Thrombopoietin Inhibits Murine Mast Cell Differentiation

Authors


Abstract

We have recently shown that Mpl, the thrombopoietin receptor, is expressed on murine mast cells and on their precursors and that targeted deletion of the Mpl gene increases mast cell differentiation in mice. Here we report that treatment of mice with thrombopoietin or addition of this growth factor to bone marrow-derived mast cell cultures severely hampers the generation of mature cells from their precursors by inducing apoptosis. Analysis of the expression profiling of mast cells obtained in the presence of thrombopoietin suggests that thrombopoietin induces apoptosis of mast cells by reducing expression of the transcription factor Mitf and its target antiapoptotic gene Bcl2.

Disclosure of potential conflicts of interest is found at the end of this article.

Introduction

Mast cells [1, [2]–3] derive from c-KitlowCD34lowSca-1pos stem/progenitor cells present in the marrow that give rise to c-Kithigh/T1/ST2pos mast cell-restricted progenitor cells (MCP) [4] and then to c-Kithigh/CD34high mast cell precursors that circulate in the blood to colonize extramedullary sites [5]. Mast cell precursors are characterized by the ability to proliferate extensively, by small cytoplasmic granules, and by little or no surface expression of the receptor that attaches to the Fc portion of IgE (FcεRIneg) with high affinity. These cells differentiate, in turn, into tissue-restricted mast cells: dermal, mucosal, and serosal in skin, gut, and peritoneal cavity, respectively [1, [2]–3].

Extrinsic and intrinsic control pathways regulate mast cell differentiation. The extrinsic pathway includes the growth factors stem cell factor (SCF) and interleukin (IL)-3 [6, [7]–8]. SCF is produced by marrow and skin fibroblasts alike [9], and mice carrying alterations in the gene encoding SCF [10] or its receptor, c-Kit [11], are mast cell-deficient. On the other hand, IL-3 is not produced in normal mice. Mast cells, however, are induced to produce IL-3 in an autocrine/paracrine fashion by IgE/FcεRI interactions [12]. The observation that IL-3-deficient mice have normal numbers of mast cells but produce an inadequate response when challenged with parasites [13] has suggested that IL-3 may be used to regulate mast cell numbers topically, in response to B cell recruitment to the site of infection [14]. Both SCF and IL-3 must be present for mast cell differentiation to occur in bone marrow-derived mast cell (BMMC) cultures [6, 8]. In 14 days, mast cell precursors with a mast cell phenotype (c-KithighCD34lowFcεRIneg with small Alcian Blueneg granules) [15] and function (reconstituting dermal and mucosal mast cells when transplanted into mast cell-deficient animals) develop within the cultures [16]. By day 21, they mature into berberine sulfateposFcεRIpos mast cells, which are biochemically similar to dermal mast cells [6, 7, 17]. In BMMC cultures, SCF permits proliferation [18, 19], whereas either SCF [20] or IL-3 [21] contributes to preventing apoptosis. At the intrinsic level, the transcription factors Mitf [22], Gata2 [23, 24], and Gata1 regulate mast cell differentiation [25]. Mitf encodes a helix-loop-helix protein that has either positive (MMCP-6) [26] or negative (MMCP-7) [27] effects on the expression of mast cell specific proteases (MMCP)s. Mitf also controls the expression of the antiapoptotic Bcl2 gene in these cells [28]. Gata2 and Gata1 are members of the GATA transcription factor family [29]. In mast cell differentiation, as in other hemopoietic lineages, Gata2 expression precedes that of Gata1 [24] and increases cell proliferation rates [23, 30]. On the other hand, given the presence of functional Gata1-binding sites in the regulatory region of mast cell carboxypepidase A (MC-CPA) [31] and that of the α [32] and β [33] chains of FcεRI, Gata1 may play a role in cell maturation.

We have recently described that treatment with thrombopoietin (TPO) affects the expression of Gata1 in mice carrying the hypomorphic Gata1low mutation and in their wild-type littermates [34, 35]. The effects exerted by TPO in the two animals are paradoxical (it increases and decreases the level of Gata1 expression in wild-type and mutant littermates, respectively) and involve cells of the erythroid, megakaryocytic, and mast cell lineages [34, 35]. The similarity of the effects of TPO on the expression of Gata1 in the cells of these three lineages (and the notion that Mpl, the TPO receptor, is expressed not only by erythroid cells and megakaryocytes [36] but by mast cells as well [34]) suggests to us that TPO/Mpl interactions might regulate mastocytopoiesis in mice. To confirm this hypothesis, we analyzed here the effects of in vivo and in vitro TPO treatments on murine mastocytopoiesis. TPO treatment of wild-type mice, or addition of TPO to BMMC cultures, favors proliferation of mast cell precursors but reduces the formation of mature cells by inducing their apoptosis. Molecular profiling analysis of cells exposed to TPO suggests that TPO affects mast cell maturation through a molecular mechanism that involves induction of apoptosis by downmodulation of cKit expression on the cell surface and suppression of Mitf and Bcl2 expression, possibly through the SCF/Mitf/PIAS3/STAT3 axis [37].

Materials and Methods

Mice

Mplnull mice [38] were provided by Dr. W. Alexander (Walter and Eliza Hall Institute for Medical Research, Melbourne, Victoria, Australia) and bred with CD1 females (Charles River Laboratories, Calco, Italy, http://www.criver.com) at the animal facilities of Istituto Superiore Sanità. Littermates were genotyped by polymerase chain reaction (PCR), and those found not to carry the mutation used as wild-type controls. All the experiments were performed with 6–10-month-old male littermates, according to protocols approved by the institutional animal care committee.

Thrombopoietin Treatment

Recombinant murine TPO was injected i.p. into wild-type mice on five consecutive days (100 μg/kg of body weight/daily), as described [35]. Mice were sacrificed either before the first injection or at 7 or 14 days after it for further analysis (described below).

Hematological Parameters

Blood was collected from the retro-orbital plexus into ethylene-diamino-tetracetic acid-coated microcapillary tubes (20–40 μl per sampling), and hematocrit and platelet counts were determined manually [35].

Histological Analysis

The ear, spleen, and femur were fixed in 10% (vol/vol) phosphate-buffered formalin (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), paraffin-embedded, and cut into 2.5–3-μm sections according to standard procedures. Slides of consecutive sections were dewaxed, rehydrated, and stained with regular and acidified toluidine blue (Sigma-Aldrich), as described [25]. Cell metachromasia is defined by the color acquired by the cytoplasmic granules after acidified toluidine blue staining. The granuli of immature and mature mast cells are, respectively, blue (metachromaticneg) and red (metachromaticpos) [25]. Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining was performed with the In Situ Cell Death Detection Kit (Boehringer Mannheim, Mannheim, Germany, http://www.boehringer.com) using, as negative controls, sections treated with a fluorescent deoxyuridine triphosphate-free TUNEL reaction mixture. Cells obtained by peritoneal lavage were cytospun onto glass slides (Cytospin 3; Thermo Shandon Inc., Pittsburgh, http://www.thermo.com) and stained with either May-Grunwald/Giemsa (Sigma-Aldrich) or Alcian Blue. For electron microscopy, samples were fixed in glutaraldehyde (2.5%), postfixed in OsO4, and embedded in SPURR resin (Polysciences Inc., Warrington, PA, http://www.polysciences.com) [25]. Light microscopy was analyzed with a Leica light microscope (Leica, Heidelberg, Germany, http://www.leica.com) equipped with a Coolsnap video camera for computerized images (RS Photometrics, Tucson, AZ, http://www.photometrics.net), whereas transmission electron microscopy was performed with the EM 109 (Carl Zeiss, Oberkochen, Germany, http://www.zeiss.com).

Flow Cytometry Analysis

Cells were labeled with phycoerythrin (PE)-CD117 (which recognizes c-Kit) coupled with either fluorescein isothiocyanate (FITC)-CD34, -Annexin V or with -CD45R/B220, as a control. Additional antibodies used in the study were FITC-Mac3, -CD61, and -CD71 and PE-Ly-6G (Gr1), -CD41, and -TER119. The expression of FcεRI was revealed by sequential incubations with the monoclonal mouse anti-DNP-IgE (clone SPE-7; Sigma-Aldrich) and FITC-conjugated rat anti-mouse IgE [25]. Unless otherwise stated, all the antibodies were from BD Pharmingen (San Diego, http://www.bdbiosciences.com/index_us.shtml) and were incubated at a concentration of 1 μg per 106 cells for 30 minutes on ice. Cell fluorescence was analyzed with either a Coulter Epix Elite ESP (Beckman Coulter, Miami, http://www.beckmancoulter.com) or a FACSAria (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com). Nonspecific fluorescent signals were gated out with appropriate fluorochrome-conjugated isotype controls, and dead cells were excluded by propidium iodide staining.

Cell Purification

Mast cells were purified by sorting, as described [25]. Briefly, peritoneal cells were first incubated with FITC-conjugated CD45R and immunodepleted with a monoclonal mouse anti-fluorescein antibody-coated microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany, http://www.miltenyibiotec.com), as described by the manufacturers. The B220-negative cell fraction was then incubated with PE-CD117, and CD117high cells (∼50% of B220neg cells) were isolated with the FACSAria (>95% CD117high after reanalysis). Cells in the prospective stem/progenitor cell gate [39] were sorted with the FACSAria from the marrow of wild-type mice incubated with PE-CD117 and FITC-CD34. The sorted CD117pos/CD34pos cells were >90% pure by fluorescence-activated cell sorting reanalysis and gave rise to >90 CFU-GM-derived colonies per 100 plated cells in clonogenic assay [40].

Culture of BMMC

BMMC were established from marrow (and spleen) light-density cells (1–2 × 105 cells per milliliter) or sorted CD117pos/CD34low cells (0.2–2 × 104 cells per milliliter), as described [11, 25, 40]. The cultures were stimulated with SCF (100 ng/ml) and IL-3 (10 ng/ml) with or without the presence of TPO (50 ng/ml), replenished with fresh growth factors, and demipopulated every 3–4 days, as necessary, to keep the cell concentration <106 cells per milliliter.

Serotonin Release Assay

Peritoneal B220neg cells or day 21 BMMC-derived cells (1 × 106 cells per milliliter) were incubated for 6 hours at 37°C with 3H-serotonin (5-hydroxy(3H)tryptamine trifluoroacetate, 2 μCi/ml) (84.0 Ci/mmol; Amersham Biosciences, Piscataway, NJ, http://www.amersham.com), washed twice, and either lysed with Triton X-100 (1% vol/vol), as a measure of the total amount of 3H-serotonin incorporated, or incubated again (20 × 106 cells per milliliter) for 1 hour on ice with the monoclonal mouse anti-DNP-IgE (10 μg/ml). After incubation, aliquots of 0.4 × 106 cells were resuspended in 50 μl and stimulated for 15 minutes at 37°C with the rat monoclonal αIgE (2 μg/ml, R35–72; BD Pharmingen). Reactions were terminated with 50 μl of cold Hanks' balanced saline solution, cells were removed by centrifugation, and the level of 3H-serotonin in the supernatants was measured with the Packard 1600TR liquid scintillator counter (PerkinElmer Life and Analytical Sciences, Boston, http://www.perkinelmer.com) and expressed as a percentage of the total amount of 3H-serotonin incorporated (according to the following algorithm: cpm in supernatants ÷ cpm in Triton lysates × 100) [5, 25].

RNA Isolation and Quantitative Reverse Transcription-PCR

Total RNA was prepared with Trizol (Gibco-BRL, Gaithersburg, MD, http://www.gibcobrl.com) and retrotranscribed with random primers using the SuperScript III kit (Invitrogen, Bethesda, MD, http://www.invitrogen.com). Gene expression was quantified with the TaqMan PCR kit (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com) and predeveloped custom-made oligos, whose sequence is available upon request, using either the ABI Prism 7700 or the 7300 Sequence Detection System (Applied Biosystems). Glyceraldehyde-3-phosphate dehydrogenase was also quantified in each reaction. Results were analyzed with the SDS program (version 1.9; Applied Biosystems) and expressed in arbitrary units, using the amplification of GPDH as calibrator, according to the following algorithm: ΔCT = [CTX − CTGPDH], where CT is the threshold cycle of the gene analyzed. Results are presented as 2math image.

Statistical Analysis

Statistical analysis was performed by analysis of variance using Origin 3.5 software for Windows (OriginLab Corp., Northampton, MA, http://www.OriginLab.com).

Results

TPO Treatment Decreases Mast Cell Differentiation in Mice

Figure 1 and Table 1 present the number and differentiation profile of dermal mast cells observed in mice treated with TPO. The platelet number in blood represents control for the efficacy of the TPO treatment.

Figure Figure 1..

The apparent increased number of immature (metachromaticneg after toluidine blue staining) and mature (Alcian Blue-Safraninepos) dermal mast cells observed in mice after TPO treatment is associated with high levels of apoptosis (terminal deoxynucleotidyl transferase dUTP nick-end labeling [TUNEL]pos). Shown are toluidine blue (A–C), Alcian Blue/Safranine (D–F), and TUNEL (G–I) staining of ear sections from untreated (A, D, G) and TPO-treated (day 7 [B, E, H] and day 14 [C, F, I]) wild-type mice. Magnification, ×40 (panels) and ×100 (insets). Abbreviation: TPO, thrombopoietin.

Table Table 1.. Frequency of mast cell precursors (metachromaticneg after toluidine blue staining) and of mature (Alcian Blue/Safraninepos or CD117pos/FcϵRIpos) mast cells and of apoptotic (TUNELpos or Annexin Vpos) cells in dermis of ear and in peritoneal cavity from wild-type mice, either untreated or at days 7 and 14 of TPO treatment, as indicated
original image

As expected [35], TPO treatment significantly increased the number of platelets present in the blood at day 7 (1.8 × 106 vs. 0.9 × 106 platelets per microliter, respectively; p < .01). The number of circulating platelets remained higher than normal, although no longer statistically different, up to 14 days but returned to pretreatment levels by day 21 (Table 1; data not shown).

TPO treatment increased the number of mast cell precursors (metachromaticneg after toluidine blue staining) by threefold at days 7–14 and that of mature mast cells (Alcian Blue/Safraninepos) by twofold at day 14 in the dermis of the ear (Fig. 1; Table 1). Many of the cells, however, underwent apoptosis. In fact, whereas TUNEL staining labeled the nuclei of a limited number of cells (<20 cells per mm2) in the dermis of the ear from untreated mice, numerous cells in the dermis just below the epithelium of the ear were TUNELpos 14 days after TPO treatment (600 ± 43 cells per mm2; Fig. 1; Table 1).

TPO treatment also increased the total cell number (by 2–4-fold) and the frequency of CD117highFcεR1pos cells (by 3–40-fold) present in the peritoneal cavity at days 7–14 (Fig. 2A; Table 1). All these values returned to pretreatment levels by day 21 (data not shown). Similar to what had been observed in the dermis, numerous serosal mast cells from TPO-treated mice were apoptotic. In fact, Annexin Vpos cells were rare (∼1%) among CD117high cells in the peritoneal cavity of untreated mice but represented ∼60% of those from TPO-treated animals (Fig. 2A; Table 1). The pattern of TUNEL and Annexin V staining of dermal and serosal mast cells is consistent with the hypothesis that TPO triggers apoptosis of mast cells in vivo between day 7 and day 14 of treatment.

Figure Figure 2..

TPO treatment reduces the number of serosal mast cells by inducing apoptosis. (A): Fluorescence-activated cell sorting analysis with CD117/FcεRI, CD117/Annexin V, or CD117/irrelevant antibodies (negative controls, gray curves and dotted lines) of serosal mast cells from wild-type mice, either untreated or 7 and 14 days after TPO treatment. Gates defining CD117pos cells are indicated in the left panels. (A1–A3): May-Grunwald staining of representative cells is presented for comparison. Magnification, ×40. (B): Expression of genes encoding growth-factor receptors (Mpl/c-Kit), transcription factors (Mitf/Gata1/Gata2), antiapoptotic proteins (Bcl2), and MMCPs (MC-CPA/MMPC-6/MMPC-7) in CD117high cells from wild-type mice either untreated or 7 and 14 days after TPO treatment (mean [±SD] of three separate experiments performed in triplicate). Values statistically different from untreated controls are indicated by * (p < .05) and ** (p < .01). (C): Serotonin released by serosal mast cells from wild-type mice upon IgE+αIgE stimulation, either untreated or 7 or 14 days after TPO treatment. Results are expressed as percentage of total serotonin incorporated (i.e., cells lysed with Triton, 100%, ∼30,000 cpm per 105 cells; mean [±SD] of three independent experiments performed in duplicate). *, statistically different (p < .01) from controls. Abbreviations: aIgE, anti IgE antibody; d, days; FITC, fluorescein isothiocyanate; FSC, forward side scatter; MC-CPA, mast cell carboxypeptidase A; MMCP, mast cell specific protease; TPO, thrombopoietin.

TPO treatment greatly affected the morphology of serosal mast cells, which became smaller and contained fewer granules than the corresponding cells from untreated controls (Fig. 2A). These morphological changes are associated with numerous alterations in differentiation profile. By flow cytometry, serosal mast cells from mice after 7 days of TPO treatment expressed 10-fold less CD117 (800 vs. 8,700 arbitrary fluorescent unit (AFU) per cell, respectively) and 1-log more FcεRI (1.1 × 103 vs. 0.1 × 103 AFU per cell, respectively) than normal (Fig. 2A). By real-time reverse transcription-PCR, serosal mast cells from TPO-treated mice expressed normal levels of Mpl, c-Kit, and Gata2 but significantly lower levels of Bcl2 and Mitf (both by 30%–50%; p < .05–.01), MMCP-6, and MMCP-7 (both by ∼1-log) and significantly higher levels of Gata1 (by 50% at day 7) (Fig. 2B). Finally, after IgE/αIgE stimulation, serosal mast cells from TPO-treated mice released significantly less (by 50%) serotonin than normal cells (Fig. 2C).

TPO Inhibits Mast Cell Differentiation in BMMC Cultures

The effects of TPO on mast cell differentiation were further analyzed in BMMC seeded with marrow cells from wild-type (and Mplnull, as negative control) mice. TPO had no effect on the number (6.2 ± 0.7 × 107 vs. 7.2 ± 0.5 × 107 cells per flask by day 21; p > .05) or morphology (Fig. 3A) of cells obtained in BMMC cultures from Mplnull mice. In contrast, in BMMC cultures from wild-type mice, TPO increased the number of mast cell precursors obtained at day 14 (4.1 ± 0.2 × 107 vs. 2.3 ± 0.2 × 107cells in flasks containing or not containing TPO, respectively; p < .05) by approximately twofold but reduced that generated at day 21 (14.2 ± 3.2 × 107 vs. 22.5 ± 5.5 × 107 cells per flask, respectively; p < .05) by approximately twofold. Furthermore, although more than 90% of the cells obtained after 21 days in BMMC cultures (both wild-type and Mplnull) were mature cells (c-KithighCD34highFcεRIhigh; Fig. 3A; [11, 13, 25]), only a minority (∼25%) of those developed at day 21 in the presence of TPO expressed CD34, and none were FcεRIpos (Fig. 3A). In addition, wild-type BMMC-derived cells obtained in the presence of TPO had few cytoplasmic granules (Fig. 3A).

Figure Figure 3..

TPO hampers mast cell differentiation in bone marrow-derived mast cell (BMMC) cultures. (A): Fluorescence-activated cell sorting analysis for CD117/CD34, CD117/FcεRI or CD117/irrelevant antibodies (gray curves), May-Grunwald staining of day 21 BMMC-derived wild-type and Mplnull cells, stimulated with SCF/IL-3 or SCF/IL-3/TPO, as indicated. Similar results were obtained in five separate experiments. (B): Serotonin released upon IgE+αIgE stimulation by BMMC-derived mast cells obtained in either the absence or the presence of TPO, as indicated. Results are expressed as percentage of total levels of serotonin incorporated (i.e., cells lysed with Triton). Shown is the mean (±SD) of five separate experiments performed in duplicate. *, statistically different (p < .05) from cultures without TPO. Abbreviations: FITC, fluorescein isothiocyanate; IL, interleukin; SCF, stem cell factor; TPO, thrombopoietin.

Table 2 presents the expression profile of wild-type mast cells obtained under BMMC conditions. All the genes analyzed were expressed by wild-type BMMC-derived cells at levels either slightly (Mitf, Bcl2, and MMCP-6, by twofold; p < .05) or greatly (c-Kit, Gata1, Gata2, MC-CPA, and MMCP-7 by 2-log; p < .01) higher than those expressed by the corresponding serosal mast cells (Fig. 2B; Table 2). The expression profile of BMMC-derived cells from cultures seeded with Mplnull mice was similar to that of the corresponding wild-type cells, with the exception of a lower expression (by ∼1.5-log; p < .01) of MMCP-7 (Table 2). TPO did not affect the expression profiling of Mplnull BMMC-derived cells but significantly reduced the levels of c-Kit, Mitf, Gata2, Bcl2, MMCP-6, and MMCP-7 expressed by wild-type BMMC-derived cells (Table 2). Finally, following IgE/αIgE stimulation, BMMC-derived wild-type mast cells obtained in the presence of TPO incorporated and released 30% less serotonin than cells obtained in control cultures without TPO (Fig. 3B).

Table Table 2.. Levels of gene expression in mast cells obtained in BMMC from wild-type and Mplnull mice in the presence or absence of TPO
original image

Effect of TPO on Mast Cell Differentiation from Purified Murine Stem/Progenitor Cells

To further assess whether TPO directly affects mast cell differentiation and to define the differentiation stage targeted by this growth factor, the effect of TPO on the number and phenotype of cells obtained in BMMC cultures seeded with wild-type CD117pos/CD34pos progenitor cells was analyzed under conditions of limiting dilution (200 to 2 × 104 cells per milliliter) (Fig. 4). By day 14, purified CD117pos/CD34pos cells generated cells with the CD117highCD34neg/lowFcεRIneg (Fig. 4C) CD41neg/CD61low, Ter119low/CD71neg, Gr1neg, and Mac3neg (not shown) phenotype. As such, these cells represent a homogeneous population of mast cell precursors completely devoid of megakaryocytic, erythroid, granulocytic, or monocytic cells. By day 21, the number of cells present in the cultures increased in proportion to the number of cells originally seeded (Fig. 4B). These day 21 BMMC-derived cells have the morphology of large granulated Alcian Bluepos cells with the antigenic profile of mature mast cells, CD117highCD34neg/highFcεRIhigh (Fig. 4B, 4C). The presence of TPO exerted no effect on the number and phenotype of mast cell precursors generated in these cultures at day 14 (Fig. 4B, 4C) but dramatically decreased the number of cells obtained at day 21 (Fig. 4B). This result confirms that TPO hampers the transition from precursor cells to mature mast cells occurring between day 14 and day 21 of BMMC culture.

Figure Figure 4..

TPO hampers mast cell differentiation of purified progenitor cells cultured under conditions of limiting dilution. (A): CD117low/CD34low gate used to purify progenitor cells from the bone marrow of wild-type mice and reanalysis of the sorted cells for purity. (B): Total cell number obtained after 14 (left) and 21 (right) days in bone marrow-derived mast cell (BMMC) cultures seeded with increasing numbers (200 to 2 × 104 cells per milliliter) of CD117low/CD34low cells and stimulated with or without TPO. The morphology of representative cells is presented in the insets. Similar results were obtained in two additional experiments. (C): Fluorescence-activated cell sorting analysis for CD117/CD34, FcεRI (gray area), and irrelevant antibody (blue curve) of cells obtained at days 14 and 21 of BMMC culture stimulated with stem cell factor/interleukin-3. The profile of the cells obtained in the presence of TPO was similar to that obtained in its absence at day 14 and could not be determined, because of the low cell number, at day 21. Abbreviations: FITC, fluorescein isothiocyanate; TPO, thrombopoietin.

Discussion

We have previously shown that Mpl, both as mRNA and as protein, is expressed in murine mast cells [34]. Here we show that forced activation of Mpl by TPO administration impaired mast cell differentiation in vivo and in vitro. In vivo, TPO apparently increased the number of dermal (Fig. 1) and serosal (Fig. 2) mast cells (Table 1). The majority of mast cells observed in TPO-treated mice, however, were either TUNELpos or Annexin Vpos and were therefore in apoptosis (Figs. 1, 2A). In addition, TPO-treated mast cells had impaired morphology (Fig. 2A), expression profiling (Fig. 2B), and serotonin-releasing activity (Fig. 2C). These effects were mimicked by TPO in vitro in BMMC cultures. In fact, addition of TPO to these cultures transiently increased, by ∼2.5-fold, the number of mast cell precursors generated by day 14 but decreased the overall number (Fig. 3) and maturation state (low FcεRI expression [Fig. 3A], poor MMCPs profiling [Table 2], and low serotonin-releasing activity [Fig. 3B]) of cells present after 21 days in BMMC cultures. Although apoptosis was not formally assessed, mast cells obtained in vitro in the presence of TPO, like those obtained in vivo, expressed low levels of Bcl2, implying the presence of high death rates in TPO-stimulated cultures.

As already observed for the Mplnull mutation, TPO treatment induces changes in expression profiling consistent with the biological alterations induced in the cells. TPO treatment modestly (approximately twofold) increases Gata1 and its target FcεRI [32, 33] but strongly reduces the expression of cKit on the cell surface and Mitf and its target genes, Bcl2 [28] and MMCP-6 [26]. This latter result indicates the SCF/Mitf/PIAS3/STAT3/Bcl2 network [37] as a possible downstream effector of the TPO-induced apoptosis of mast cells.

According to two independent studies, BMMC cultures seeded with human CD34pos cells increase cellular output when treated with TPO [41, 42]. The increases in the number of mast cell precursors induced by TPO in human [41, 42] and murine (this study) BMMC cultures are similar and probably due to expansion of the stem/progenitor cell compartments. Furthermore, TPO-exposed human mast cells, like the murine cells described here, express lower levels of FcεRI [42]. Although the expression of Mpl in human mast cells is debatable [41, 42], Mpl expression has been determined in murine cells [34]. In contrast with the human studies, however, the major effect of TPO observed here is induction of apoptosis. It is possible, though, that the effects of TPO on mast cell maturation are species-specific. Additional studies in humans, comparing mast cell differentiation in normal donors and in patients carrying either an inherited [43] or acquired [36, 44] mutation of Mpl or of its downstream target Jak2, will clarify the full spectrum of TPO activity in human mastocytopoiesis.

TPO not only is an important regulator of megakaryocytopoiesis but also serves as an important component of the stem cell niche [45]. As recently shown by Arai and Suda [46], TPO, produced by the osteoblasts on the marrow endosteum, regulates, together with SCF, the stem cell activity in the niche. The high apoptotic rates displayed by mast cells from TPO-treated mice suggest that TPO might represent the factor preventing stem cell differentiation toward the mast cell lineage in the niche. Therefore, TPO may act as the physiological regulator limiting the proliferation of mast cells to extramedullary sites. On the other hand, the observation that TPO treatment alters Gata1 and Mitf expression, as well as the expression of their target genes MC-CPA and MMCP-6, suggests that TPO might be involved in the regulation of the lineage switch between serosal and dermal mast cell differentiation in response to parasite infection [47]. This interesting possibility will be addressed in dedicated experiments that will couple the use of prospectively isolated progenitor cells with functional studies in the C57BL/6-KitW-sh transplantation model [4].

Conclusion

Our data are consistent with the regulatory model of mast cell differentiation described in Figure 5, according to which TPO boosts the number of mast cell precursors but prevents their final differentiation output. It achieves the former by promoting expansion of the hematopoietic stem/progenitor cell compartments and the latter by prohibiting the transition between precursor and mature cells.

Figure Figure 5..

A model for the biological activity of thrombopoietin (TPO) on mast cell differentiation. The model is based on results presented in this paper and by Maeda et al. [33] and suggests that, in addition to the proliferative effect exerted at the stem/progenitor cell level, TPO affects mastocytopoiesis by blocking the transition from precursor (CD117posCD34neg/pos) to mature (CD117posCD34highFcεRIhigh) cells. Such inhibition involves repression of Mitf expression that in turn alters the apoptotic (by reducing Bcl2 expression) and differentiative (by reducing MMCP-6 expression) program of the cells [33]. It is suggested that TPO affects Mitf expression indirectly by downmodulating cKit expression on the cell surface and suppressing the stem cell factor/Mitf/PIAS3/STAT3 network described in mast cell by Sonnenblick et al. [37]. Abbreviations: MC-CPA, mast cell carboxypeptidase A; MCP, mast cell-restricted proliferator cells; MMCP, mast cell specific protease; TPO, thrombopoietin.

Disclosure of Potential Conflicts of Interest

M.N. is employed by a company (Kirin) whose potential products (thrombopoietin and the AMM2 anti-Mpl antibody) were used in the present work.

Acknowledgements

Recombinant rat SCF was provided by Amgen (Thousand Oaks, CA, USA, MTA No. 19982634-005). This work was partially presented at the Hematopoietic Stem Cell VI meeting (September 14–16, 2006), Tubingen, Germany. This study was supported by National Cancer Institute Grant P01-CA108671. Fabrizio Martelli performed cytofluorimetric analysis. Barbara Ghinassi performed cell purification and culture. Rodolfo Lorenzini supervised the mouse colony and discussed results. Alessandro M. Vannucchi performed histologic analysis and discussed results. Rosa Alba Rana performed electron and optical microscopy. Mitsuo Nishikawa analyzed the data and discussed results. Sandra Partamian analyzed the data and wrote the paper. Giovanni Migliaccio designed research, analyzed the data, and wrote the paper. Anna Rita Migliaccio designed research, supervised the experiments, analyzed the data, and wrote the paper.

Ancillary