip analysis using nucleotide arrays suggest that C/EBPε expression is associated with the induction of expression of several antiapoptotic proteins and absence of expression of several chemokines.


  1. 1

    Yamanaka R, Barlow C, Lekstrom-Himes J, Castilla LH, Liu PP, Eckhaus M, Decker T, Wynshaw-Boris A, Xanthopoulos KG . Impaired granulopoiesis, myelodysplasia and early lethality in CCAAT/enhancer binding protein ε-deficient mice Proc Natl Acad Sci USA 1997 94: 13187–13192

  2. 2

    Chumakov AM, Grillier I, Chumakowa E, Chih D, Slater J, Koeffler HP . Cloning of the novel human myeloid-cell-specific C/EBPε transcription factor Mol Cell Biol 1997 17: 1375–1386

  3. 3

    Morosetti R, Park DJ, Chumakow AM, Grillier I, Shiohara M, Gombart AF, Nakamaki T, Weinberg K, Koeffler HP . A novel, myeloid transcription factor, C/EBP epsilon, is upregulated during granulocytic, but not monocytic differentiation Blood 1997 90: 2591–2600

  4. 4

    Antonson P, Stellan B, Yamanaka R, Xanthopoulos KG . A novel human CCAAT/enhancer binding protein gene, C/EBPε is expressed in cells of lymphoid and myeloid lineages and is localized on chromosome 14q11.2 close to the T-cell receptor α/δ locus Genomics 1996 35: 30–38

  5. 5

    Chih DY, Chumkov AM, Park DJ, Silla AG, Koeffler HP . Modulation of mRNA expression of a novel human myeloid selective CCAAT/enhancer binding protein gene (C/EBP epsilon) Blood 1997 90: 2987–2994

  6. 6

    Lekstrom-Himes J, Xanthopoulos KG . CCAAT/enhancer binding protein epsilon is critical for effective neutrophil-mediated response to inflammatory challenge Blood 1999 93: 3096–3105

  7. 7

    Verbeek W, Gombart AF, Chumakov AM, Müller C, Friedman AD, Koeffler HP . C/EBPε directly interacts with the DNA binding domain of c-myb and cooperatively activates transcription of myeloid promoters Blood 1999 93: 3327–3337

  8. 8

    Verbeek W, Lekstrom-Himes J, Park DJ, Dang PM, Vuong PT, Kawano S, Babior BM, Xanthopolous K, Koeffler HP . Myeloid transcription factor C/EBPε is involved in the positive regulation of lactoferrin gene expression in neutrophils Blood 1999 94: 3141–3150

  9. 9

    Timchenko NA, Harris TE, Wilde M, Bilyeu TA, Burgess-Beusse BL, Finegold MJ, Darlington GJ . CCAAT/enhancer binding protein alpha regulates p21 protein and hepatocyte proliferation in newborn mice Mol Cell Biol 1997 17: 7353–7361

  10. 10

    Timchenko NA, Wilde M, Nakanishi M Smith JR, Darlington GJ . CCAAT/enhancer-binding protein alpha(C/EBPα) inhibits proliferation through the p21 (WAF-1/CIP-1/SDI-1) protein Genes Dev 1996 10: 804–815

  11. 11

    Nakajima H, Cleveland JL, Nagata S, Ihle JN . Granulocyte colony stimulating factor regulates myeloid differentiation through CCAAT enhancer binding protein epsilon Blood 1998 92: (Suppl. 1) 713a

  12. 12

    Broxmeyer HE, Smithyman A, Egerr RR, Meyers PA, de Sousa M . Identification of lactoferrin as the granulocyte-derived inhibitor of colony stimulating activity production J Exp Med 1978 148: 1052–1067

  13. 13

    Gentile P, Broxmeyer HE . Suppression of mouse myelopoiesis by administration of human lactoferrin in vivo and the comparative action of human transferrin Blood 1983 61: 982–993

  14. 14

    Kawakami M, Tsutsumi H, Kumakawa T, Hirai M, Kurosawa S, Mori M, Fukushima M . Serum granulocyte colony stimulating factor in patients with repeated infections Am J Hematol 1992 41: 190–193

  15. 15

    Dale DC, Lau S, Nash R, Boone T, Osborne W . Effect of endotoxin on serum granulocyte and granulocyte–macrophage colony stimulating factor levels in dogs J Infect Dis 1992 165: 689–694

  16. 16

    Munker R, Gasson J, Ogawa M, Koeffler HP . Recombinant human TNF induces production of granulocyte–monocyte colony stimulating factor Nature 1986 323: 79–82

  17. 17

    Furman WL, Lumm W Arnold B, Pratt CB, Ihle JN . Recombinant human TNFα stimulates the secretion of granulocyte colony-stimulating factor in vivo Leukemia 1992 6: 319–322

  18. 18

    Lieschke GJ, Burgess AW . Granulocyte colony colony-stimulating factor and granulocyte-macrophage colony stimulating factor New Engl J Med 1992 327: 28–35

  19. 19

    Layton JE, Hockman H, Sheridan WP, Morstyn G . Evidence for a novel in vivo control mechanism of granulopoiesis: mature cell-related control of a regulatory growth factor Blood 1989 74: 1303–1307

  20. 20

    Terashi K, Oka M, Ohdo S, Furukubo T, Ikeda C, Fukuda M, Soda H, Higuchi S, Kohno S . Close association between clearance of recombinant human granulocyte colony stimulating factor (G-CSF) and G-CSF receptor on neutrophils in cancer patients Antimicrob Agents Chemother 1999 43: 21–24

  21. 21

    Broxmeyer HE, Williams DE, Hangoc G, Cooper S Gentile P . The opposing actions on murine myelopoiesis of purified preparations of lactoferrin and the colony stimulating factors Blood Cells 1987 13: 31–48

  22. 22

    Sledge GW Jr, Bicknell DC, Harris EL, Zegarra G, Broxmeyer HE . Monoclonal antibody II2c to human lactoferrin inactivates the myelopoiesis suppressive effect of human lactoferrin in vitro Exp Hematol 1986 14: 333–337

  23. 23

    Broxmeyer HE, Williams DE, Boswell HS, Cooper S, Shadduck RK, Gillis S, Waheed A, Urdal DL . The effects in vivo of purified preparations of murine macrophage colony stimulating factor-1, recombinant granulocyte–macrophage colony stimulating factor and natural and recombinant murine interleukin 3 without and with pretreatment of mice with iron saturated human lactoferrin Immunobiology 1986 172: 168–174

  24. 24

    Broxmeyer HE, Bicknell DC, Gillis S, Harris EL, Pelus LM, Sledge GW Jr . Lactoferrin: affinity purification from human milk and polymorphonuclear neutrophils using monoclonal antibody (II2C) to human lactoferrin, development of an immunoradiometric assay using II2C and myelopoietic regulation and receptor binding characteristics Blood Cells 1986 11: 429–446

  25. 25

    Broxmeyer HE, Platzer E . Lactoferrin acts on Ia and IE/C antigen + subpopulations of mouse peritoneal macrophages in the absence of T-lymphocytes and other cell types to inhibit production of granulocyte-macrophage colony stimulating factors in vitro J Immunol 1984 133: 306–314

  26. 26

    Zucali JR, Broxmeyer HE, Levy D, Morse C . Lactoferrin decreases monocyte-induced fibroblast production of myeloid colony stimulating activity by suppressing monocyte release of interleukin 1 Blood 1989 74: 1531–1536

  27. 27

    Broxmeyer HE, Kim CH . Regulation of hematopoiesis in a sea of chemokine family members with a plethora of redundant activities Exp Hematol 1999 27: 1113–1123

  28. 28

    Broxmeyer HE, Kim CH, Cooper SH, Hangoc G, Hromas R, Pelus LM . Effects of CC, CXC, C and CX3C chemokines on proliferation of myeloid progenitor cells, and insights into SDF1 induced chemotaxis of progenitors Ann NY Acad Sci 1999 872: 142–162

  29. 29

    Maze R . Myelosuppressive effects of purified recombinant murine MIP 1 alpha J Immunol 1992 149: 1004–1009

  30. 30

    Youn BS, Jang IK, Broxmeyer HE, Cooper S, Jenkins NA, Gilbert DJ, Copeland NG, Elick TA, Fraser MJ Jr, Kwon BS . A novel chemokine, macrophage inflammatory related protein 2 inhibits colony formation of bone marrow myeloid progenitors J Immunol 1995 155: 2661–2667

  31. 31

    Broxmeyer HE, Cooper S, Calcalano G, Hague NL, Bailish E, Moore MW . Involvement of interleukin (IL)-8 receptor in negative regulation of myeloid progenitor cells in vivo: evidence from mice lacking the murine IL-8 receptor homologue J Exp Med 1996 184: 1825–1832

  32. 32

    Broxmeyer HE, Cooper S, Hanggoc G, Gao JL, Murphy PM . Dominant myelopoietic effector functions mediated by chemokine receptor CCR1 J Exp Med 1999 189: 1987–1992

  33. 33

    Reid S, Ritchie A, Boring L, Gosling J, Cooper S, Hangoc G, Charo IF, Broxmeyer HE . Enhanced myeloid progenitor cell cycling in mice lacking the chemokine receptor CCR2 Blood 1999 93: 1524–1533

  34. 34

    Williams SC, Yang D, Schwartz RC, Weiler SR, Ortiz M, Keller JR, Johnston PF . C/EBPε is a myeloid-specific activator of cytokine, chemokine, and macrophage colony-stimulating factor receptor genes J Biol Chem 1998 273: 13493–13501

Download references


This work was in part supported by the following NIH grants: NIH Genetic Core, Lymphoma Foundation of America, Parker Hughes Fund and C and H Koeffler Grant. Walter Verbeek was supported by a Grant of the Deutsche Forschungsgemeinschaft.

Author information

Correspondence to HP Koeffler.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Verbeek, W., Wächter, M., Lekstrom-Himes, J. et al. C/EBPε −/− mice: increased rate of myeloid proliferation and apoptosis. Leukemia 15, 103–111 (2001) doi:10.1038/sj.leu.2401995

Download citation


  • CCAAT enhancer binding protein epsilon
  • myeloid proliferation
  • apoptosis

Further reading