Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Thrombopoietin (TPO), the ligand of c-mpl, was molecularly cloned in 1994 and has been reported to be a main cytokine in the regulation of thrombopoiesis (Kaushansky, 1998). Subsequently, TPO, alone or in combination with other cytokines, such as interleukin-3 (IL-3), kit-ligand (KL), flk2/flt3 ligand, or erythropoietin (EPO), has also been shown to support the proliferation of both erythroid progenitors (Kobayashi et al., 1995) and primitive hematopoietic progenitor cells that yield granulocyte/erythrocyte/macrophage/megakaryocyte (GEMM) colonies (Itoh et al., 1996; Ku et al., 1996; Ramsfjell et al., 1996; Young et al., 1996; Tanimukai et al., 1997; Yoshida et al., 1997) and long-term repopulating hematopoietic stem cells (Sitnicka et al., 1996). Administration of TPO in vivo expanded not only colony-forming unit-megakaryocyte, but also burst forming unit-E and granulocyte/macrophage (GM)-colony-forming unit (Kaushansky et al., 1996). Conversely, in c-mpl deficient mice, the total number of hematopoietic progenitor cells, including multipotential and committed progenitors of multiple lineages, was reduced (Kimura et al., 1998). These data, therefore, show that TPO can stimulate primitive hematopoietic progenitor cells (Kaushansky et al., 1998).

The signal transduction pathways of TPO have been studied extensively. As with the EPO receptor (Ren et al., 1994; Li et al., 1996), phospholipase C- (PLC-) (Drachman et al., 1995) was phosphorylated on stimulation with TPO. As had been suspected, studies on UT-7/TPO cells demonstrated the existence of a PLC-protein kinase C (PKC) pathway, in addition to the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) and Ras-mitogen-activated protein kinase (MAPK) pathways (Kunitama et al., 1997). However, as signals resulting from stimulation of the same ligand are transduced along different pathways among physiological and transformed cells (Drachman et al., 1995; Ihle, 1996), it is uncertain whether these observations are applicable to normal hematopoiesis.

PKC (Nishizuka, 1988, 1995) has been reported to consist of at least 11 isoenzymes [, I, II, , , , , , , , (), and µ] and play an important role in the proliferation and differentiation of many kinds of cells. As to the localization and distribution of PKC isoforms in mouse hematopoietic cell lines, PKC- and PKC- were found to be abundant in both T- and B-lymphocytes but were not detected in myeloid cells. PKC- was restricted to the nervous system and was not detected in any of the hematopoietic cells studied. The -isoform of PKC was expressed at a high level in all lines and was predominant in B-cells and myeloid cells, whereas PKC- and - were detected in most cells, but only at a rather low level (Mischak et al., 1991). Using GM-colony-forming cells enriched with normal murine bone marrow cells by elutriation, it was shown that PKC-, -, and - were highly expressed, whereas the - and -isoforms were expressed at only relatively low levels (Whetton et al., 1994). However, the exact significance of each isoenzyme is not yet known (Goodnight et al., 1994). In the present study, we address the role of six PKC isoforms (, , , , , and ) among the signaling pathways of TPO-mediated proliferation of primitive hematopoietic progenitors to obtain a more accurate picture of the action mechanism of TPO.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Preparation. Male BDF1 mice (10-15 weeks old) were obtained from Charles River Japan (Atsugi, Japan). Cell preparation was conducted according to the method described previously (Musashi et al., 1997). Briefly, a single cell suspension was prepared from the pooled femurs of the mice. They had been intravenously injected with 150 mg/kg body weight of 5-fluorouracil (5-FU; Kyowa Hakko Kogyo Co, Tokyo, Japan) through their tail veins 2 days before examination (Day-2 post 5-FU marrow cells) to enrich their noncycling hematopoietic primitive progenitors. Lineage-negative (Lin), stem cell antigen-1-positive (Sca-1⁺) Day-2 post 5-FU marrow cells were isolated as described (Shih et al., 1992), with some minor modifications. Briefly, light density cells were separated from the Day-2 post 5-FU marrow cells by density centrifugation above a Ficoll-Conray (specific gravity, 1.077). They were then incubated at 4°C for 45 min in a cocktail of antibodies: anti-CD4 (Pharmingen, San Diego, CA), anti-CD8 (Pharmingen), B220 (CD45R, Pharmingen), Gr-1 (Pharmingen), and Mac-1 (CD11b, Pharmingen). After washing twice, sheep anti-rat IgG (Fc)-conjugated immunomagnetic beads (Dinabeads, Dynal A.S., Oslo, Norway) were added to the cell suspension and incubated at 4°C for 45 min. Lineage-specific antigen-positive (Lin⁺) cells were removed using a magnetic particle concentrator (Dynal), and Lin cells were recovered from the supernatant. The cell:bead ratio applied was 1:30. The Lin cells were then incubated with Sca-1 for 45 min at 4°C. After washing, magnetic beads were added to the cell suspension to a ratio of 3:1 and incubated for 45 min. Lin Sca-1⁺ cells were separated using a magnetic particle concentrator (positive selection), and cultured without stripping their magnetic particles. The average incidences of Lin⁺ cells in the Lin fraction and the Lin Sca-1⁺ fraction were less than 11 and less than 4%, respectively. We used three kinds of cells, including Day-2 post 5-FU marrow cells, Lin cells, and Lin Sca-1⁺ cells, according to the purpose of the experiments and the required cell number for each experiment.

Factors and Agents. Purified recombinant human TPO and rabbit neutralizing anti-human TPO polyclonal antibody were prepared by the TPO Production Group (Kirin Brewery Co., Ltd, Takasaki, Japan) (Kato et al., 1995; Tahara et al., 1998). Recombinant human EPO + recombinant murine IL-3 and recombinant murine KL were provided by the Kirin Brewery Co., Ltd. (Tokyo, Japan). Calphostin C (Kobayashi et al., 1989) and bisindolylmaleimide GF 109203X (Tollec et al., 1991; Wilkinson et al., 1993), selective inhibitors of PKC, were obtained from Kyowa Medics Co. (Tokyo, Japan) and Wako Junyaku Co. (Tokyo, Japan), respectively. PKC- translocation inhibitor peptide and its negative control (Mayne and Murray, 1998) were purchased from Calbiochem (San Diego, CA). The inhibitor peptide consists of octapeptide derived from the receptor for the activated C kinase binding site (V1 region) of PKC- and has been reported to specifically block translocation of PKC-. Its negative control consists of a scrambled peptide with an identical amino acid composition to the PKC- translocation inhibitor peptide. Lin cells were incubated in 300 µl of medium containing 75 µg of inhibitor peptide for 2 h at room temperature without cell permeabilization step with saponin or other detergents to avoid toxicity (Bassini et al., 1999). Although the inhibitor peptide used here was obtained through amplification by polymerase chain reaction of rat cDNA library, it is also available in mouse and human cells (Mayne and Murray, 1998; Bassini et al., 1999). Cycloheximide was purchased from Sigma (St. Louis, MO) and was used at 50 µg/ml (Wang et al., 1999).

Clonal Cell Culture. Methylcellulose cell culture was performed in 35-mm Lux suspension culture dishes (No. 5221R; Nunc, Inc., Naperville, IL) as described previously (Musashi et al., 1997). Each milliliter of culture contained 5 × 10⁴ Day-2 post 5-FU marrow cells, 2000 Lin cells, or 400 Lin Sca-1⁺ cells, -medium (Flow Laboratories, Inc., Rockville, MD), 1.2% methylcellulose (Wako Junyaku Co), 30% fetal bovine serum (Hyclone Laboratories, Logan, UT), 1% fraction V bovine serum albumin (BSA; Sigma), 2 U/ml recombinant human EPO, cytokines, and agents. Inhibitors were added to the cell suspensions at least 60 min before the addition of TPO. After incubation with TPO alone or in combination with inhibitors, cells were cultured in the presence of IL-3 and EPO after being washed twice. Dishes were incubated in a humidified atmosphere flushed with 5% CO₂ at 37°C. Colonies consisting of 50 or more cells were counted on specific days under an inverted microscope according to colony type, as indicated in the tables.

Serial Observation of Colony Formation from Progenitors. A total of 5 × 10⁴ Day-2 post 5-FU marrow cells were placed in plastic dishes of serum- and EPO- and IL-3-containing culture with or without TPO. The location and proliferation of emerging blast cell colonies were recorded daily as described previously (Musashi et al., 1997). Kinetic properties were calculated only for those colonies that later showed GEMM lineages. Cell doubling time, and the average number of days required for the colonies to reach 100 cells, were estimated by connecting the endpoints of their growth curves.

PKC Isoform Staining. Staining of PKC isoform was conducted according to the general method (Mischak et al., 1991) with some minor modifications. Briefly, Lin cells were incubated with TPO or medium as the control for 2 h. After washing twice, the cells were smeared onto glass slides using a Cytospin (Shandon Scientific Ltd., Pittsburgh, PA), fixed with formalin containing Triton X-100, and rinsed in phosphate-buffered saline. After blocking with normal goat serum, the smears were incubated with anti-PKC isoform , , , , , and  antibodies (Life Technologies, Rockville, MD) for 45 min followed by staining with fluorescein isothiocyanate-conjugated anti-rabbit IgG antibody (Life Technologies). The specificity of the antibodies was determined by antigen competition. Cell nuclei were stained with propidium iodide after treatment with RNase (Sigma), and the slides were mounted in glycerol containing p-phenylenediamine to retard fading during microscopy. Smears were observed under an MRC 1024 laser scanning confocal imaging system (Bio-Lad Laboratories, Hercules, CA) and photographed.

Statistical Analysis. Statistical analysis was performed using an ANOVA and a Student's t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

TPO Stimulation of Primitive Hematopoietic Progenitors. First, we determined the optimal concentrations of IL-3 (0.2 ng/ml) by conducting dose-response studies on the colony formation derived from the marrow cells of the 5-FU-treated mice. We also analyzed the colony formation derived from Day-2 post 5-FU marrow cells supported by IL-3 in combination with TPO at a concentration of between 1 and 100 ng/ml. We found that, although TPO alone hardly stimulated colony growth at all, it did augment colony formation up to 2-fold when combined with IL-3; this included not only megakaryocyte colonies, but also GM and GEMM colonies (Table 1). Maximal stimulation was observed at 100 ng/ml of TPO (Table 1). However, 10 ng/ml of TPO also exerted a synergistic effect. Although we had added EPO to the cultures to enable a differential count of the colonies, including erythrocyte-mixed colonies, TPO and IL-3 also exerted a synergistic effect in the absence of EPO (data not shown). Next, we studied the synergistic effect of TPO and IL-3 under serum-attenuated culture conditions, showing that the synergistic effect was independent of factors contained in the serum (Table 2). KL was required for maximal stimulation of colony formation under these conditions, whereas it was not in serum-containing culture, suggesting KL as the serum factor.

View this table: [in this window] [in a new window]   TABLE 1 Effect of TPO alone or in combination with IL-3 on colony formation of primitive hematopoietic progenitors Day-2 post 5-FU marrow cells (5 × 104) and 400 LinSca-1+ marrow cells were incubated with cytokines for 14 days in the presence of 2 units/ml EPO. Colonies consisting of 50 or more cell aggregates were counted according to colony type. Data represent mean ± S.D. of four dishes.

View this table: [in this window] [in a new window]   TABLE 3 Effect of neutralizing anti-TPO antibody on colony formation augmented by TPO After 30 min of incubation of TPO with anti-TPO antibody at 37°C, 2 × 103 Lin Day-2 post 5-FU marrow cells were added to the mixture of TPO and the antibody, and were further incubated for 1 h. After washing twice, the cells were cultured with IL-3 and EPO. Colonies were counted on day 14 of culture. Data represent mean ± S.D. of four dishes.

Serial Observation of Blast Cell Development in Day-2 Post 5-FU Marrow Cells. To examine the time course of colony development, we examined the culture dishes daily, and recorded the development of new blast cell colonies and their subsequent proliferation in the serum containing cultures with added IL-3 and IL-3 + TPO ("mapping studies", Fig. 1). The doubling time of individual colonies, based on lines connecting the two endpoints of their growth curves, was estimated to be 13.0 ± 3.1 h and 13.9 ± 4.1 h in cultures supported by IL-3 and IL-3 + TPO, respectively. The average number of days required for colonies to reach 100 cells was calculated to be 9.2 ± 2.2 and 8.0 ± 2.0 in cultures supported by IL-3 and IL-3 + TPO, respectively. Although the average number of days required for the colonies to reach 100 cells was statistically different between the two groups (p < 0.05), their growth rates were not. Therefore, these data indicate that TPO stimulated proliferation of the primitive progenitors by shortening their G₀ phase (Musashi et al., 1997) in accordance with a former report (Itoh et al., 1996).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Graphic presentation of change in cell number in individual blast cell colonies that later showed GEMM expression. The emergence of new blast cell colonies and their subsequent proliferation in four dishes in each group were recorded daily.

Effect of Preincubation with TPO Alone or in Combination with PKC Inhibitors on IL-3-Dependent Colony Formation. To infer the signal transduction pathways involved in conveying this proliferation effect of TPO, and focusing on PKC, Lin Day-2 post 5-FU marrow cells were preincubated for 2 h with TPO alone or in combination with calphostin C or bisindolylmaleimide GF 109203X, which have different IC₅₀ values for classical PKC and novel PKC. After washing twice, cells were cultured with IL-3. As shown in Table 4, calphostin C abrogated the enhancing effect of TPO at concentrations of 50 nM or more. However, the possibility that calphostin C exerted its inhibitory effect by blocking the signal of IL-3 could not be ruled out, because it suppressed the generation of colonies supported by IL-3 alone. On the other hand, a concentration of 200 nM GF 109203X or more did not affect IL-3-dependent colony growth, but did suppress TPO-augmented IL-3-dependent colony formation, partially but significantly. IC₅₀ of GF 109203X was reported to be lower than 20 nM against PKC-, -, and -, 130 to 210 nM against - and -, and 5800 nM against PKC- (Tollec et al., 1991; Hong et al., 1998). Because the TPO-augmented colonies were reduced at 200 nM, but not at 100 nM, it is possible that TPO might activate novel PKC but not classical PKC or atypical PKC. However, IC₅₀ was determined in kinase activity in a cell free system, hence its simple application to intact cells should be avoided. To ascertain the involvement of PKC-, we used the PKC--specific translocation inhibitor peptide. As shown in Table 5, the inhibitor peptide did not alter colony formation supported by IL-3, whereas it did suppress TPO-augmented colony growth partially compared with that augmented by TPO preincubated with the control peptide.

View this table: [in this window] [in a new window]   TABLE 4 Effect of selective PKC inhibitors on colony formation augmented by TPO 2 × 103 Lin Day-2 post 5-FU marrow cells were preincubated with PKC inhibitors for 1 h followed by another 2-h incubation with TPO. After washing twice, the cells were cultured with IL-3 and EPO. Colonies were counted on day 14 of culture. Data represent mean ± S.D. of four dishes.

View this table: [in this window] [in a new window]   TABLE 5 Effects of PKC- translocation inhibitor peptide on colony formation of Lin Day-2 post 5-FU marrow cells Lin or LinSca-1+ cells were preincubated with inhibitor peptide or its control peptide for 2 h. Without washing, the cells were incubated for a further 2 h with or without TPO. After washing, the cells were cultured with IL-3 plus EPO. Colonies were counted on day 14 of culture. Data represent mean ± S.D. of four dishes.

Effect of TPO on PKC Isoform in Primitive Progenitor Cells. To obtain direct evidence of the involvement of PKC in TPO signaling, we evaluated the change in intracellular distribution of each PKC isoform by TPO stimulation of Lin Day-2 post 5-FU marrow cells.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Expression of PKC isoforms in Lin Day-2 post 5-FU marrow cells. Cells were stained with polyclonal anti-PKC- (A), - (B), - (C), - (D), - (E), and - (F) antibodies and fluorescein isothiocyanate-labeled second antibody and observed under a confocal laser scanning microscope.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Translocation of PKC- on stimulation with TPO and its cancellation by calphostin C, GF 109203X, or PKC- translocation inhibitor peptide. Lin Day-2 post 5-FU bone marrow cells were incubated with calphostin C or GF 109203X for 1 h and another 2 h in combination with TPO. After washing twice, cells were stained with anti-PKC- antibody followed by nuclear staining with propidium iodide. In the case of the inhibitor peptide or its control peptide, the incubation time was prolonged for 2 h. A, control; B, TPO + vehicle; C, TPO + control peptide; D, TPO + calphostin C (100 nM); E, TPO + GF 109203X (100 nM); and F, TPO + inhibitor peptide.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we have shown that TPO can exert a synergistic effect in combination with IL-3 on the proliferation of primitive progenitors, at least, in part, through activation of PKC-. With regard to the mechanism of this synergy, two different theories have been postulated: first, that TPO triggers the recruitment of progenitors during the dormant stage of the cell cycle (G₀) (Itoh et al., 1996; Ku et al., 1996); and second, that TPO promotes the viability and inhibits the apoptosis of primitive progenitors (Ramsfjell et al., 1996). Based on the mapping study in this report (Fig. 1), TPO does not appear to influence the cell doubling time of blast cells, but does shorten the G₀ period of blast cell colony-forming cells. This observation is in accordance with the former mechanism, despite the difference in kind of progenitor cell involved. Although we cannot completely rule out the possibility that TPO exerts a synergistic effect through prevention of the apoptotic process, it would appear likely that this effect is due, at least in part, to a triggering of the recruitment of progenitors from their dormant phase into the cell cycle.

With regard to the postreceptor signaling of TPO, the roles of both the JAK/STAT and MAPK pathways in mitogenic response to TPO are still open to debate (Ihle et al., 1996; Dorsch et al., 1997). Therefore, we have focused on PKC, investigating which kind of PKC isoform was involved in its signaling pathways. The precise subcellular distribution and function of each PKC isoform is an active area of ongoing investigation. This study on 5-FU-resistant, Lin bone marrow cells showed that all six isoforms studied were expressed, and that although the amount of PKC- was abundant, PKC- was low. This distribution pattern is somewhat different from that of mouse GM-colony-forming cells, in which , , and  were predominant, whereas  and  were expressed at low levels (Whetton et al., 1994). In human CD34-positive cells, PKC-, -I, -II, -, -, and - were all detected by immunofluorescence analysis. PKC-, however, was unfortunately not studied (Marchisio et al., 1999). Taken together, these data suggest that there may be a characteristic distribution of PKC isoforms for each cell type.

PKC- is classified as a novel PKC, requiring micelles composed of phosphatidylserine and diacylglycerol, or phorbol ester, but not Ca²⁺, in its activation (Nishizuka, 1988, 1995). As activation of PKC usually results in translocation of the enzyme from cytosol to plasma membrane, it can be evaluated by serine/threonine kinase activity or intracellular translocation. We observed that PKC- was finely stained in the cytoplasm of unstimulated cells, and that stimulation with TPO caused changes in its fluorescent intensity and in its pattern, from fine granulation to blocky aggregation.

Since PKC isoform-specific inhibitors are not available to date, we have to combine inhibitors with different IC₅₀ values against each isoform to deduce its involvement. Although IC₅₀ of calphostin C against each PKC isoform has not yet been reported, it does appear to prevent novel PKC preferentially, rather than classical PKC (Mayne and Murray, 1998). Indeed, calphostin C reduced PKC- activity in plasma membrane (Mischak et al., 1993). GF 109203X derived from the structural lead provided by the nonselective protein kinase inhibitor, staurosporine, has the tendency to inhibit classical PKC preferentially, rather than novel and atypical PKC (Hong et al., 1998). We have shown that a low concentration of calphostin C and a relatively high concentration of GF 109203X canceled the changes in intracellular expression pattern of PKC- by TPO, and suppressed the synergistic effect of TPO on IL-3-dependent colony formation, implying that PKC- might be activated in the signaling pathways of TPO. Recently developed PKC- translocation inhibitor peptide is a great tool in evaluating the role of PKC-. As with GF 109203X, this inhibitor peptide blocked both TPO-stimulated changes in the expression pattern of PKC- and enhancement of colony growth. These observations are in accordance with previous reports that TPO activated PKC (Kunitama et al., 1997; Hong et al., 1998). The degree of suppression of TPO-enhanced colony formation by the inhibitor peptide was somewhat low, compared with that by GF 109203X, suggesting the possibility of incomplete cell permeabilization with the inhibitor peptide.

Since the significance of the activation of PKC in TPO signaling pathways remains unknown (Kunitama et al., 1997), the elucidation of the role of PKC- in cellular response to TPO signaling is an issue of some importance. PKC- has been reported to be important in cell proliferation and transformation: first, overexpression of PKC- in fibroblasts and epithelial cells resulted in an increase in the formation and growth rate of tumors in nude mice (Cacace et al., 1996; Gubiana et al., 1998); second, overexpression of PKC- in IL-3-dependent leukemia TF-1 cells prevented the apoptotic process in the absence of the cytokine through induction of bcl-2 expression (Baxter et al., 1992); third, GM-cerebrospinal fluid activated PKC- in GM-cerebrospinal fluid-dependent human leukemia cell line (Cai et al., 1997); and fourth, EPO also activated PKC- through PLC- in murine erythroleukemia cells (Li et al., 1996). In our previous studies, we showed that 12-O-tetradecanoyl phorbol-13-acetate (TPA) in combination with IL-3 increased colonies without triggering the traverse of the progenitors into cell cycling (Musashi et al., 1997). In this case, TPA did not appear to alter the subcellular localization of PKC- in Lin Day-2 post 5-FU marrow cells (data not shown). Taking this activation of PKC- by TPO, together with the triggering of cell cycle traverse by TPO, the  isoform of PKC may well correlate with cell proliferation rather than cell differentiation and maturation. However, it was reported recently that in c-Mpl-expressed UT-7 cells, PKC- and - mediated the mitogenic action of TPO, but not PKC- (Hong et al., 1998). The reason for this difference in TPO-activated PKC isoform is not clear, but it may result from the difference in the cells used in those other experiments or the relatively low concentration of TPO used (>10 ng/ml), compared with ours (100 ng/ml).

In summary, this study furnishes further evidence that TPO acts on the primitive hematopoietic progenitors by triggering their cell cycle traverse from G₀ into G₁/S, resulting in cellular proliferation. PKC- is involved in this signaling pathway from the TPO receptor (c-Mpl) to the cell cycle traverse and the stimulation of DNA synthesis. Since c-Myc (Ren et al., 1994; Li et al., 1996; Kunitama et al., 1997), Raf-1 (Gubiana et al., 1998), and Bcl-2 (Baxter et al., 1992) have been postulated to be the downstream of PKC-, further study is needed to clarify this point and the cross-talk between PKC- and the JAK/STAT or MAPK pathways.