We have identified a series of novel non-peptide compounds that activate the thrombopoietin-dependent cell line Ba/F3-huMPL. The compounds stimulated proliferation of Ba/F3-huMPL in the absence of other growth factors, but did not promote proliferation of the thrombopoietin-independent parent cell line Ba/F3. The thrombopoietin-mimetic compounds elicited signal-transduction responses comparable with recombinant human thrombopoietin, such as tyrosine phosphorylation of the thrombopoietin receptor, JAK (Janus kinase) 2, Tyk2 (tyrosine kinase 2), STAT (signal transducer and activator of transcription) 3, STAT5, MAPKs (mitogen-activated protein kinases), PLCγ (phospholipase Cγ), Grb2 (growth-factor-receptor-bound protein 2), Shc (Src homology and collagen homology), Vav, Cbl and SHP-2 (Src homology 2 domain-containing protein tyrosine phosphatase 2) and increased the number of CD41+ cells (megakaryocyte lineage) in cultures of human CD34+ bone-marrow cells (haematopoietic stem cells). These findings suggest that this series of compounds are novel agonists of the human thrombopoietin receptor and are possible lead compounds for the generation of anti-thrombocytopaenia drugs.
Patients with thrombocytopaenia currently have no choice but to undergo platelet transfusion, the only known effective treatment for preventing life-threatening haemorrhage. However, this entails problems such as viral and bacterial infections, transfusion reactions, refractoriness or alloimmunization . Although administration of recombinant human IL (interleukin)-11 has been approved to decrease the frequency of blood transfusions, its primary physiological function is not regulation of haematopoiesis , and it is associated with undesirable side-effects . Therefore there has long been a need to develop a more specific treatment that increases platelet number.
Thrombopoietin, which is produced primarily in the liver, was isolated in 1994 as a major cytokine that regulates the proliferation and differentiation of megakaryocyte-lineage cells and promotes platelet production. Thrombopoietin binds to receptors expressed upon the surface of haematopoietic stem cells, megakaryotic precursor cells, megakaryocytes and platelets. Subsequently, it triggers activation of JAK (Janus kinase) 2, STAT (signal transducer and activator of transcription) 3, STAT5, PI3K (phosphoinositide 3-kinase), the Ras-MAPK (mitogen-activated protein kinase) pathway and other regulatory proteins, resulting in the induction of megakaryocytopoiesis and thrombopoiesis [4,5]. Thus administration of recombinant human thrombopoietin was considered as a novel effective treatment for thrombocytopaenia . Several clinical studies have demonstrated that recombinant human thrombopoietin is useful for decreasing the extent of thrombocytopaenia [7–9]. However, adverse side-effects, such as the appearance of neutralizing antibodies against thrombopoietin that cause thrombocytopaenia, have been reported [10–12]. Therefore compounds that induce endogenous thrombopoietin or compounds with thrombopoietin-like activities should be safer alternative therapies and would present an important medical advantage over current treatments for thrombocytopaenia.
In the present study, we describe the discovery of novel non-peptide compounds that induce megakaryocytopoiesis in vitro using a high-throughput cell proliferation assay. The characterization of these compounds will be discussed.
MATERIALS AND METHODS
Recombinant human thrombopoietin, recombinant human GM-CSF (granulocyte/macrophage colony-stimulating factor) and recombinant mouse IL-3 were purchased from Genzyme. Rabbit anti-(human thrombopoietin receptor) antibody was purchased from IBL (Takasaki-Shi, Gunma, Japan). Affinity-purified rabbit polyclonal antibodies against JAK1, JAK2, JAK3, STAT1, STAT3, STAT5, Tyk2 (tyrosine kinase 2), SOS (son of sevenless), Grb2 (growth-factor-receptor-bound protein 2) Vav, SHIP (Src homology 2 domain-containing inositol phosphatase), SHP-2 (Src homology 2 domain-containing protein tyrosine phosphatase 2) and PLCγ (phospholipase Cγ), affinity-purified goat polyclonal antibodies against STAT6 and Cbl, and mouse monoclonal antibodies against JAK2, Tyk2, STAT6, Shc (Src homology and collagen homology), Grb2, Cbl, Vav, SHIP, SHP-2 and PI3K [p85 (85 kDa subunit of PI3K)] were all from Santa Cruz Biotechnology. Affinity-purified rabbit polyclonal antibodies against JAK3, Shc and PLCγ, mouse monoclonal antibodies against PI3K (p85), and a mouse monoclonal biotinylated anti-phosphotyrosine antibody (4G10) were purchased from Upstate Biotechnology. Mouse monoclonal antibodies against JAK1, STAT1, STAT3 and STAT5 were purchased from BD Transduction Laboratories.
Cells and cultures
The IL-3-dependent mouse pre-B-cell line Ba/F3 was obtained from the Riken Cell Bank (Tsukuba, Ibaraki, Japan). The GM-CSF- or erythropoietin-dependent human erythroleukaemic cell line TF-1 and the mouse T-cell lymphoma cell line EL4 were purchased from the A.T.C.C.
Ba/F3 cells were maintained in RPMI 1640 medium containing 300 mg/l L-glutamine, 10% (v/v) FBS (fetal bovine serum; HyClone) and 10% (v/v) WEHI-3-conditioned medium (as a source of mouse IL-3). TF-1 cells were maintained in RPMI 1640 medium supplemented with 5 ng/ml recombinant human GM-CSF. Ba/F3 cells transfected with human proto-oncogene c-mpl cDNA (Ba/F3-huMPL) were maintained in RPMI 1640 medium supplemented with 10% (v/v) WEHI-3-conditioned medium and 2 mg/ml G418 (Invitrogen). EL-4 cells were maintained in RPMI 1640 medium containing 10% (v/v) FBS. Human CD34+ bone-marrow haematopoietic progenitor cells were purchased from Cambrex (Lot 041047). These cells were isolated from human tissue obtained with informed consent in accordance with the Declaration of Helsinki (2000) of the World Medical Association. The present study was approved by the institutional ethics committee.
Construction of the
c- mpl expression plasmid
Full-length cDNA encoding human c-mpl was amplified by PCR using cDNA derived from the human leukaemia cell line KU812 and the following two pairs of oligonucleotide primers: 5′-ATGCCCTCCTGGGCCCTCTT-3′ and 5′-CACAGTCACAGGGAGGGA-3′; and 5′-ATCAGTGATTTCCTGAGG-3′ and 5′-TCAAGGCTGCTGCCAATAGC-3′. Each amplified fragment was subcloned into a TA-cloning vector, pGEM-T (Promega). The 5′-end of the c-mpl gene was digested with NotI and HindIII, the 3′-end was digested with HindIII and SalI and the fragments were ligated and inserted into pGEM3-SRα-neo at the NotI and SalI restriction sites respectively. The selected pGEM3-SRα-neo-c-mpl clone was verified by DNA sequencing.
Transfection and cloning of stable transfectants
Ba/F3 cells were washed twice and 107 cells were resuspended in 0.6 ml of PBS. The cells were mixed with 20 μg of the expression plasmid carrying human c-mpl cDNA, and then transfected by electroporation at 220 V and 960 μF. After 24 h incubation at 37°C in RPMI 1640 medium containing 10% (v/v) FBS and 10% (v/v) WEHI-3-conditioned medium, the cells were collected by centrifugation, resuspended in the same medium containing 2 mg/ml G418 as a selection marker and incubated for 10–14 days at 37°C. Subsequently, cloning was performed by dilution of transfected cells into 96-well plates. The cells were incubated for an additional 10 days at 37°C. Functional clones were selected by a cell proliferation assay as described below.
Cell proliferation assay
Cells were grown to exponential phase, collected by centrifugation and washed once in RPMI 1640 medium (without supplements) to completely remove IL-3, and resuspended in RPMI 1640 medium containing 10% (v/v) FBS (without IL-3) at a density of 1×105–5×105 cells/ml. A 100 μl aliquot of the resuspended culture was dispensed into each well of a 96-well tissue-culture plate containing either growth factors, one of 50000 compounds being tested (see the Results section for details) or DMSO as a vehicle at a variety of concentrations. After incubation for 20 or 68 h at 37°C, 10 μl of WST-1 reagent (TaKaRa) was added to each well and the plates were incubated for a further 4 h at 37°C. Finally, the absorbance at 450 nm was measured to detect cell viability. This high-throughput assay was carried out using a Biomek 2000 SL robotic system (Beckman Coulter). Two small-molecule compounds, termed compound A and compound B, induced cell proliferation and so a cytotoxicity test for these compounds using the mouse T-cell lymphoma cell line EL4 was carried out in the same way as the cell proliferation assay. In the time course experiment, the number of cells was counted under a microscope after they had been stained with Trypan Blue.
Preparation of whole-cell lysates for immunoprecipitation
Exponentially growing Ba/F3-huMPL cells were washed once with IL-3-free RMPI-1640 medium and resuspended in the same medium at a density of 107 cells/ml. After starvation for 4 h, the cells were incubated for 10 min at 37°C in the presence of 0.53 nM (25 ng/ml) recombinant human thrombopoietin, 0.2 μM (98 ng/ml) compound B or 0.05% DMSO as a vehicle. The cells were centrifuged, washed twice with ice-cold PBS and resuspended at a density of 108 cells/ml in ice-cold lysis buffer [50 mM Hepes (pH 7.5), 1% Triton X-100, 150 mM NaCl, 1 mM EGTA, 1 mM sodium pyrophosphate, 1 mM sodium fluoride, 1 mM sodium vanadate, 1 mM PMSF, 10 μg/ml leupeptin, 10 μg/ml aprotinin and 1 μM pepstatin A]. Cells were incubated on ice for 20 min, centrifuged at 10000 g for 10 min at 4°C in order to remove insoluble material and the clear lysates were stored at −80°C for immunoprecipitation.
Immunoprecipitation and immunoblotting
Whole-cell lysates (50 μl), equivalent to 5×106 cells, were transferred into a 1.5 ml microfuge tube and diluted with 950 μl of ice-cold lysis buffer (see above), 1 μg of the desired polyclonal or monoclonal antibody and 15 μg of Protein A–Sepharose (Pharmacia Biotech). After incubation at 4°C overnight and subsequent centrifugation, the immunoprecipitates were washed five times in ice-cold lysis buffer and boiled for 3 min in Laemmli sample buffer. Samples were separated by SDS/PAGE (10% gels) (Daiichi Pure Chemicals, Tokyo, Japan) and electrophoretically transferred on to a nitrocellulose membrane. The phosphorylation of signal transduction molecules was detected using a mouse monoclonal biotinylated anti-phosphotyrosine antibody (4G10, 1:1000 dilution), followed by the addition of an avidin-biotin horseradish peroxidase complex (Vecta Stain) and Super Signal chemiluminescence reagent (Pierce). To determine the identity of the protein that was tyrosine phosphorylated and to compare the amount of tyrosine-phosphorylated protein after treatment with compound B or thrombopoietin, each membrane was reprobed with a primary antibody against the protein of interest [anti-JAK1 antibody (1:1250 dilution), anti-JAK2 antibody (1:400 dilution), anti-JAK3 antibody (1:500 dilution), anti-Tyk2 antibody (1:1000 dilution), anti-STAT1 antibody (1:12500 dilution), anti-STAT3 antibody (1:3000 dilution), anti-STAT5 antibody (1:1250 dilution), anti-STAT6 antibody (1:1000 dilution), anti-SOS antibody (1:1000 dilution), anti-Grb2 antibody (1:1000 dilution), anti-Shc antibody (1:1000 dilution), anti-PLCγ antibody (0.2 μg/ml), anti-PI3K (p85) antibody (1:400 dilution), anti-Vav antibody (1:830 dilution), anti-Cbl antibody (1:500 dilution), anti-SHIP antibody (1:1000 dilution) and anti-SHP-2 antibody (1:1000 dilution)].
Measurement of p44/p42 MAPK activity
Activation of p44/p42 MAPK was evaluated by measuring the phosphorylation of Elk-1 using a p44/p42 MAPK Assay kit (New England Biolabs) following the manufacturer's instructions. Active p44/p42 MAPK was immunoprecipitated from whole-cell lysates with a mouse monoclonal anti-phospho-(p44/p42 MAPK) antibody. Next, an in vitro kinase assay was carried out by adding GST (glutathione transferase)-tagged Elk-1 protein as a substrate. Furthermore, Western blotting using a rabbit polyclonal anti-(phospho-Elk-1) antibody was performed to detect phosphorylation of Elk-1.
Measurement of SAPK (stress-activated protein kinase)/JNK (c-Jun N-terminal kinase) activity
The activity of SAPK/JNK was measured by detection of c-Jun phosphorylation using a SAPK/JNK Assay kit (New England Biolabs) following the manufacturer's instructions. SAPK/JNK was pulled down from cell lysates using a GST–c-Jun fusion protein bound to glutathione–Agarose beads. The beads were washed with a reaction buffer and the in vitro kinase assay was performed using the beads as a substrate. A rabbit polyclonal anti-(phospho-c-Jun) antibody was used to detect c-Jun phosphorylation by Western blotting.
Measurement of p38 MAPK activity
The activity of p38 MAPK was assessed by measurement of ATF-2 (activating transcription factor 2) phosphorylation using a p38 MAPK Assay kit (New England Biolabs) following the manufacturer's instructions. First, activated p38 MAPK was immunoprecipitated from cell lysates with a mouse monoclonal anti-(phospho-p38 MAPK) antibody. Subsequently, an in vitro kinase assay was performed using GST–ATF-2 protein as a substrate. ATF-2 phosphorylation was detected by Western blotting using a rabbit polyclonal anti-(phospho-ATF-2) antibody.
Megakaryocytopoiesis from human CD34+ bone-marrow cells
Megakaryocyte colony formation assays were performed using a MegaCult™-C kit (StemCell Technologies) following the manufacturer's instructions. Briefly, CD34+ cells from human bone marrow were washed and resuspended at a concentration of 3.3×103 cells/ml in IMDM (Iscove's modified Dulbecco's medium) supplemented with 10 μg/ml recombinant human insulin, 200 μg/ml human transferrin (iron-saturated), 2 mM L-glutamine, 0.1 mM 2-mercaptoethanol and 1.1 mg/ml collagen. An aliquot (0.75 ml) was dispensed into each well of a double-chamber slide (StemCell Technologies) in the presence of various concentrations of recombinant human thrombopoietin, compound B or vehicle (0.1% DMSO). Cells were incubated at 37°C in 95% air/5% CO2 for 12 days. Cells were dehydrated, fixed and blocked with 5% (v/v) human serum in Tris-buffered saline before incubation with 10 μg/ml mouse monoclonal anti-[human GPIIb/IIIa (CD41a)]. Next, biotin-conjugated goat anti-mouse IgG antibody, avidin-conjugated-alkaline phosphatase and alkaline phosphatase substrate were applied in sequence. Evans Blue counterstaining was performed for each slide. The number of megakaryocyte colonies was counted under a stereoscopic microscope (OPTIHOT-2; Nikon, Japan). Results are means±S.D. for four determinations.
Identification of thrombopoietin-mimetic compounds
We screened approx. 50000 chemical compounds for molecules that stimulated proliferation of the thrombopoietin-dependent cell line Ba/F3-huMPL, and identified compound A (Figure 1). A structure–activity relationship study for compound A identified a derivative (compound B) with greater efficacy and less toxicity than compound A. The proliferative activity of compounds A and B, compared with recombinant human thrombopoietin, is shown in Figure 2(A). Compound A resulted in a ‘bell-shaped’ growth curve, indicating that its cytotoxicity, rather than its proliferation activity, had an effect on Ba/F3-huMPL cells beyond the concentration of 3 μM. The maximal response of Ba/F3-huMPL cells after stimulation with compound A did not even reach 25% of the maximal proliferative activity exhibited by recombinant human thrombopoietin. However, the maximal response of Ba/F3-huMPL cells when exposed to compound B (70 nM) was equivalent to that of recombinant human thrombopoietin in this proliferation assay. The cytotoxicities of compound A and compound B against the mouse T-cell lymphoma cell line EL4 were measured (Figure 2B). Although the LC50 (half lethal concentration) value of compound A was determined by this test to be 5.5 μM, compound B did not show severe cytotoxicity against the EL4 cell line at concentrations below 50 μM.
Chemical structures of thrombopoietin-mimetic compounds
Compounds A and B are novel activators of the thrombopoietin-dependent cell line Ba/F3-huMPL
As shown in Figure 2(C), neither compound A nor compound B stimulated the proliferation of parental IL-3-dependent Ba/F3 cells that do not express the thrombopoietin receptor. Furthermore, compounds A and B did not promote the growth of the GM-CSF- or erythropoietin-dependent cell line TF-1 (results not shown).
We next examined the time course of Ba/F3-huMPL cell proliferation in the presence of 2.3 μM compound A (Figure 2D). On day 1, no difference was observed in cell number between cells stimulated with compound A and with vehicle; however, compound A stimulated the growth of Ba/F3-huMPL cells continuously from day 2, such that the cell number had increased 5-fold after four days. The vehicle had no effect on the proliferation of the Ba/F3-huMPL cell line, and all cells had died by day 7. Compound A maintained the survival of Ba/F3-huMPL cells for more than 2 weeks (results not shown).
Activation of the human thrombopoietin receptor by stimulation with compound B
When bound to thrombopoietin, the thrombopoietin receptor undergoes rapid tyrosine phosphorylation . We investigated whether or not compound B could cause tyrosine phosphorylation of the thrombopoietin receptor. Ba/F3-huMPL cells were stimulated for 10 min with 25 ng/ml recombinant human thrombopoietin, 0.2 μM compound B or 0.05% DMSO (vehicle) and whole-cell lysates were immunoprecipitated with an antibody against the thrombopoietin receptor. Western blotting with an anti-phosphotyrosine antibody was then performed. Stimulation with compound B led to tyrosine phosphorylation of the thrombopoietin receptor, although the intensity of phosphorylation was weaker than that observed for recombinant human thrombopoietin (Figure 3).
Tyrosine phosphorylation of the human thrombopoietin receptor in response to recombinant human thrombopoietin or compound B
Activation of thrombopoietin signalling molecules by stimulation with compound B
Stimulation by thrombopoietin triggers several signal-transduction cascades, such as tyrosine phosphorylation of JAK2, STAT3 and Shc in Ba/F3-huMPL cells. To verify that the signalling molecules activated by compound B are the same as those stimulated by thrombopoietin, we prepared Ba/F3-huMPL whole-cell lysates under the conditions described above. First, immunoprecipitation of cell lysates with antibodies against JAK/STAT proteins and subsequent Western blot analysis with an anti-phosphotyrosine antibody was performed. Figure 4 shows that tyrosine phophorylation of JAK2, Tyk2, STAT3 and STAT5 was elicited in Ba/F3-huMPL cells by stimulation with either recombinant human thrombopoietin or compound B. However, recombinant human thrombopoietin activated these molecules slightly more strongly than compound B. There was no activation of these proteins in the unstimulated whole-cell lysate (vehicle).
Tyrosine phosphorylation of JAK/STAT proteins in response to recombinant human thrombopoietin or compound B
The MAPK signal-transduction cascade in Ba/F3-huMPL cells was examined to analyse the activities of p44/p42 MAPK, SAPK/JNK and p38 MAPK. As shown in Figure 5, p44/p42 MAPK, SAPK/JNK and p38 MAPK were activated by stimulation of cells with recombinant human thrombopoietin or compound B. Compound B activated SAPK/JNK and p38 MAPK to almost the same degree as recombinant human thrombopoietin, whereas the activation of p44/p42 MAPK by compound B was weaker than that by recombinant human thrombopoietin. Signalling molecules that are regarded as adaptor proteins in the Ras-MAPK pathway, Grb2 and Shc, were tyrosine phosphorylated in cells stimulated with compound B as well as when cells were stimulated with recombinant human thrombopoietin, but SOS was not tyrosine phosphorylated in either case.
Activation of MAPK family proteins and adaptor proteins related to the Ras-MAPK pathway in response to recombinant human thrombopoietin or compound B
The effects of recombinant human thrombopoietin or compound B on PLCγ or PI3K were assessed by detection of tyrosine phosphorylation. As shown in Figure 6, compound B elicited tyrosine phosphorylation of PLCγ, although to a lesser extent than that induced by recombinant human thrombopoietin. Tyrosine phosphorylation of p85 was not induced by recombinant human thrombopoietin or by compound B.
Tyrosine phosphorylation of PLCγ, PI3K and related proteins in response to stimulation with recombinant human thrombopoietin or compound B
We also evaluated the phosphorylation of regulatory proteins by immunoprecipitation and Western blot analysis with anti-phosphotyrosine antibodies. Figure 6 shows that tyrosine phosphorylation of Vav, Cbl and SHP-2, but not SHIP, was induced by compound B as well as by recombinant human thrombopoietin.
These results suggest that recombinant human thrombopoietin and compound B induce cell growth through the same signal-transduction pathway.
Compound B stimulation of human megakaryocytopoiesis
To examine whether compound B could promote proliferation and maturation of megakaryocyte-lineage cells, a megakaryocyte colony formation assay was performed using human CD34+ haematopoietic progenitor cells, and colonies were verified to be of the megakaryocyte lineage by immunocytostaining for the CD41 antigen, a specific marker of this lineage. Treatment with 1 nM recombinant human thrombopoietin for 12 days increased the number of CD41+ colonies in a semi-solid culture of human CD34+ bone-marrow cells (Figure 7). Compound B increased the number of CD41+ colonies in a dose-dependent manner, similar to recombinant human thrombopoietin-induced megakaryocytopoiesis.
Effects of compound B on the differentiation of human CD34+ bone-marrow cells into the megakaryocytic lineage
Thrombopoietin is the most important cytokine regulating the proliferation and maturation of megakaryocytes and formation of platelets . Clinical studies have demonstrated that two recombinant forms of human thrombopoietin [full-length recombinant human thrombopoietin and PEG-rhMGDF (PEGylated recombinant human megakaryocyte growth and development factor) might be effective in reducing thrombocytopaenia caused by myelosuppressive chemotherapy [7–9]. However, administration of either form of recombinant human thrombopoietin had led to immunogenicity in some cases and the risk of developing neutralizing antibodies that cross-react with endogenous thrombopoietin, which could lead to secondary thrombocytopaenia [10–12]. Artificial ligand proteins of the thrombopoietin receptor that share no sequence homology with thrombopoietin have been suggested as a therapeutic approach to avoid this problem [13–16]. Peptide agonists of the thrombopoietin receptor have been reported to activate the thrombopoietin receptor, regardless of the fact that they are far smaller than native thrombopoietin [17,18]. Several non-peptide small molecules have been identified previously as functional thrombopoietin receptor agonists [19–25].
We screened a compound library for molecules with anti-thrombocytopaenia activity, and found a series of small molecular-mass non-peptide thrombopoietin-mimetic compounds. Compound A and its chemically modified derivative, compound B, promoted proliferation of Ba/F3-huMPL cells in a dose-dependent manner. Compound A allowed proliferation of Ba/F3-huMPL cells to continue for more than 2 weeks in the absence of other growth factors. However, proliferation of the parent Ba/F3 cells, which do not express the thrombopoietin receptor, was not stimulated by these compounds, even though they express the IL-3 receptor. It is possible that compound A induces the production of endogenous thrombopoietin in Ba/F3-huMPL cells and promotes autocrinal cell growth, therefore we measured the concentration of thrombopoietin in the supernatant of Ba/F3-huMPL cells grown in the presence of compound A. However, we did not detect any thrombopoietin (results not shown), indicating that these compounds do not promote the production of thrombopoietin, but stimulate the proliferation of Ba/F3-huMPL cells through activating the thrombopoietin receptor directly. What is especially important is that neither parent IL-3-dependent Ba/F3 cells nor GM-CSF- or erythropoietin-dependent TF-1 cells were affected by these compounds, meaning they were highly specific to the thrombopoietin receptor.
Chemical modification of compound A led to great success in generating more active and less toxic derivatives. Table 1 summarizes the activity and cytotoxicity of compounds A and B. Since compound A did not promote the proliferation of Ba/F3-huMPL cells to levels that were 50% of the maximal proliferation induced by recombinant human thrombopoietin, the EC20 (concentration required for 20% of the maximal response exhibited by recombinant human thrombopoietin) was expediently used as an additional parameter. The EC20 of compound A was approx. 1.5 μM and the EC50 and EC20 of compound B in Ba/F3-huMPL cells were calculated to be 29 nM and 9 nM respectively. The EC20 of compound A was more than 150 times that of compound B. The LC50 of compound A on the mouse T-cell lymphoma EL4 cell line was 5.5 μM in this proliferation test, whereas that of compound B was more than 50 μM. The separation of proliferative activity from cytotoxicity made it possible to elucidate the signal-transduction pathways involved in proliferation and evaluate the potency of in vitro megakaryocytopoiesis to further advance the structure–activity relationship study of thrombopoietin-mimetic compounds.
|.||Ba/F3-huMPL .||EL4 mouse T-cell lymphoma .|
|Compound .||EC50 (nM) .||EC20 (nM) .||Emax (%) .||LC50(μM) .|
|.||Ba/F3-huMPL .||EL4 mouse T-cell lymphoma .|
|Compound .||EC50 (nM) .||EC20 (nM) .||Emax (%) .||LC50(μM) .|
With regards to signal transduction, compound B clearly activated the classical thrombopoietin signal-transduction pathway in Ba/F3-huMPL cells, as the thrombopoietin receptor, JAK-STAT, MAPK family, PLCγ and other related proteins were all activated, as summarized in Table 2. These results suggest that compound B acts via the thrombopoietin receptor and activates the same intracellular signal-transduction responses as recombinant human thrombopoietin.
|Protein .||Vehicle (DMSO) .||Compound B .||Thrombopoietin .|
|Protein .||Vehicle (DMSO) .||Compound B .||Thrombopoietin .|
Compound B did not activate molecules such as the thrombopoietin receptor or STAT3 to same extent as their activation by recombinant human thrombopoietin. The concentration of compound B or recombinant human thrombopoietin required for stimulation was determined as the minimal concentration required to induce maximal proliferation of Ba/F3-huMPL cells. Sufficient activation of some molecules might not be essential for Ba/F3-huMPL proliferation. Actually, transformants that express deletion mutants of the thrombopoietin receptor, in which tyrosine phosphorylation of signalling molecules such as STAT3 or Shc is not sufficiently elicited by thrombopoietin, retain the ability to proliferate when stimulated by thrombopoietin [26,27]. These signalling molecules would otherwise have more of a role in differentiation rather than proliferation in megakaryocytes. We have not investigated all of the possible signalling mechanisms involved in thrombopoietin-mediated activation, such as PI3K activity or serine/threonine-phosphorylation of the thrombopoietin receptor  and other regulatory proteins. They may play important roles in transducing the proliferation and maturation signals and need to be studied further.
A comparison of the signal transduction profile of compound B with that of recombinant human thrombopoietin should be helpful in predicting its biological effects and assessing its relevance as a lead compound for an anti-thrombocytopaenia drug. The complete overlap in phosphorylation profile of signalling molecules induced by compound B and recombinant human thrombopoietin in Ba/F-huMPL cells suggested that compound B might induce megakaryocytopoiesis in vitro.
We performed a colony-formation assay and found that compound B stimulated human CD34+ cells and increased the number of megakaryocytic colonies in the absence of other growth factors. It is not clear why the efficacy of 5 μM compound B in the megakaryocyte colony formation assay was lower than that of 1 nM thrombopoietin, whereas 1 μM compound B was equivalent to 1 nM thrombopoietin in the proliferative assay using Ba/F3-huMPL cells. One possible explanation for the weaker activity for megakaryocytopoiesis may be that some signalling molecules are not as well activated by compound B, as described above. Another possibility is that compound B binds to the thrombopoietin receptor at a different site from thrombopoietin. Recently, the existence of a motif that prevents autonomous activation of the thrombopoietin receptor at the transmembrane-cytoplasmic junction has been reported . We need to clarify how our compounds bind to the thrombopoietin receptor.
Several thrombopoietin-mimetic compounds are reported to exhibit not only in vitro but also in vivo activity in various animal models [22,24]. SB-497115, a non-peptide thrombopoietin-mimetic compound was shown to increase the number of platelets in a clinical trial . It was necessary to confirm the in vivo activity of compound B or its derivatives but this was not possible in the majority of common experimental animals and we speculated that this was the result of species specificity . However, we recently succeeded in evaluating the in vivo activity of these compounds by transplanting human precursor cells into mice .
In conclusion, we have discovered a series of novel non-peptide thrombopoietin mimetic small molecules which stimulate the proliferation and maturation of megakaryocytes from human CD34+ cells by directly activating the thrombopoietin receptor.
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
activating transcription factor 2
concentration required for 20% of the maximal response exhibited by recombinant human thrombopoietin
fetal bovine serum
granulocyte/macrophage colony-stimulating factor
growth-factor-receptor-bound protein 2
c-Jun N-terminal kinase
half lethal concentration
mitogen-activated protein kinase
85 kDa subunit of PI3K
stress-activated protein kinase
Src homology and collagen homology
Src homology 2 domain-containing protein tyrosine phosphatase 2
Src homology 2 domain-containing inositol phosphatase
son of sevenless
signal transducer and activator of transcription
tyrosine kinase 2
These authors contributed equally to this work.