Impaired myelopoiesis in mice devoid of interferon regulatory ... - Nature

Leukemia (2004) 18, 1864–1871 & 2004 Nature Publishing Group All rights reserved 0887-6924/04 $30.00 www.nature.com/leu

Impaired myelopoiesis in mice devoid of interferon regulatory factor 1 U Testa1, E Stellacci2, E Pelosi1, P Sestili3, M Venditti3, R Orsatti2, A Fragale2, E Petrucci1, L Pasquini1, F Belardelli3, L Gabriele3 and A Battistini2 1 Department of Hematology, Oncology and Molecular Medicine, Istituto Superiore di Sanita`, Rome, Italy; 2Department of Infectious, Parasitic and Immunomediated Diseases, Istituto Superiore di Sanita`, Rome, Italy; and 3Department of Cell Biology and Neurosciences, Istituto Superiore di Sanita`, Rome, Italy

Interferon regulatory factor (IRF)-1 is a transcription factor controlling the expression of several genes, which are differentially induced depending on the cell type and signal. IRF-1 modulates multiple functions, including regulation of immune responses and host defence, cell growth, cytokine signalling and hematopoietic development. Here, we investigated the role of IRF-1 in granulocytic differentiation in mice with a null mutation in the IRF-1 gene. We show that IRF-1/ bone marrow cells exhibit an increased number of immature granulocytic precursors, associated with a decreased number of mature granulocytic elements as compared to normal mice, suggestive of a defective maturation process. Clonogenetic analyses revealed a reduced number of CFU-G, CFU-M and CFU-GM colonies in IRF-1/ mice, while the number of BFU-E/CFU-E colonies was unchanged. At the molecular level, the expression of CAAT-enhancer-binding protein (C/EBP)-e, -a and PU.1 was substantially lower in the CD11b þ cells from the bone marrow of IRF-1/ mice as compared to cells from wild-type mice. These results, together with the fact that IRF-1 is markedly induced early during granulo-monocytic differentiation of CD34 þ cells, highlight the pivotal role of IRF-1 in the early phases of myelopoiesis. Leukemia (2004) 18, 1864–1871. doi:10.1038/sj.leu.2403472 Published online 16 September 2004 Keywords: hematopoiesis; transcription factors; granulocytes; gene expression; normal cell development

Introduction Pluripotent stem cells undergo progressive restriction in their lineage potential to give rise to mature, terminally differentiated cells. Transcription factors regulating lineage-specific genes play a major role in the commitment and development of specific hematopoietic lineages. Accordingly, some of hematopoietic disorders and leukemias are due to the dysregulation of the activities of specific transcription factors.1 The role of a number of transcription factors in hematopoietic development has been elucidated by knockout studies and has lead to the identification of GATA-1 and -2, Tal-1/SCL, Ikaros, c-myb, AML-1, PU.1 genes, among others, as key regulators of hematopoiesis. (reviewed in Tenen et al1, and Shivdasani and Orkin2). Of the numerous transcription factors involved in myelopoiesis, PU.1,3 and CAAT-enhancer-binding protein (C/EBP)-a and -e are critical for normal granulocytic differentiation;3–6 while Egr-1, a zincfinger protein, drives macrophage differentiation.7 Interferon regulatory factor 1 (IRF-1) was originally identified as a protein binding to DNA sequences termed IRF-E present on the promoters of the IFN-a and -b genes.8 These sequences, detected on most IFN-stimulated gene promoters, bind with Correspondence: Dr A Battistini, Department of Infectious, Parasitic and Immunomediated Diseases, Istituto Superiore di Sanita`, Viale Regina Elena, 299, 00161 Rome, Italy; Fax: þ 3906 49902082; E-mail: [email protected] Received 4 February 2004; accepted 2 July 2004; Published online 16 September 2004

different affinities at least nine cellular members belonging to the IRF family, which share a homologous DNA-binding domain located at the NH2 terminus. Beside their key role in gene regulation by IFNs and viral infections, IRFs have been recognized as regulators of genes expressed during inflammation, immune responses, hematopoiesis, cell proliferation and differentiation (for a review, see Nguyen et al9 and Taniguchi et al10). Specifically, clinical studies indicated that IRF-1 may function as tumor suppressor by preventing the development of some forms of leukemia. In particular, the IRF-1 gene maps to the chromosomal region 5q31.1, which is usually deleted in acute promyelocytic leukemia or leukemias and preleukemic myelodysplastic syndrome.11–14 Similarly, in chronic myeloid leukemia, a marked decrease of full-length IRF-1 mRNA, associated with high levels of aberrant spliced IRF-1 mRNA generated by exon skipping, has been recently reported.15 Intensive functional analysis of IRF-1 revealed a remarkable functional diversity of this transcription factor in the regulation of cellular responses, through the modulation of different sets of genes, depending on the cell type, state of the cell and/ or the nature of the stimuli.16 Specifically, studies in knockout mice implicated IRF-1 in the regulation of various immune processes such as T-cell selection and maturation, as well as leukemogenic development. Impairment in CD8 þ cell maturation, defective Th1 responses associated with defects in macrophage production of IL-12 and maturation of NK cells have all been observed in immune cells from IRF-1/ mice.17–21 Recent studies suggest that transcription factors belonging to the IRF family can play an important role in the myeloid differentiation. In particular, studies in knockout mice have shown that IRF-8 has a key role in the myeloid cell lineage selection and macrophage maturation,22 while in vitro studies have provided some evidence on the importance of IRF1 in granulocytic differentiation, its induction by G-CSF representing a limiting step in the early events of differentiation.23 In this study, we investigated the role of IRF-1 in normal granulocytic differentiation in vivo, in mice with a null mutation in IRF-1. Even though several defects in the lymphoid lineage of these mice had been described, a detailed analysis of medullar myelopoiesis was missing. Here, we show that in the IRF-1/ mice there is a reduction in the CFU-GM and CFU-M colonies and a decrease of the Gr-1bright/CD11b cells, associated with an increase of the Gr-1dim fraction. This phenotype correlates with a decreased expression of key master genes in granulocytic differentiation such as C/EBP-e and -a, whose expression is accordingly stimulated by IRF-1. These results represent the first in vivo evidence of the involvement of IRF-1 in the early phases of granulo-monocytic differentiation and further extend the role of this pleiotropic transcription factor in hematopoietic differentiation.

Interferon regulatory factor-1 and granulocytic differentiation U Testa et al

1865 Materials and methods

Mice IRF-1-deficient mice17 (IRF-1/) (Jackson Laboratory Bar Harbor, ME, USA) were kindly provided by Yutaka Tagaya (NCI, NIH Bethesda, USA). The / mice, which are on a hybrid C57BL/6X129/SV background, have been crossed with C57BL/6 and the / and þ / þ mice were obtained by back cross. Homozygous-deficient and wild-type (wt) ( þ / þ ) mice were bred and maintained under specific pathogen-free conditions. Genotyping was performed by polimerase chain reaction (PCR) using standard procedures and the following primers: IRF-1 gene: SN: AS:

ACA AAG CAG GAG AAA AAG AGC CAG TTC CTG GTG AGG GGT GGC AGC ATC

colonies by 10% WEHI-3-conditioned medium, 20 ng/ml SCF and 3 IU of recombinant human erythropoietin (kindly provided by Amgen, Thousand Oaks, CA, USA; specific activity, 1.1 106 U/mg). At days 8 and 10, CFU-GM and BFU-E colonies were scored and counted under an inverted microscope by two independent investigators. In some experiments, individual colonies have been pickedup and analyzed for cell morphology after cytocentrifugation and staining with May–Gru¨nwald–Giemsa.

Bone marrow and spleen cell morphology Bone marrow or spleen cells were counted in a hemocytometer. Approximately, 3 104 cells in 0.1 ml physiological solution containing 2 mg/ml bovine serum albumin were cytocentrifuged. Slides were stained with May–Gru¨nwald–Giemsa and examined under a miscroscope as previously described.

Neo gene: SN: AS:

TTC CAG ATT CCA TGG AAG CAC GC ATT CGC CAA TGA CAA GAC GCT GG

Mice were treated in accordance with Italian and European guidelines and were killed between 6 and 8 weeks of age.

Cells and culture CD11b þ cells were isolated from bone marrow of IRF-1/ and wt mice by direct magnetic cell sorting with MACS CD11b þ microbeads (Miltenyi Biotec Gmbh, Germany) according to the manufacturer’s instructions. 32Dcl3 cells were grown in Iscove’s modified Dulbecco’s medium supplemented with 15% heat-inactivated fetal calf serum and 2 ng/ml recombinant murine IL-3. Culture conditions were at 321C in 5% CO2 atmosphere with twice weekly passages at low cell density. Stable transfectants were prepared as described elsewhere.23

CFU assay Bone marrow cells were isolated by flushing tibia and femur with phosphate-buffer saline from 10 wt or IRF-1/ mice. Spleens were removed aseptically and washed with a syringe in Iscove’s modified Dulbecco’s medium. The cell suspension was then passed through a single layer of 100-gauge nylon mesh, centrifuged and resuspended in Iscove’s modified Dulbecco’s medium at an appropriate cell dilution for colony assay. Bone marrow cells derived from tibias were passed into a 2-ml syringe through a 11-gauge needle; cells were then counted and resuspended at an appropriate dilution for colony assay. The methylcellulose culture method was used for the assay of BFU-E, CFU-G, CFU-M and CFU-GM and progenitors. Bone marrow and spleen cells were cultured at a concentration of 5 104 and 1 105 cells/ml/dish, respectively (three plates/point) in 0.9% methylcellulose (cell culture grade; Sigma, St Louis, MO, USA), 40% fetal calf serum (Flow Laboratories, Glasgow, Scotland) and a-thioglycerol (104 mol/liter; Sigma) at 371C in a 5% CO2 humidified atmosphere. The growth of CFU-G, CFU-M and CFU-GM colonies was sustained by the addition of recombinant murine growth factors, specifically G-CSF (Peprotech, London, UK), M-CSF (Peprotech, UK), GM-CSF (Peprotech, UK), added alone or in combination in the absence or in the presence of murine kit ligand (Peprotech, UK); the growth of erythroid

White blood cells (WBC) and granulocytic cell count Anesthetized animals were bled by the tail. In all, 5 ml of blood was diluted in 50 ml of Turk’s solution, and WBC were counted in a hemocytometer. Blood was smeared on a glass slide and stained with May–Gru¨nwald–Giemsa for granulocyte identification and counting.

Flow cytometry analysis For flow cytometry analysis, cells were resuspended in PBS containing bovine serum albumin (2 mg/ml) and incubated 30 min at 41C in the presence of an appropriate dilution of one of these antibodies: anti-CD11b/Mac-1, anti-Gr-1/Ly-6G, anti Sca-1/Ly6A/E, anti-c-kit, anti-flt3 and anti-TER-119. All these antibodies were directly conjugated with either fluoresceine (FITC) or phycoerythin (PE) and were purchased from Becton Dickinson/Pharmingen Co (San Jose, CA, USA). After two washings in cold PBS, the cells were resuspended in PBS and analyzed by a FACS Scan (Becton-Dickinson) flowcytometer. Negative controls were stained with appropriate isotypematched negative controls. In some experiments, bone marrow cells of wt and IRF-1/ mice were labelled with FITC-labelled anti-CD11b and PE-labelled Gr-1; after labelling, cells were sorted into two populations according to the intensity of Gr-1 labelling (CD11b þ /Gr-1dim and CD11b þ /Gr-1bright) using a FACS vantage (Becton-Dickinson) flow cytometer. In some experiments, bone marrow cells have been labelled with a set of anti-B-lymphocyte antibodies (anti-B220, -CD19, -CD21, CD22, -CD40 and -IgM) and analyzed by flow cytometry as above.

Western blot assay Nuclear cell extracts were prepared as described.24 Total cell extracts from CD11b þ cells were prepared by resuspending cell pellets in lysis buffer containing 60 mM Tris-HCl pH 6.8, 2% (v/v) SDS, 1% (v/v) glycerol. Cell extracts were then boiled for 10 min to dissolve DNA and membranes. Aliquots of 20–50 mg of nuclear or total cell extracts were separated on 10% SDS PAGE. Blots were incubated with polyclonal anti-c/EBP-a (Sc61 Santa Cruz Biotechnology), antic/EBP-e (Sc158, Santa Cruz Biotechnology) and anti-PU.1 antibodies (Sc352, Santa Cruz Biotechnology) and then with Leukemia


1866 horseradish peroxidase-coupled secondary antibody (Amersham, Buckingamshire, UK), using the enhanced chemiluminescence system.

Table 1 Evaluation of the number of cells present in the peripheral blood of wt and IRF-1/ mice Peripheral blood counts

Results

Blood cell counts in adult mice The IRF-1/ mice have been previously described. 17 The IRF-1 gene contains a deletion of aminoacids 63–223 corresponding to the exons that encode the DNA-binding domain. The mutated IRF-1 gene was confirmed by PCR of tail DNA (Figure 1a), as described in Material and methods. The number of blood cells present in the peripheral blood of wt and IRF-1/ mice was evaluated (Table 1). White blood cell counts were significantly decreased in the group of IRF-1/ mice as compared to wt controls, consistently with literature data, while both groups of animals exhibited comparable levels of red blood cells. The analysis of subpopulations of WBCs showed that IRF-1/ mice displayed a marked decrease in the number of lymphocytes and a moderate, but significant, decrease of granulocytes as compared to the levels observed in wt mice.

IRF-1/ mice show an impaired granulocytic differentiation The analysis of the cellularity of the bone marrow and the spleen (Figure 1b) showed that: (i) IRF-1/ mice exhibited a moderate,

White blood cells/ml ( 103) Erythrocytes/ml ( 106) Lymphocytes/ml ( 103) Monocytes/ml ( 103) Granulocytes/ml ( 103)

Wt mice (N ¼ 6)

IRF-1/ mice (N ¼ 6)

P

5.3670.29 3.470.5 4.270.19 0.4570.05 0.7170.04

3.2570.43 3.2770.17 2.0970.08 0.5170.017 0.5470.05

o0.05 NS o0.01 NS 0.05

Mean values 7s.e. observed in six different mice. NS ¼ not significant.

but significant (Po0.05), decrease in the number of total hemopoietic cells present at the level of bone marrow, as compared to the levels observed in wt mice; (ii) at the level of the spleen, the cellularity was comparable in both wt and IRF1/ mice. Given the decrease in the bone marrow cellularity observed in IRF-1/ mice, we asked whether this phenomenon was due to a generalized or selective decrease in the different hemopoietic lineages. Thus, bone marrow cells were labelled with specific antibodies and the number of positive cells was evaluated. As shown in Figure 1c, the total number of Gr-1 (Ly6G)- or CD11b (Mac-1)- positive cells per tibia was decreased from a mean of 4.0 10670.3 and 4.3 10670.45, respectively, for control mice to a mean of 2.21 10670.19 and 2.43 10670.27 for IRF-1/ mice. The effect of IRF-1 deficiency was specific for the granulocytic lineage, in that the total number of erythroid cells (identified as Ter-119-positive cells) present in the bone marrow of IRF-1/ mice was similar to that observed in wt mice (Figure 1c). These observations indicate that IRF-1 deficiency selectively affects the granulocyte lineage in the bone marrow.

Lack of IRF-1 results in a reduction in granulo-monocytic progenitors

Figure 1 IRF-1/ mice show an impaired granulocytic differentiation. (a) Genotyping of the IRF-1/ mice. (b) The number of total cells from bone marrow and spleens was evaluated by counting cells in a hemocytometer. Cells were recovered from three wt and three IRF-1/ mice 6–8 weeks old, and the mean values of three separate experiments are indicated. Po0.05. (c) Bone marrow cells from tibia of three wt or null mice were labelled with specific anti GR-1 (Ly 6G), CD11b (Mac-1) and Ter 119 antibodies and analyzed in a FACS scan flow cytometer. The total number of positive cells are expressed as mean values of three independent experiments using five animals for each group. Leukemia

The decreased numbers of granulocytes in the bone marrow of IRF-1-deficient animals could be due to a decrease in the size of the pool of granulocyte precursors, impaired granulocytic maturation or death of immature and/or mature granulocytic cells. To distinguish between these possibilities, granulopoiesis was explored at the level of progenitor compartment by clonogenic assay in semisolid medium, as well as at the level of different maturation compartments defined according to double labelling procedures. Bone marrow cells were plated in semisolid medium and granulocytic (CFU-G), monocytic (CFU-M) and granulo-monocytic (CFU-GM) progenitors were evaluated using a standard colony-forming assay. Bone marrow cells have been grown in the presence of either G-CSF added alone or togheter with KL or M-CSF added alone or together KL and GM-CSF þ G-CSF þ MCSF with and without KL (Figure 2). This study provided evidence that the number of CFU-G and CFU-GM was reduced in all culture conditions in IRF-1/ mice compared to wt mice (Figure 2a). Interestingly, IRF-1/ mice displayed a very low response to M-CSF. Moreover, the size of CFU-G and CFU-GM colonies in cultures of bone marrow IRF-1/ cells was significantly lower than that observed in the corresponding colonies of bone marrow wt cells (data not shown). Finally, the analysis of the cellular composition of individual CFU-G colonies at day 8 of culture provided evidence that these


1867

Figure 2 Colony forming ability of IRF-1 þ / þ (wt) and IRF-1/ bone marrow cells. (a) Bone marrow cells were cultured in the methylcellulose medium in the presence of M-CSF (10 ng/ml) or G-CSF (10 ng/ml) or M-CSF þ G-CSF þ GM-CSF (10 ng/ml) at 5 104 cells/ml for 7 days. In some plates to these growth factor combinations, KL (20 ng/ml) was added. Colonies were classified into granulocyte (CFU-G), macrophage (CFU-M) and granulocyte-macrophage (CFU-GM) colonies according to standard morphologic criteria. Values are the mean 7 s.d. of quadruplicate cultures. The difference in the number of colonies between wt and IRF-1/ mice were statistically significant in all cases, except for CFU-G when grown in the presence of KL þ M-CSF and CFU-M when grown in the presence of G-CSF and G-CSF þ KL. (b) Cell morphology of individual CFU-G colonies grown in the presence of G-CSF as reported in (a). Cytospin preparations of individual colonies picked up under an inverted microscope have been stained with May–Gru¨mwald–Giemsa. Original Magnification 200. (c) Cell morphology of bone marrow cells obtained from wt and IRF-1/ mice. Cytospin preparation of bone marrow samples obtained from wt and IRF-1/ mice were stained with May–Grunwald–Giemsa. Representative cells from three fields are shown. The photographs were taken at an original magnification of 400 using a Nikon microscope equipped with a camera and a device for digital imaging elaboration.

colonies when derived from wt mice bone marrow are composed by mature granulocytic elements, while when they are derived from IRF-1/ mice bone marrow are composed by a significant proportion of immature granulocytic precursors (Figure 2b).

In contrast to these observations, similar number of erythroid progenitors (BFU-E) were observed in IRF-1/ mice and wt mice (data not shown). At the spleen level, a comparable number of both CFU-GM and BFU-E was observed in wt and IRF-1/ mice (data not shown). Leukemia


1868

Impaired granulocytic maturation in IRF-1/ mice In line with the observations made on single colonies, the inspection of cytospin preparations of bone marrow stained with May-Grumwald-Giemsa showed that, in the bone marrow of IRF-1/ mice, a marked increase of immature granulocytic elements was observed, compared to wt bone marrow (Figure 2c). In particular, the bone marrow from wt mice contained cells in various stages of differentiation, from promyelocytes to neutrophils, with a predominant frequency of mature elements, whereas bone marrow from IRF-1/ mice contained mostly promyelocytes and myelocytes. To evaluate at which step of granulocytic maturation a defect could occur, we performed double labelling experiments for different granulocytic markers. Cells from the bone marrow

were labelled with a panel of antibodies for either CD11b or Gr1, which allow the identification of all granulocytic elements present in the bone marrow, and c-kit, Flt3 or Sca-1, which allow the identification of the more immature elements of the granulocytic lineage. The immature granulocytic cells were also identified by double labelling experiments with CD11b and Gr-1 antibodies, in that the cell population CD11b þ /Gr-1dim is composed by immature granulocytic elements, while CD11b þ /Gr-1bright cells are composed by more mature granulocytic elements, as previously shown.25 Representative double labelling experiments are illustrated in Figure 3, where the results obtained by a total gating (Figure 3a) as well as by a gating limited to the cells with a scatter corresponding to granulocytic cells are shown (Figure 3b). This analysis provided evidence that immature granulocytic cells identified either as

Figure 3 Impaired granulocytic maturation and increase of immature elements in the bone marrow of IRF-1/ mice. Bone marrow cells derived from wt and IRF-1/ mice were labelled with different couples of mAbs (CD11b and Gr-1; c-kit and Gr-1; flt3 and CD11b, Scan-1 and CD11b) and analyzed for fluorescence by flow cytometry in a FACscan. One representative analysis is shown. (a) Gating on total bone marrow cells. (b) Gating on cells with a myeloid scatter. The number of total positive cells for each marker (c-kit/Gr-1; Sca-1/CD11b; flt3/CD11b) was calculated as mean value 7 s.e. The difference in the number of these populations between wt and IRF-1/ mice was statistically significant: for c-kit þ /Gr-1 þ , P ¼ p0.05; for Sca-1 þ /CD11b þ , P ¼ p0.02; for Flt3 þ /CD11b þ , Po0.05. Leukemia


1869 Table 2 cells

Immunophenotypic analysis of bone marrow B-lymphoid

Mice B220+/IgM B220+/IgM+ B220+/BP-1 B220+/BP-1+ CD19+/IgM CD19+/IgM+ CD19+/BP-1 CD19+/BP-1+ CD19+/CD21 CD19+/CD21+ CD19+/CD22 CD19+/CD22+ CD19+/CD40 CD19+/CD40+

Wt

IRF-1/

1874 1973 2773.8 1173 1572.4 1171.2 3173 9.770.9 19.671 1471.5 1571.7 1372.6 19.371.1 1271.1

2571.8 1970.5 3472.1 1171.6 2072 1470.7 3972.3 12.370.9 20.671 1771.6 23.372.4 14.372.1 20.370.6 1671.2

Bone marrow cells were harvested from four wt and four IRF-1/ mice, labelled with couples of mAbs anti-B lymphocytes, and the proportion of positive cells (mean values 7 s.e.) has been determined by flow cytometry with a cell gating in the lymphocyte area of the cell scatter.

CD11b/Gr-1dim or c-kit þ /Gr-1 þ or Sca-1 þ /CD11b þ or Flt3 þ / CD11b þ cells are clearly increased in IRF-1/ mice compared to the levels observed in wt mice. The quantitative analysis of the proportion of c-kit þ /Gr-1 þ , Sca-1 þ /CD11b þ and Flt3 þ / CD11b þ cells indicated that their number is significantly increased in IRF-1-deficient mice compared to wt mice (Po0.02, o0.02 and o0.05, respectively) (Figure 3b).

B-lymphocyte differentiation is normal in IRF-1/ mice The effect of IRF-1 deficiency on B lymphopoiesis was also investigated. The total number of B lymphocytes present in the bone marrow of IRF-1/ mice was comparable to that observed in wt mice (data not shown). The proportion of immature B-lymphoid precursors (identified as B220 þ /IgM þ , B220 þ /BP1 þ , 19 þ /IgM þ , 19/BP-1 þ , 19 þ /21, 19 þ /22, 19 þ /40) and B-lymphoid mature elements (identified as B220 þ /IgM, B220 þ /BP-1, 19 þ /IgM, 19 þ /21 þ , 19 þ /22 þ and 19 þ /40 þ ) was comparable in wt and IRF-1 mice (Table 2).

C/EBP-a and -e expression is reduced in CD11b þ cells from IRF-1/ mice bone marrow C/EBP-a and -e are transcription factors playing a critical role in granulocytic lineage development. Expression of C/EBP-a can induce granulocytic differentiation of multipotential myeloid cells, and mechanisms of C/EBP-a inactivation in acute myelogenous leukemia have been described.26,27 Similarly, C/EBP-e-null mice fail to generate functional neutrophils; notably, it has been suggested that C/EBP-e acts temporally downstream of C/EBP-a, blocking late steps in terminal differentiation of mature segmented granulocytes.28,29 Considering the impaired granulopoiesis and, in particular, the increase of immature granulocytic elements observed in IRF-1/ mice, we asked whether bone marrow cells from these animals could show a decreased C/EBP-a and -e expression. The expression of C/EBP-a and C/EBP-e was, therefore, tested in CD11b þ cells purified from bone marrow of IRF-1/ mice. As shown in Figure 4a, a significant decrease in both C/EBP-a and -e

Figure 4 C/EBP-a and -e expression is reduced in CD11b þ cells isolated from IRF-1/ mice. (a) CD11b þ cells were purified by magnetic isolation from the bone marrow of six wt or IRF-1/ mice, as indicated in Materials and methods. Total cell extracts were subjected to Western blot analysis with C/EBP-a, c/EBP-e and PU.1specific antibodies. The represented data are from one out of three different experiments. (b) Nuclear cell extracts from 32DCl3 cells transfected with an empty vector (RcCMV) or an IRF-1-expressing vector were subjected to Western blot analysis with anti-C/EBP-a, -e and anti IRF-1-specific antibodies, as described in Materials and methods. Cell extracts from HL-60 cells were used as a positive control for the C/EBP-a and -e expression. The same membranes were stripped and reprobed with anti-TFIIH or b-actin-specific antibodies (bottom panels).

expression was observed in CD11b þ cells of IRF-1/ mice as compared to the levels observed in cells isolated from wt mice. Moreover, PU.1, a downstream target of c/EBP-a, whose overexpression has been shown to potentiate terminal myeloid differentiation,23,30 was substantially inhibited in IRF-1/ mice. Finally, to further evaluate the link between IRF-1 andC/EBP-a and -e, we evaluated the expression of these transcription factors in 32Dcl3 cells overexpressing IRF-1 (Figure 4b). Of interest, the Western blot analysis of C/EBP-a and -e expression in these cells showed a substantial increase in both transcription factors, which was even greater than that observed when parental cells were allowed to differentiate in the presence of G-CSF (data not shown). These results suggest that IRF-1 is able to modulate the expression of transcription factors, such as C/EBP-a and -e, playing a key role in granulocytic differentiation.

Discussion In this study, we have demonstrated that the loss of IRF-1 results in the in vivo impairment of granulocytic differentiation. In fact, we have found that the bone marrow of IRF-1/ mice shows a decrease of about 50% in CFU-GM colonies number, which is associated with a marked increase in immature granulocytic cells identified as CD11b þ /Gr-1dim, c-kit þ /Gr-1 þ , Sca-1 þ / CD11b þ or Flt3 þ /CD11b þ cells. Leukemia


1870 A role of IRF-1 in granulocytic differentiation was suggested by some in vitro studies: IRF-1 is expressed in immature myeloid bone marrow cells31 and its expression significantly increases during granulocytic maturation of human progenitors;23 moreover, IRF-1 gene expression is rapidly induced in M1 myeloblastic leukaemia cells induced to terminal differentiation by IL-6 or leukaemia inhibitory factor (LIF),32 and finally enforced IRF-1 expression in the 32Dcl3 myeloid progenitor cell line stimulates spontaneous granulocytic differentiation while potentiating G-CSF-induced maturation.23 While these findings support the involvement of IRF-1 in the control of granulopoiesis, the demonstration of the in vivo relevance of this transcription factor was still missing. Thus the present study, based on the analysis of granulopoiesis in IRF-1/ mice, represents the first in vivo evidence on the role of IRF-1 in controlling normal granulopoiesis. Two different mechanisms seem responsible for the impaired granulopoiesis observed in the knockout mice. The first defect occurs at the level of the progenitor compartment and the other one at the level of granulocytic precursors. Bone marrows of IRF-1/ mice show, in fact, a decrease in CFU-GM colonies of about 50% (Figure 2). This decrease in granulo-monocytic progenitors is associated with a marked increase in immature granulocytic precursor cells, identified as CD11b þ /Gr-1dim, cKit þ /Gr-1 þ , Sca-1 þ /CD11b þ or Flt3 þ /CD11b þ cells (Figure 3). Only a moderate decline of granulocytes is instead observed in the peripheral blood, indicating that compensatory mechanisms can occur during late stages of granulopoiesis. Of interest, the phenotype observed at level of bone marrow in IRF-1/ mice is reminiscent of that observed in the C/EBP-e/ mice.28,29 In particular, these mice showed a pronounced increase of immature myeloid elements in bone marrow, associated with a significant decrease in the number of mature granulocytes and the presence of immature myeloid elements in peripheral blood. The similarities between the phenotypes of these two knockout mice suggest that the effect of IRF-1 on granulopoiesis could be mediated through C/EBP-e. In line with this hypothesis, we observed that (i) there was a significant decrease (more than 70%) of C/EBP-e levels in cell extracts derived from bone marrow of IRF-1/ mice, as compared to normal mice; (ii) C/EBP-e levels were upmodulated in 32Dcl3 cells overexpressing IRF-1, as compared to the levels observed in wt 32Dcl3 cells. Of note, the analysis of IRF-1/ mice also showed an effect at the level of the CFU-G compartment, where a moderate, but significant, decrease in the number of these progenitors, compared to that in wt mice, was observed. This effect on the granulocytic compartment could be due to a reduction in C/ EBP-a expression that is, in fact, expected to give rise to a decrease in the number of granulocytic progenitors. According to this hypothesis, 32Dcl3 cells overexpressing IRF-1 display a pronounced increase in C/EBP-a levels, while bone marrow cells derived from IRF-1/ animals exhibit a decrease in C/EBPa expression as compared to wt mice. Interestingly, as reported in,23 IRF-1 overexpressing cells are able to partially undergo granulocytic differentiation even in the absence of growth factors and this effect is not related to an inhibition of cell growth. The inhibitory effect on the granulocytic differentiation observed after C/EBP-a gene targeting is markedly more pronounced than that observed in IRF-1/ mice,33 in spite of the marked decrease in C/EBP-a observed in these mice. This discrepancy could be related to the fact that the decrease of C/ EBP-a (and PU.I) observed in IRF-1/ CD11b þ cells, in comparison to the corresponding normal cells, is overestimated, due to the accumulation of immature granulocytic precursors in Leukemia

IRF-1/ compared to wt mice. Additional experiments are required to assess whether the effect of IRF-1 deficiency on C/ EBP-a and PU.1 is due to a primary effect of IRF-1 on the expression of these genes or to the general effect on granulopoiesis. It is of interest to note that we also observed a markedly reduced response of IRF-1/ bone marrow cells to M-CSF. This finding is similar to the defective M-CSF response observed in IRF-8/ mice. In IRF-8/ mice, however, a markedly reduced CFU-M colony formation in response to M-CSF is associated with a markedly increased CFU-G colony formation in response to M-CSF and G-CSF.34,35 Therefore, the result obtained in IRF1/ mice is in some way unexpected considering that the two lineages, macrophage and granulocyte are coupled and the development of one is associated with the repression of the other. The reason of the reduced response of IRF-1/ bone marrow cells to M-CSF is at the moment unclear and could be related, as shown for IRF-8/ mice, to a reduced M-CSFR expression34 and signalling.35 We could however speculate that IRF-1 acts at a earlier step in lineage commitment with respect to IRF-8, affecting a common GM progenitor. This hypothesis is also supported by the highly decreased PU.1 expression observed in IRF-1/ mice. Previous analyses of IRF-1 knockout mice36 did not revealed severe defect in the myeloid lineages differentiation. However, in previous studies, maximal emphasis was generally given to immunological abnormalities of these mice and no detailed analysis of bone marrow hematopoiesis was performed. Moreover, compensatory mechanisms occurring during late stages of granulopoiesis may result in a mild phenotype at the level of peripheral blood. This has been reported for knockout mice for other transcription factors. Stat5a/ 5b/ mice, for example, despite their almost normal hematocrit, exhibit marked abnormalities of erythropoiesis at the level of bone marrow, with expansion and premature death of immature erythroblasts.37 In conclusion, our results demonstrate a previously unrecognized role of IRF-1 in granulopoiesis and indicate that IRF-1 acts at several steps in this process, via a modulation of early specific transcription factors including C/EBP-a, -e and PU.1, and through induction of lineage-specific markers as previously reported.23 Moreover, the observation that changes in marrow granulopoiesis are associated with only a moderate decline of granulocytes in the blood stream is in line with previous observations, suggesting that multiple independent molecular pathways may stimulate granulocyte differentiation and alternative pathways can compensate for the absence of specific transcription factors, stressing again the remarkable degree of redundancy and robustness involved in the regulation of granulopoiesis.

Acknowledgements We thank Sabrina Tocchio for secretarial and editorial assistance and Roberto Gilardi for preparing graphs. This work was partially supported by Institutional grants from the Istituto Superiore di Sanita`, 1% Projects, and Special Project on AIDS to AB.

References 1 Tenen DG, Hromas R, Licht J, Zhang DE. Transcription factors, normal myeloid development, and leukemia. Blood 1997; 90: 489–519.


1871 2 Shivdasani RA, Orkin SH. The transcriptional control of hematopoiesis. Blood 1996; 87: 4025–4039. 3 Lloberas J, Soler C, Celada A. The key role of PU.1/SPI-1 in B cells, myeloid cells and macrophages. Immunol Today 1999; 20: 184–189. 4 Wang X, Scott E, Sawyers CL, Friedman AD. C/EBPa bypasses granulocyte colony-stimulating factor signals to rapidly induce PU.1 gene expression, stimulate granulocytic differentiation, and limit proliferation in 32D cl3 myeloblasts. Blood 1999; 94: 560–571. 5 Radomska HS, Huettner CS, Zhang P, Cheng T, Scadden DT, Tenen DG. CCAAT/enhancer binding protein a is a regulatory switch sufficient for induction of granulocytic development from bipotential myeloid progenitors. Mol Cell Biol 1998; 18: 4301–4314. 6 Lekstrom-Himes J, Xanthopoulos KG. Biological role of the CCAAT/enhancer-binding protein family of transcription factors. J Biol Chem 1998; 273: 28545–28548. 7 Nguyen HQ, Hoffman-Liebermann B, Liebermann DA. The zinc finger transcription factor Egr-1 is essential for and restricts differentiation along the macrophage lineage. Cell 1993; 72: 197–209. 8 Fujita T, Sakakibara J, Sudo Y, Miyamoto M, Kimura Y, Taniguchi T. Evidence for a nuclear factor(s), IRF-1, mediating induction and silencing properties to human IFN-b gene regulatory elements. EMBO J 1988; 7: 3397–3405. 9 Nguyen H, Hiscott J, Pitha PM. The growing family of interferon regulatory factors. Cytokine Growth Factor Rev 1997; 8: 293–312. 10 Taniguchi T, Ogasawara K, Takaoka A, Tanaka N. IRF family of transcription factors as regulators of host defense. Annu Rev Immunol 2001; 19: 623–655. 11 Boultwood J, Fidler C, Lewis S, MacCarthy A, Sheridan H, Kelly S et al. Allelic loss of IRF1 in myelodysplasia and acute myeloid leukemia: retention of IRF1 on the 5q-chromosome in some patients with the 5q-syndrome. Blood 1993; 82: 2611–2616. 12 Green WB, Slovak ML, Chen IM, Pallavicini M, Hecht JL, Willman CL. Lack of IRF-1 expression in acute promyelocytic leukemia and in a subset of acute myeloid leukemias with del(5)(q31). Leukemia 1999; 13: 1960–1971. 13 Willman CL, Sever CE, Pallavicini MG, Harada H, Tanaka N, Slovak ML et al. Deletion of IRF-1, mapping to chromosome 5q31.1, in human leukemia and preleukemic myelodysplasia. Science 1993; 259: 968–971. 14 Harada H, Kondo T, Ogawa S, Tamura T, Kitagawa M, Tanaka N et al. Accelerated exon skipping of IRF-1 mRNA in human myelodysplasia/leukemia; a possible mechanism of tumor suppressor inactivation. Oncogene 1994; 9: 3313–3320. 15 Tzoanopoulos D, Speletas M, Arvanitidis K, Veiopoulou C, Kyriaki S, Thyphronitis G et al. Low expression of interferon regulatory factor-1 and identification of novel exons skipping in patients with chronic myeloid leukaemia. Br J Haematol 2002; 119: 46–53. 16 Taniguchi T, Lamphier MS, Tanaka N. IRF-1: the transcription factor linking the interferon response and oncogenesis. Biochem Biophys Acta 1997; 1333: M9–M17. 17 Matsuyama T, Kimura T, Kitagawa M, Pfeffer K, Kawakami T, Watanabe N et al. Targeted disruption of IRF-1 or IRF-2 results in abnormal type I IFN gene induction and aberrant lymphocyte development. Cell 1993; 75: 83–97. 18 Reis LF, Harada H, Wolchok JD, Taniguchi T, Vilcek J. Critical role of a common transcription factor, IRF-1, in the regulation of IFN-b and IFN-inducible genes. EMBO J 1992; 11: 185–193. 19 Ogasawara K, Hida S, Azimi N, Tagaya Y, Sato T, Yokochi-Fukuda T et al. Requirement for IRF-1 in the microenvironment supporting development of natural killer cells. Nature 1998; 391: 700–703. 20 Penninger JM, Sirard C, Mittrucker HW, Chidgey A, Kozieradzki I, Nghiem M et al. The interferon regulatory transcription factor IRF1 controls positive and negative selection of CD8+ thymocytes. Immunity 1997; 7: 243–254.

21 Lohoff M, Ferrick D, Mittru¨cker HW, Duncan GS, Bischof S, Ro¨llinghoff M et al. Interferon regulatory factor-1 is required for a T helper 1 immune response in vivo. Immunity 1997; 5: 531–539. 22 Tamura T, Nagamura-Inoue T, Shmeltzer Z, Kuwata T, Ozato K. ICSBP directs bipotential myeloid progenitor cells to differentiate into mature macrophages. Immunity 2000; 13: 155–165. 23 Coccia EM, Stellacci E, Valtieri M, Masella B, Feccia T, Marziali G et al. Ectopic expression of interferon regulatory factor-1 potentiates granulocytic differentiation. Biochemical J 2001; 360: 285–294. 24 Saura M, Zaragoza C, Bao C, McMillan A, Lowenstein CJ. Interaction of interferon regulatory factor-1 and nuclear factor kB during activation of inducible nitric oxide synthase transcription. J Mol Biol 1999; 289: 459–471. 25 Walkley CR, Yuan YD, Chandraratna RA, McArthus GA. Retinoic acid receptor antagonism in vivo expands the numbers of precursor cells during granulopoiesis. Leukemia 2002; 16: 1763–1772. 26 Westendorf JJ, Yamamoto CM, Lenny N, Downing JR, Selsted ME, Hiebert SW. The t(8;21) fusion product, AML-1-ETO, associated with C/EBP-alpha, inhibits C/EBP-alpha-dependent transcription, and blocks granulocytic differentiation. Mol Cell Biol 1998; 18: 322–333. 27 Pabst T, Mueller BU, Harakawa N, Schoch C, Haferlach T, Behre G et al. AML1-ETO downregulates the granulocytic differentiation factor C/EBPalpha in t(8;21) myeloid leukemia. Nat Med 2001; 7: 444–451. 28 Yamanaka R, Kim GD, Radomska HS, Lekstrom-Himes J, Smith LT, Antonson P et al. CCAAT/enhancer binding protein epsilon is preferentially up-regulated during granulocytic differentiation and its functional versatility is determined by alternative use of promoters and differential splicing. Proc Natl Acad Sci USA 1997; 94: 6462–6467. 29 Lekstrom-Himes J. The role of C/EBPe in the terminal stages of granulocyte differentiation. Stem Cells 2001; 19: 125–133. 30 Bellon T, Perrotti D, Calabretta B. Granulocytic differentiation of normal hematopoietic precursor cells induced by transcription factor PU.1 correlates with negative regulation of the c-myb promoter. Blood 1997; 90: 1828–1839. 31 Liebermann DA, Hoffman B. Myeloid differentiation (MyD) primary response genes in hematopoiesis. Oncogene 2002; 21: 3391–3402. 32 Abdollahi A, Lord KA, Hoffman-Liebermann B, Liebermann DA. Interferon regulatory factor 1 is a myeloid differentiation primary response gene induced by interleukin 6 and leukemia inhibitory factor: role in growth inhibition. Cell Growth Differ 1991; 2: 401–407. 33 Zhang DE, Zhang P, Wang ND, Hetherington CJ, Darlington GJ, Tenen DG. Absence of granulocyte colony-stimulating factor signaling and neutrophil development in CCAAT enhancer binding protein alpha-deficient mice. Proc Natl Acad Sci USA 1997; 94: 569–574. 34 Scheller M, Foerster J, Heyworth CM, Waring JF, Lohler J, Gilmore GL et al. Altered development and cytokine responses of myeloid progenitors in the absence of transcription factor, interferon consensus sequence binding protein. Blood 1999; 94: 3764–3771. 35 Tsujimura H, Nagamura-Inoue T, Tamura T, Ozato K. IFN consensus sequence binding protein/IFN regulatory factor-8 guides bone marrow progenitor cells toward the macrophage lineage. J Immunol 2002; 169: 1261–1269. 36 Sato M, Taniguchi T, Tanaka N. The interferon system and interferon regulatory factor transcription factors – studies from gene knockout mice. Cytokine Growth Factor Rev 2001; 12: 133–142. 37 Socolovsky M, Nam H, Fleming MD, Haase VH, Brugnara C, Lodish HF. Ineffective erythropoiesis in Stat5a(/)5b(/) mice due to decreased survival of early erythroblasts. Blood 2001; 98: 3261–3273.

Leukemia