Haematopoietic progenitor cells utilise conventional PKC to suppress PKB/Akt activity in response to c-Kit stimulation

Authors


Bengt Hallberg, Department of Medical Biosciences/Pathology, Build. 6M, Umeå University, S-901 87 Umeå, Sweden. E-mail: bengt.hallberg@medbio.umu.se

Summary

Receptor tyrosine kinase (RTK) c-Kit signalling is crucial for the proliferation, survival and differentiation of haematopoietic stem cells (HSCs). To further understand the mechanisms underlying these events we explored how the downstream mediators interact. The present study investigated the function of conventional protein kinase Cs (c-PKC) in c-Kit mediated signalling pathways in HSC-like cell lines. This analysis supported earlier findings, that steel factor (SF) activates c-PKC, extracellular signal-regulated kinase (Erk) and protein kinase B (PKB). The present results were consistent with an important role of c-PKC in the positive activation of Erk and for proliferation. Further, it was observed that c-PKC negatively regulated PKB activity upon SF stimulation, indicating that c-PKC acts as a suppressor of c-Kit signalling. Finally, these observations were extended to show that c-PKC mediated the phosphorylation of the endogenous c-Kit receptor on serine 746, resulting in decreased overall tyrosine phosphorylation of c-Kit upon SF stimulation. This report showed that this specific feedback mechanism of c-PKC mediated phosphorylation of the c-Kit receptor has consequences for both proliferation and survival of HSC-like cell lines.

The receptor tyrosine kinase (RTK) c-Kit (Yarden et al, 1987) is expressed by the majority of haematopoietic stem and progenitor cells in the human bone marrow and also in several non-haematopoietic tissues (Ashman, 1999). The ligand for the c-Kit RTK is steel factor (SF), also known as Stem Cell Factor. Mutations in the loci of either the receptor or the ligand (White spotted and Steel mutants respectively) (Chabot et al, 1988; Geissler et al, 1988; Huang et al, 1990; Zsebo et al, 1990) lead to lethal anaemia, haematopoietic stem cell (HSC) defects, mast cell deficiency and a series of non-haematological defects in the mouse (Russell, 1979; Lev et al, 1994). In humans, heterozygous mutations resulting in the alteration of c-Kit function are associated with diseases, such as piebaldism (Spritz et al, 1994; Akin & Metcalfe, 2004), which is a syndrome equivalent to the white spotting observed in rodents carrying heterozygous loss-of-function mutations (Geissler et al, 1988; Tsujimura et al, 1991). In addition, aberrant expression and gain-of-function mutations of the c-Kit RTK are associated with proliferation and survival in combination with oncogenic potential (Akin & Metcalfe, 2004).

Upon ligation of SF to c-Kit the receptor dimerises and is autophosphorylated on specific tyrosine residues in the cytoplasmic domain (Lev et al, 1992). This activation leads to the recruitment of intracellular proteins containing Src homology 2 (SH2) domains, which recognise and bind the specific phosphotyrosine motifs on the receptor and in turn become activated. Subsequently, various specific signalling pathways are activated (Pawson & Scott, 1997; Linnekin, 1999; Pawson & Nash, 2003).

Signalling downstream from c-Kit has been studied in several different systems, for example mast cells, different immortalised haematopoietic cell lines and also in transient transfection systems using both the original and chimaeric c-Kit receptors (Tsai et al, 1994; Wong et al, 1994; Yanai et al, 1999; Varnum-Finney et al, 2000; Sundstrom et al, 2003; Lennartsson & Ronnstrand, 2006). The present study used immortalised HSC-like cell lines, known as haematopoietic progenitor cells (HPC), to investigate c-Kit signalling in a physiologically relevant environment. These cell lines are derived by expression of the LIM-homeobox gene LHX2 in haematopoietic precursors generated from embryonic stem (ES) cells differentiated in vitro (Pinto do, et al 1998) The advantage of using this model system is that these HSC-like cell lines share many properties in vitro and in vivo with normal HSCs. They are dependent on SF for their maintenance in an immature state and can be long-term repopulated (Pinto do et al, 2002). The HPCs express cell-surface markers consistent with early fetal haematopoietic progenitors and they possess multipotentiality (Pinto do et al, 1998, 2001; Utsugisawa et al, 2006). These characteristics make the HSC-like cell lines unique when compared with previously described cell lines and therefore provide a valuable model system to study intracellular signalling in haematopoietic progenitor/stem cells.

We have previously reported that c-Kit mediated activation of the mitogen-activated protein kinase (MAPK) cascade was dependent on the activation of phosphoinositide-3 (PI-3) kinase in HPCs (Wandzioch et al, 2004). In contrast, differentiation of these HPCs to mast cells resulted in a signalling switch where Raf/MAPK/Erk kinase/extracellular signal-regulated kinase (Raf/Mek/Erk) activation changed from PI-3 kinase dependent to PI-3 kinase independent (Wandzioch et al, 2004). To further examine the signalling events downstream from the c-Kit receptor, we have now studied the role of the protein kinase C (PKC) family in this complex network of signalling pathways.

Protein kinase C activity was first described in the 1970s (Nishizuka, 2003) and was originally cloned by Parker et al (1986). Since then several isoforms of these serine/threonine kinases have been discovered and the PKC family is now divided into three groups (conventional, novel and atypical) depending on activation requirements (Liu & Heckman, 1998). The conventional PKCs (α, βI, βII and γ) (c-PKC) are Ca2+-dependent kinases, while the novel PKCs (δ, ε, η, θ and μ) function independently of Ca2+ (Ohno et al, 1988). The conventional and novel isoforms are activated by phosphatidylserines and phorbol esters, which, for some of the isoforms (PKCα, δ, ε,), results in degradation and depletion. The atypical PKC isoforms (ζ, λ, τ) function independent of Ca2+, phosphatidylserines and phorbol esters (Liu & Heckman, 1998; Parker & Murray-Rust, 2004). The large variety of PKC isoforms and the multitude of effects described have complicated studies concerning the biological role of PKC in the proliferation and differentiation of haematopoietic cells. Clarification of the signalling events by PKC downstream of the c-Kit receptor in haematopoietic progenitor/stem cells is important for the understanding of stem cell physiology but also important at many different levels in the haematopoietic hierarchy.

This study investigated the interconnection of Erk, protein kinase B (PKB, also known as Akt) and c-PKC pathways in HPC lines. As HPCs have the ability to self-renew as well as differentiate into all haematopoietic cell lineages, we reasoned that the interactions between these pathways might underlie crucial events, such as self-renewal, proliferation and cell-death. This study will help us to shed some light on the understanding of the mechanisms responsible for these events.

Materials and methods

Generation and maintenance of the HSC-like cell lines

Haematopoietic progenitor cell and bone marrow (BM)–HPC lines were generated by expressing LHX2 in haematopoietic progenitor/stem cells derived from ES cells differentiated in vitro or in haematopoietic progenitor/stem cells derived from adult BM, as previously described (Pinto do et al, 1998). The characteristics of these cell lines in vitro and in vivo have been thoroughly investigated previously (Pinto do et al, 1998, 2001, 2002). The cells were maintained in Iscove's Modified Dulbecco's Medium (IMDM; Gibco, Grand Island, NY, USA) supplemented with 5% fetal calf serum (FCS; Gibco), 1·4 mmol/l monothioglycerol (MTG; Sigma, St Louis, MO, USA), 50 U/ml Penicillin, 50 μg/ml Streptomycin and 50 ng/ml recombinant murine steel factor (R&D Systems, Minneapolis, MN, USA) at a cell density between 5 × 105 and 4 × 106/ml.

Antibodies, GST fusion proteins and inhibitors

Anti-c-Kit antibody and phospho-c-Kit (S746) were custom made at Agrisera (Vännäs, Sweden). The antigen used to generate the phospho-c-Kit (S746) was a phosporylated serine peptide (CSVRIGpSYIERDVTP) corresponding to the human c-Kit amino acids 741 and 754. Anti-Ras (# R02120) and anti-Erk2 (# 610123) antibodies were from BD Transduction Laboratories (Lexington, KY, USA). The antibodies phospho-p44/42 MAPK (Thr202, Tyr 204), p44/42 MAPK, phospho-PKB (Ser473), phospho-PKB (Thr308), PKB, phospho-PKC (pan), phospho-c-Raf (Ser338) and phospho tyrosine antibody PY100 were obtained from Cell Signaling Technology (Beverly, MA, USA). Anti c-Raf was from BD Biosciences (San Jose, CA, USA). 4G10 phospho tyrosine antibody was from Upstate (Lake Placid, NY, USA). Ro318220 was from Sigma, LY294002 was from Calbiochem (San Diego, CA, USA) and PP1 was from Alexis (Lausen, Switzerland). Glutathione S-transferase fusion proteins encompassing the Ras-binding domain of Raf-1 (GST-Raf-RBD) and the C-terminal SH2 domain of p85α (Blume-Jensen et al, 1994) have been described and were used according to Henriksson et al (2002) and Yonezawa et al (1992) respectively.

Cell lysis, immunoprecipitation (IP), Western blotting and in vitro PKB kinase activity assay

The cells were starved overnight (in IMDM without FCS and SF) and treated with inhibitors (as described in the figure legends) for 30 min and then stimulated with SF (50 ng/ml) for various times at 37°C. Thereafter, the cells were lysed on ice in lysis buffer (100 mmol/l NaCl, 50 mmol/l Tris-HCl pH 7·5, 1% Triton X-100, 1 mmol/l EGTA, 1 mmol/l EDTA, 10 mmol/l Benzamidine, 2 mmol/l ZnCl, 15 mmol/l MgCl, 1 mmol/l Na3VO4 supplemented with Complete protease inhibitors (Roche, Basel, Switzerland). Lysates were cleared by centrifugation at 21 000 g for 10 min at 4°C. Equal amounts of lysate were incubated with primary antibody for 2 h prior to incubation with protein G- or A-agarose (Amersham Biosciences, Uppsala, Sweden) for an additional hour. Alternatively the lysates were incubated with the described fusion proteins on glutathione sepharose beads. After four washes in lysis buffer, the samples were boiled in sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer and separated on 10% SDS-PAGE. The gels were transferred to Immobilion-P polyvinylidenedifluoride membranes (Millipore, Billerica, MA, USA) and blocked according to the manufacturers protocol before incubation with primary antibodies overnight. After incubation with horseradish peroxidase-conjugated secondary antibodies (Pierce, Rockford, IL, USA) the membranes were washed 6 × 10 min with PBSA-Tween 20 0·1% and the proteins were visualised using enhanced chemiluminescence (ECL, Amersham Biosciences). PKB kinase activity assays were performed using the Akt1/PKBα Immunoprecipitation Kinase Assay Kit from Upstate (#17-188) according to protocol. γ32P -ATP (3000 Ci/mmol) was obtained from Amersham Biosciences. For the evaluation of the phospho-c-Kit(S746) antibody, COS-1 cells were transfected with pcDNA3 plasmids containing wild type and mutated c-Kit RTK. 24 h post-transfection the medium was exchanged and the cells were incubated for another 10 h prior to starvation overnight and stimulation with SF for 5 min.

Fluorescent-activated cell sorting (FACS) analysis and proliferation study

For the flow cytometry analysis the cells were diluted to 5 × 105 cells/ml and treated as described in the figure legends. Subsequently, the cells were washed once in PBS and incubated for 30 min in Vindellövs propidium iodide (PI) solution (20 mmol/l Tris-HCL pH 7·5, 100 mmol/l NaCl, 50 μg/ml PI, 0·1% Nonidet P40, and 20 μg/ml RNAse) before analysis on a Becton Dickinson FACSCalibur. The percentage of dead cells in the control sample (cells in culture) was deducted from the starved cells. For the proliferation study the cells were resuspended in fresh medium with the addition of 0·5 mmol/l Ro318220 as indicated, at a density of 1·0 × 106 cells/ml. Prior to counting using a Bürkner chamber, the cells were stained with trypan blue (Gibco).

Results

c-PKC activates Erk downstream of Ras in HPCs

Initially, we investigated whether PKC activity, via stimulation with phorbol 12-myristate 13-acetate (PMA) in HPCs, is important for our earlier observation that Erk activation in HPCs is PI-3 kinase dependent (Wandzioch et al, 2004). As expected, PMA induced phosphorylation of the downstream targets Erk 1 and 2 (Fig 1A, compare lane 4 with 1) (Ueda et al, 1996). The activation of Erks was not dependent on PI-3 kinase because LY294002, a known physiological PI-3 kinase inhibitor (Vlahos et al, 1994), did not block this phosphorylation (Fig 1A, compare lane 4 with 5). Consistent with this, phosphorylation of Thr308 and Ser473 on PKB was not stimulated with PMA, as was observed upon stimulation with SF (Fig 1C and D, compare lane 2 with 4).

Figure 1.

 Phorbol 12-myristate 13-acetate (PMA) activates the mitogen-activated protein kinase (MAPK) cascade but not protein kinase B (PKB). Haematopoietic progenitor cells were incubated for 30 min with 20 μmol/l Ly294002 (lanes 3 and 5) prior to 5 min stimulation with SF 50 ng/ml (lane 2 and 3) or PMA 100 ng/ml (lane 4 and 5). Total cell lysates were analysed by Western blot with indicated antibodies (A–D and I). C-Raf immunoprecipitations were immunoblotted with phospho-c-Raf(S338) antibody (E) and stripped and reblotted with c-Raf antibody (F). GST-Raf-RBD affinity precipitations were immunoblotted with Ras antibody (G). Ras in total lysate was analysed as loading control (H).

Raf was also phosphorylated on Ser338 in a PI-3 kinase dependent manner upon stimulation with SF, because LY294002 impaired the phosphorylation of Raf (Fig 1E, compare lane 2 with 3). Raf was also phosphorylated on Ser338 upon stimulation with PMA, although in a PI-3 kinase independent manner (Fig 1E, compare lane 4 with 5). To monitor Ras activation we employed the Ras-GTP-dependent pulldown assay using the GST-Raf-RBD fusion protein (Taylor & Shalloway, 1996). In contrast to SF stimulation, we observed that Ras was not activated by PMA (Fig 1G, compare lane 4 with 2). As expected, PKCs were phosphorylated upon stimulation with both SF and PMA (Fig 1I). As loading controls, the filters were stripped and reprobed with Erk1/2 and C-Raf antibodies (Fig 1B and F). Ras in total cell lysate was used as control for the Ras pulldown assay (Fig 1H). As published previously (Wandzioch et al, 2004), SF stimulation of the Erks was PI-3 kinase dependent because LY294002 blocked both phosphorylation of PKB, Raf and Erks, but not the activation of Ras in HPCs. We also observed that PMA mediated the activation of Raf and Erks, but not Ras and PKB in HPCs.

PKC activity is necessary for SF-induced Erk phosphorylation and suppresses PKB phosphorylation

To investigate the role of c-PKC activity in SF/c-Kit mediated downstream signalling we used a pharmacological inhibitor, Ro318220, which is a general ATP analogue inhibitor of the PKC isoforms (Davies et al, 2000; Way et al, 2000). In addition to pharmacological inhibition, we employed overnight pretreatment of HPCs with PMA, a treatment known to downregulate PKCs (Liu & Heckman, 1998).

Cells were treated with PMA overnight or with the c-PKC inhibitor Ro318220 for 30 min prior to stimulation with SF for 5 min. As expected, Erk activation was blocked. However, we were intrigued to find that the phosphorylation of PKB was more pronounced if the cells were pretreated with PMA or with Ro318220 (data not shown). From these results we conclude that, firstly, PKC can stimulate Erk activation and, secondly, PKC negatively regulates PKB activity.

To further investigate SF signalling in HPCs with impaired PKC activity, we compared untreated cells with cells pretreated with Ro318220. Prior to harvest, cells were stimulated with SF for various times (Fig 2). Our results showed that inhibition of c-PKC mediated an increased and more sustained phosphorylation of both Thr308 and Ser473 of PKB compared with untreated cells (Fig 2A and B, compare lanes 7–11 with 2–6). Furthermore, Erk phosphorylation was drastically downregulated at all time points upon pretreatment with Ro318220 (Fig 2D). Thus, we conclude that in HPCs, c-PKC act on one hand to suppress PKB activity, and on the other hand to drive activation of Erks. To validate the amounts of inhibitor used during our investigation, we tested increasing doses of Ro318220 from 0·01 μmol/l to 50 μmol/l, which showed that the amount used was within its boundaries for suppressing c-PKC (data not shown).

Figure 2.

 Conventional protein kinase C (PKCs) regulate Erk, protein kinase B (PKB) and c-Kit activation. Haematopoietic progenitor cells were incubated for 30 min with 5 μmol/l Ro318220 (lane 7–12) prior to steel factor (SF) stimulation for 0–120 min. Total cell lysates were analysed by Western blot as indicated (A–E). C-Kit and phospho-tyrosine (P-Y) antibodies were used for immunoprecipitations (F–I). C-Kit immunoprecipitations were immunoblotted with P-Y (F), phospho-c-Kit(S746) (G) and c-Kit (H) antibodies. P-Y immunoprecipitations were immunoblotted with c-Kit (I).

PKC mediates serine phosphorylation of c-Kit and suppresses c-Kit activation

We then turned our attention to activation of the c-Kit RTK. In agreement with our earlier results, c-Kit phosphorylation peaked between 15 and 30 min (Fig 2F and Wandzioch et al (2004)). However, when cells were pretreated with the PKC inhibitor, increased tyrosine phosphorylation of the c-Kit receptor was observed in comparison with controls (Fig 2F and I, lanes 7–12 vs. 1–6).

Another approach for investigating whether c-Kit receptor activation is enhanced upon treatment with PKC inhibitor is to employ pull down assays with the SH2-domain of p85, the regulatory domain of PI-3 kinase (Reedijk et al, 1992). In these experiments we observed increased binding of c-Kit RTK in cells pretreated with Ro318220 prior to stimulation with SF compared with cells stimulated with SF only (Fig 3, compare lane 3 with 2).

Figure 3.

 The p85 subunit of PI-3 kinase binds c-Kit more strongly upon c-PKC inhibition. Haematopoietic progenitor cells were incubated for 30 min with 5 μmol/l Ro318220 (lane 3) prior to 15 min stimulation with steel factor (SF) (lanes 2 and 3). GST-p85-SH2 affinity precipitations were immunoblotted with c-Kit antibody (A). C-Kit (B) and GST-p85-SH2 (C) in total lysates were analysed as loading controls.

It has been shown that the exogenously expressed human c-Kit receptor can be phosphorylated on serine 741 and serine 746 in the kinase insert region in porcine aortic endothelial cells stably transfected with cDNA for the human c-Kit receptor (Blume-Jensen et al, 1995). To investigate if the endogenous mouse c-Kit RTK in our HPCs is similarly regulated, we developed a phospho-serine specific antibody towards serine 746 on the c-Kit RTK (Fig 4; see Materials and methods). To examine the specificity of this antibody, cells were transfected with plasmids expressing wild type c-Kit receptor or a substitution mutant c-Kit receptor (c-Kit S746A). Prior to harvest, the cells were stimulated with SF, and cell lysates were subjected to immunoprecipitation with c-Kit antibody followed by immunoblotting using tyrosine-phosphorylation specific antibody, serine 746 specific antibody or a c-Kit antibody (Fig 4A–C). Wild type c-Kit receptor was clearly phosphorylated both on tyrosine and on serine 746 in response to stimulation with SF (Fig 4A and B, lane 2). However, the c-Kit S746A receptor was not recognised by the serine 746 specific antibody, although the receptor was tyrosine phosphorylated upon stimulation with SF (Fig 4A and B, lane 4). Taken together, we concluded that this novel antibody displays specificity for c-Kit, which is phosphorylated on serine 746 within the intracellular kinase domain.

Figure 4.

 Demonstration of phospho-c-Kit Serine 746 antibody specificity. Cos-1 cells were transiently transfected with wild type c-Kit or c-KitS746A. Transfected cells were untreated (lanes 1 and 3) or stimulated with steel factor (SF) for 5 min (lanes 2 and 4) and c-Kit was immunoprecipitated. C-Kit IPs were immunoblotted with P-Y (4G10) (A), phospho-c-kit(S746) (B) and c-Kit (C) antibodies.

Upon examination of endogenous c-Kit in HPCs it was observed that the c-Kit receptor was phosphorylated on serine 746 upon stimulation with SF in vivo, and that this phosphorylation was abrogated if the cells were pretreated with Ro318220 (Fig 2G). In summary, inhibition of PKC activity mediated both a more sustained tyrosine phosphorylation of the endogenous c-Kit RTK, as well as a specific decrease in S746 phosphorylation, which may explain the resultant increased PKB activity. This is in agreement with findings by Blume-Jensen et al (1994) who demonstrated increased association of PI-3 kinase with c-Kit upon inhibition of PKC in stably transfected porcine aortic cells with the human c-Kit RTK.

Inhibition of PKC mediates an increased PKB activity

As analysis of PKB phosphorylation of the threonine 308 and serine 473 sites reflects phosphorylation status (Fig 2A and B), but not enzymatic activity, a PKB kinase activity assay was performed. PKB was immunoprecipitated from cells pretreated with different inhibitors and the activity was measured with the PKB-specific target peptide RPRAATF as substrate (Bozinovski et al, 2002). We approached the question of whether c-PKC activity negatively regulates PKB by comparing the endogenous PKB kinase activity in vitro in cells treated with Ro318220, PP1, an inhibitor shown to impair Src and c-Kit protein activities (Tatton et al, 2003), and LY294002. Here, SF-dependent BM-HPCs (generated by expressing LHX2 in HSCs derived from adult bone marrow) were employed (see Materials and methods).

Protein kinase B in vitro kinase activity was roughly sixfold higher when the cells were pretreated with the c-PKC inhibitor, when compared with cells that were stimulated with SF in the absence of inhibitor (Fig 5). Consistent with earlier reports, LY294002 impaired the SF-induced activation of PKB drastically in HPCs (Fig 5 and Wandzioch et al (2004)). Similar downregulation of PKB activity was observed using the PP1 inhibitor, which impairs c-Kit activity, in undifferentiated BM-HPCs.

Figure 5.

 Protein kinase B (PKB) kinase activity is increased upon protein kinase C (PKC) inhibition. In vitro PKB activity assay. Serum-starved haematopoietic progenitor cells were pretreated with 10 μmol/l PP1, 20 μmol/l LY294002 or 5 μmol/l Ro318220 for 30 min and stimulated with 50 ng/ml steel factor (SF) for 5 min. Mean fold activity compared with SF stimulated cells were calculated from four independent experiments, error bars indicate standard deviation.

Taken together, these data indicate that activation of c-PKC negatively regulates PKB activity in haematopoietic progenitors. Thus, c-PKC acts as a suppressor of PKB activity and this may have importance for HPC line proliferation or survival.

Downregulation of PKC influences survival and proliferation

To investigate the biological consequences of c-PKC inhibition we monitored the proliferation and survival of HPCs in the presence or absence of inhibitor. As shown previously (Wandzioch et al, 2004) it was observed that SF was necessary for HPC line viability, because no cells survived upon withdrawal of SF (data not shown and Wandzioch et al (2004). HPCs in fresh culture media treated with Ro318220 exhibited dramatically inhibited proliferation. In fact similar numbers of cells were observed per ml at the start of the experiment and after 24 and 48 h(Fig 6). These results clearly indicated that c-PKC activity was absolutely necessary for HPC proliferation, because cell numbers in the untreated HPC controls increased by 1·0 × 106 cells after 24 h and another 1·0 × 106 cells after 48 h. During the ongoing assay the percentage of viable cells was also monitored and no increase in dead cells could be detected. This can be compared with the withdrawal of SF, which induced cell death in HPCs and could be observed as early as after 4 and 8 h of starvation. However, HPCs treated with Ro318220 during starvation showed around 50% fewer dead cells both after 4 and 8 h (Fig 7). Thus, these observations clearly indicated that PKC activity was absolutely necessary for proliferation and that the enzymatic activity of c-PKC suppressed the PKB-mediated survival signal and the viability of the HPCs.

Figure 6.

 Proliferation of haematopoietic progenitor cells (HPCs) requires protein kinase C (PKC) in vivo. HPCs were cultured in the presence of fetal calf serum (FCS) and SF (bsl00001) or FCS, SF and 0·5 μmol/l Ro318220 (bsl00066). Cells were counted after 8, 24 and 48 h. The numbers represent the mean of three independent experiments in duplicate. Error bars indicate standard deviation.

Figure 7.

 Protein kinase C (PKC) inhibition rescues haematopoietic progenitor cells (HPCs) from starvation-induced apoptosis. PI incorporation was analysed by fluorescent-activated cell sorting. HPCs were deprived of fetal calf serum (FCS) and steel factor (SF) for 4 h and 8 h in the presence or absence of 0·5 μmol/l Ro318220. Cells in culture were included as control. (A) DNA histograms (4 h) with the percentage of apoptotic cells indicated. Control (left), starved cells (middle) and Ro318820-treated cells (right) are shown. (B) Percentage of apoptotic cells after 4 and 8 h starvation (26% and 35%; light grey bars) or starvation with Ro318820 (12% and 21%; dark grey bars).

Discussion

The present study aimed to determine whether the conventional PKC family is involved or regulates any of the important signalling pathways downstream of the c-Kit RTK in HSCs regarding self-renewal and differentiation. The molecular mechanisms of c-Kit signal transduction in HSCs are poorly characterised, mainly due to their rarity in the bone marrow and the requirement of large numbers of cells for the functional analysis of signal transduction. However, the importance of the c-Kit receptor is not in question, as haematopoietic cells depend on SF for growth, survival and function (Oliveira & Lukacs, 2003; Ronnstrand, 2004). Most of the CD34+ cells in the bone marrow express c-Kit, indicating an important feature for haematopoietic cells (Simmons et al, 1994). Generally, c-Kit expression is lost during differentiation, although mast cells retain a high level of expression even when fully differentiated (Oliveira & Lukacs, 2003).

This study employed HPCs, which are SF-dependent and share many properties with primary HSCs in vitro and in vivo, including surface markers, multipotentiality, capacity for self-renewal and long-term repopulation (Pinto do, et al 2002). These criteria make HPCs useful for the investigation of signalling mechanisms responsible for self-renewal, cell death and differentiation upon stimulation with SF.

It was observed that PMA treatment mediated phosphorylation of Raf and Erks, but not activation of Ras and PKB in HPCs. Furthermore, inhibition of c-PKC impaired the ability of c-Kit to induce phosphorylation of Erk, suggesting that c-PKC activity is necessary for Erk activation. Our results are consistent with an important role of c-PKC in the positive activation of Erk and proliferation, indicating a necessary role for PKC in HPC signalling (Schonwasser et al, 1998). This is in contrast to earlier studies, which have shown that Erk activation both in cell lines and in vitro kinase assays from c-Kit expressing non-haematopoietic cells is intact after PKC inhibition (Blume-Jensen et al, 1994). The reason for this dissimilarity is unknown, but differences in experimental conditions, together with the fact that non-haematopoietic cells might have a different mechanism for Erk activation compared with haematopoietic cells, may explain this difference. An additional explanation might be the difference in signalling, depending on c-Kit splice forms (Caruana et al, 1999; Voytyuk et al, 2003). Both of the described splice forms of c-Kit were expressed in our cells (Wandzioch et al, 2004), however, Blume-Jensen et al (1994) analysed endothelial cells transfected with the human c-Kit GNNK+ isoform.

In an effort to elucidate the mechanism behind our observations, it was observed that c-Kit was more extensively tyrosine phosphorylated in cells pretreated with c-PKC inhibitor when compared with untreated cells. In addition, this increased c-Kit phosphorylation appeared to be sustained for a longer period. Moreover, upon c-PKC inhibition, increased c-Kit RTK binding was found in SH2-p85α pull down experiments. This indicated an increased ability of c-Kit to activate PI-3 kinase. This is of interest, as PI-3 kinase generates the second messenger PtdIns(3,4,5)P3, an activator of PKB activity (Cantley, 2002).

It has been shown that the human c-Kit receptor can be phosphorylated on S741 and S746 in the kinase insert region of porcine aortic endothelial cells stably transfected with cDNA for the human c-Kit receptor (Blume-Jensen et al, 1995). Consequently, a SF-stimulated feedback loop involving c-PKC, leading to inhibition of c-Kit kinase activity, has been proposed. Utilising a phospho-serine specific antibody towards serine 746 on the c-Kit receptor, we observed that the endogenous c-Kit receptor was serine phosphorylated upon stimulation with SF. If cells were pretreated with Ro318220 prior to stimulation no serine phosphorylation could be observed. Thus, inhibition of c-PKC results in abrogated serine phosphorylation, together with a higher level of tyrosine phosphorylation on the c-Kit receptor for an extended time period, which may explain the observed increase in PKB activity. In agreement with this, PKB in vitro kinase activity was dramatically increased when the cells were pretreated with c-PKC inhibitors. Thus, in HPCs, activation of c-PKC negatively regulates stimulated PKB activity in response to SF.

The functional biological assays clearly demonstrated that the activity of c-PKC was absolutely necessary for proliferation and that the enzymatic activity of c-PKC was required to simultaneously suppress the PKB-mediated survival signal and the viability of the HPCs.

The present study found a novel role for c-PKC in the regulation of c-Kit signalling in HPCs. In conclusion, it was shown that signalling events by PKC downstream of the c-Kit receptor in haematopoietic progenitor/stem cells are important for the activation of Erk and PKB, which mediate signals for proliferation and survival. These results support a negative feedback loop mechanism where c-Kit activates c-PKC, which mediates a negative serine phosphorylation of c-Kit on residue 746. This feedback loop results in suppressed PKB activation and subsequent impaired cell viability. Our data also establish that c-PKC is necessary for Erk activation and, as a probable consequence, proliferation. Thus, this report clearly demonstrated the importance of PKC activity for c-Kit signalling in HPCs, which should contribute to the understanding of HSC physiology.

Acknowledgements

The financial support for this work was from the Swedish Cancer Society. RHP is a Swedish Cancer Society Senior Research Fellow. We are grateful to Jeanette Blomberg for expertise assistance with the FACS analyses.

Ancillary