Section of Orthopedic and Spinal Surgery, Center of Excellence Program for Frontier Research on Molecular Destruction and Reconstruction of Tooth and Bone, Tokyo Medical and Dental University Graduate School, Bunkyo-ku, Tokyo, Japan
To clarify the role of ghrelin in bone metabolism, we examined the effect of ghrelin in vitro and in vivo. Ghrelin and its receptor, GHS-R1a, were identified in osteoblasts, and ghrelin promoted both proliferation and differentiation. Furthermore, ghrelin increased BMD in rats. Our results show that ghrelin directly affects bone formation.
Introduction: Ghrelin is a gut peptide involved in growth hormone (GH) secretion and energy homeostasis. Recently, it has been reported that the adipocyte-derived hormone leptin, which also regulates energy homeostasis and opposes ghrelin's actions in energy homeostasis, plays a significant role in bone metabolism. This evidence implies that ghrelin may modulate bone metabolism; however, it has not been clarified. To study the role of ghrelin in skeletal integrity, we examined its effects on bone metabolism both in vitro and in vivo.
Materials and Methods: We measured the expression of ghrelin and growth hormone secretagogue receptor 1a (GHS-R1a) in rat osteoblasts using RT-PCR and immunohistochemistry (IHC). The effect of ghrelin on primary osteoblast-like cell proliferation was examined by recording changes in cell number and the level of DNA synthesis. Osteoblast differentiation markers (Runx2, collagen α1 type I [COLI], alkaline phosphatase [ALP], osteocalcin [OCN]) were analyzed using quantitative RT-PCR. We also examined calcium accumulation and ALP activity in osteoblast-like cells induced by ghrelin. Finally, to address the in vivo effects of ghrelin on bone metabolism, we examined the BMD of Sprague-Dawley (SD) rats and genetically GH-deficient, spontaneous dwarf rats (SDR).
Results: Ghrelin and GHS-R1a were identified in osteoblast-like cells. Ghrelin significantly increased osteoblast-like cell numbers and DNA synthesis in a dose-dependent manner. The proliferative effects of ghrelin were suppressed by [D-Lys3]-GHRP-6, an antagonist of GHS-R1a, in a dose-dependent manner. Furthermore, ghrelin increased the expression of osteoblast differentiation markers, ALP activity, and calcium accumulation in the matrix. Finally, ghrelin definitely increased BMD of both SD rats and SDRs.
Conclusions: These observations show that ghrelin directly stimulates bone formation.
GHRELIN IS A 28 amino acid peptide that was initially isolated from rat stomach as a growth hormone secretagogue peptide.(1,2) Ghrelin and the synthetic growth hormone secretagogues, such as GHRP-6, hexarelin, and MK-0677,(3,4) activate G-protein-coupled receptors (GPCRs) and stimulate growth hormone (GH) secretion from pituitary somatotropes.(1) The ghrelin receptor, known as the growth hormone secretagogue receptor (GHS-R), belongs to the GPCR family, and has two subtypes produced by alternative splicing, which are the fully functional type 1a receptor (GHS-R1a) and the biologically inactive GHS-R type 1b (GHS-R1b).(5–7) GHS-R1a mRNA is found in a variety of organs, including the stomach, heart, lung, pancreas, intestine, kidney, testis, and ovary, as well as in the hypothalamus region and in adipose tissue.(1,8–13) The wide distribution of this receptor indicates that ghrelin, which is produced in and secreted mainly from the stomach, may have a variety of regulatory functions both in the brain and peripheral tissues. In fact, emerging evidence indicates that ghrelin performs an array of additional biological actions: it stimulates appetite, promotes adipogenesis, decreases energy metabolism, improves cardiovascular function, and stimulates prolactin and cortisol releases.(13–19) These findings, together with the evidence that GH is well known to promote bone formation,(20–22) suggest that ghrelin may play a role in bone metabolism; however, this has not been comprehensively studied, and the role of ghrelin in bone metabolism remains unknown.
Recently, leptin, another energy-regulating hormone, was shown to have a role in bone metabolism.(23–29) Physiologically, ghrelin's actions oppose those of leptin; ghrelin stimulates food intake and suppresses energy expenditure, whereas leptin suppresses food intake and increases energy expenditure.(30) Moreover, in the hypothalamus, ghrelin activates neuropeptide Y (NPY)/agouti-related peptide (AgRP) neurons to stimulate appetite, whereas leptin suppresses these neurons.(31) These evidence also suggest that ghrelin may have a role in bone formation.
Here, we report the expression of GHS-R1a in rat osteoblasts and show that ghrelin stimulates cell proliferation and differentiation. In addition, we show that ghrelin increases BMD both in normal and GH-deficient rats. Based on these observations, we conclude that ghrelin directly promotes bone formation mediated by GHS-R1a in vitro and in vivo. This mechanism of action for ghrelin provides a new insight in our understanding of bone formation.
MATERIALS AND METHODS
Sprague-Dawley (SD) rats (Charles River Co., Yokohama, Japan) were used for osteoblast-like cell culture experiments and in vivo studies. Spontaneous dwarf rats (SDRs), a specific model of severe isolated GH deficiency(32) (kindly provided by Prof Ishikawa), were used for in vivo studies. The rats were housed in a regulated environment (22 ± 2°C, 55 ± 10% humidity, 12-h light, 12-h dark cycle with light on at 7:00 a.m.) with free access to food and water. All experiments were conducted in accordance with the Japanese Physiological Society's guidelines for animal care.
α-MEM (containing L-glutamine and nucleosides), DMEM, penicillin/streptomycin (10,000 IU/ml and 10,000 μg/ml), and trypsin/EDTA were purchased from Life Technologies-GIBCO (Cergy Pontoise, France). FBS was purchased from Thermo Trace (lot B10152-500; Melbourne, Australia). BSA (A-7888) was purchased from Sigma Chemical (St Louis, MO, USA). Fungizone was acquired from Life Technologies (Rockville, MD, USA). Collagenase, ascorbic acid, and β-glycerophosphate were of reagent grade and were purchased from Sigma Chemical. Trizol was purchased from Invitrogen (Carlsbad, CA, USA). Superscript II reverse transcriptase, 5× first stand buffer, and oligo(dT)12-18 were purchased from Life Technologies-GIBCO. Taq DNA polymerase and dNTP mix were obtained from TAKARA BIO (Shiga, Japan). A QIAquick gel extraction kit and SYBR Green were purchased from QIAGEN (Hilden, Germany). Rat ghrelin and GHRP-6 were kindly provided by Prof Kangawa. [D-Lys3]-GHRP-6 was obtained from WAKO Pure Chemical Industry (Osaka, Japan). Alizarin red S was obtained from Sigma Chemical. The cell count reagent SF was purchased from Nacalai Tesque (Kyoto, Japan). A BrdUrd cell proliferation kit was purchased from Roche (Mannheim, Germany). An alkaline phosphatase (ALP) kit was obtained from WAKO Pure Chemical Industry, and a protein assay kit was purchased from BIO-RAD Laboratories (Hercules, CA, USA). Polyclonal anti-ghrelin and anti-GHS-R1a antibodies were purchased from Phoenix Pharmaceuticals (Belmont, CA, USA). Alexa-Fluor 488 goat anti-rabbit antibody was purchased from Molecular Probes (Eugene, OR, USA). Eliet ABC kit for immunohistochemistry (IHC) was obtained from Vector Laboratories (Burlingame, CA, USA).
Osteoblast-like cell culture
Primary osteoblast-like cells were isolated by collagenase digestion from 21-day fetal rat calvaria, as previously described.(33,34) Digests 3-5 were pooled and grown in 10-cm cell culture plates in primary culture media consisting of α-MEM supplemented with 10% FBS and antibiotics, including 100 μg/ml penicillin G, 50 μg/ml streptomycin sulfate, and 0.3 μg/ml Fungizone. Cells were grown to confluence before being subjected to experimentation.
UMR106 cells, a rat osteoblastic cell line, were obtained from Dainippon Pharmaceutical Co. (Osaka, Japan). Cells were plated in 10-cm plates at a density of 2 × 105 cells/plate and maintained in DMEM supplemented with 10% FBS and antibiotics.
All cultures were incubated at 37°C in a humidified atmosphere of 95% air and 5% CO2.
Expression of ghrelin and GHS-R1a
Ghrelin and GHS-R1a were identified by RT-PCR and IHC.
RNA extraction and RT-PCR were performed as follows. Total RNA was extracted from cell pellets using Trizol according to the manufacturer's instructions. cDNA was synthesized from 2 μg of total RNA using the Super Script Preamplification System for First-Strand cDNA Synthesis Kit. Primer sequences were as follows: rat ghrelin (254-bp product; GenBank accession no. AB029433) sense, 5′-CCAGAGGACAGAGGACAAGC-3′, and antisense, 5′-AGTTGCAGAGGAGGCAGAAGCT-3′; and rat GHS-R1a (314-bp product; GenBank accession no. U94321) sense, 5′-GAGATCGCTCAGATCAGCCAG-3′, and antisense, 5′-AGAACCTCAGTTTGGGGATTA-3′. The PCR conditions used were as follows: denaturation at 94°C for 30 s, annealing at 60°C for 30 s, and extension at 72°C for 1 minute, 35 cycles. The PCR products were purified from agarose gels using QIAquick gel extraction kits, and their sequences were determined using an ABI PISM 310 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).
In vitro IHC was performed as follows. Cells were grown to 70% confluence in 6-well plates (NUNC), washed in PBS, and fixed with 4% paraformaldehyde for 15 minutes. The cells were permeabilized by incubation with 0.1% Triton-X in PBS for 5 minutes. Blocking of nonspecific antibody binding was performed by incubating with 10% normal goat serum for 1 h. The cells were incubated with the polyclonal rabbit anti-rat ghrelin antibody (1:2000 dilution) or the rabbit anti-human GHS-R1a antibody (1:2000 dilution) in PBS.(11,12) After a 1-h incubation with the primary antibody, the cell layer was washed with PBS and incubated with the Alexa Fluor 488 goat anti-rabbit antibody (1:200 dilution) for 30 minutes. All incubations described above were performed at room temperature (22°C). Immunoreactivity was measured using a laser scanning confocal microscope (Micro Radiance; BIO-RAD Laboratories).
For in vivo IHC, 3-day-old male SD rats were used (n = 3). Bones were fixed using 5% paraformaldehyde and embedded in paraffin. Paraffin-embedded femoral sections (40 μm thick) were incubated for 1 day with rabbit anti-rat ghrelin antibody (1:50,000 dilution) or rabbit anti-human GHS-R1a antibody (1:2000 dilution). We stained the sections by the avidin-biotin complex method using Eliet ABC kits as described previously.(35)
Cell proliferation assays
Primary osteoblast-like cells were seeded onto 96-well plates at a density of 6000 cells/well. Twenty-four hours after the creation of the subcultures, the cells were changed to serum-free medium with 1% BSA for a further 24 h before the addition of the experimental compounds. Cells were stimulated with ghrelin or growth hormone-releasing peptide-6 (GHRP-6), a synthetic agonist for the GHS-R1a, at concentrations between 10−11 and 10−8 M, or with the GHS-R1a antagonist, [D-Lys3]-GHRP-6, at concentrations between 10−10 and 10−6 M, before stimulation with 10−8 M ghrelin.
The relative number of viable cells in each well was determined 24 h after addition of compounds using the cell count reagent SF, which is 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt (WST-8). Briefly, 10 μl of WST-8 solution was added to each well, including three wells containing medium alone, to be used background subtraction. The cells were incubated at 37°C for 1 h. The absorbance at 450 nm in each well was determined using a Multiskan JX (Thermo Labsystems, Helsinki, Finland). This technique produces a linear relationship between the number of viable cells and the absorbance at 450 nm.
A BrdUrd incorporation assay was performed using a colorimetric BrdUrd cell proliferation kit according to the manufacturer's instructions. For the last 2 h of the 24-h stimulation period, the cells were pulsed with BrdU. Absorbance at 450 nm was measured with a microplate reader. FCS was reduced to 1% for all treatment conditions.
Calcified nodule formation and ALP activity
Calcified nodule formation was induced as previously described.(33) Briefly, osteoblast-like cells were plated in 6-well plates at a density of 4 × 104 cells/well. Cells were grown until confluent, which was designated as day 0, and nodule formation was induced by the addition of 50 μg/ml ascorbic acid and 10 mM β-glycerophosphate to the medium. Cells were incubated for 28 days with or without ghrelin (10−9 and 10−8 M). Media was changed every 3 days. Cells were rinsed three times with PBS and fixed with 10% paraformaldehyde for 10 minutes. Subsequently, calcified nodules were stained with 1% Alizarin red S at pH 6.4 for 2 minutes and washed with distilled water.(36) The area of the nodules was analyzed using NIH Image 1.63 software.
The time course of ALP activity was analyzed after 0, 3, 5, and 7 days of culture (with or without 10−8 M of ghrelin). Dose-dependent effects of ghrelin were analyzed after a 5-day incubation with or without ghrelin (10−9 or 10−8 M). ALP activity was analyzed histochemically using an ALP kit. In brief, assay mixtures containing 0.1 M 2-amino-2-methyl-1-propanol, 1 mM MgCl2, 8 mM p-nitrophenol phosphate disodium, and cell homogenate were incubated for 5 minutes at 37°C, at which point the reaction was stopped with 0.1 N NaOH, and the absorbance at 405 nm was measured. A standard curve was prepared with p-nitrophenol. Each value was normalized to the protein concentration. Total cellular protein in the cell layer was measured by the Lowry method using a protein assay kit and BSA as a standard.
Assessment of expression of osteoblast differentiation markers
Cells in 6-well plates treated with or without 10−8 M of ghrelin were analyzed after 0, 3, 5, 7, 14, and 21 days of in vitro culture. Cells were plated at a density of 4 × 104 cells/well and were grown until confluent, which was designated as day 0. Cells were grown in primary culture media with 50 μg/ml ascorbic acid and 10 mM β-glycerophosphate. cDNA obtained as described above was subjected to quantitative real-time PCR analysis using the ABI PRISM 7700 sequence detection system (Applied Biosystems). Specific primers yielding short PCR products suitable for SYBR-green detection were designed using Primer Express software (version 1.0; PE Applied Biosystems). Primer sequences were as follows: Runx2 (67-bp product; GenBank accession no. AF053953), 5′-GCTTCATTCGCCTCACAAACA-3′ (sense) and 5′-TGCTGTCCTCCTGGAGAAAGTT-3′ (antisense); collagen α1 type I (COLI; 65-bp product; GenBank accession no. Z78279), 5′-TTCACCTACAGCACGCTTGTG-3′ (sense) and 5′-GATGACTGTCTTGCCCCAAGTT-3′ (antisense); ALP (101-bp product; GenBank accession no. J03572), 5′-CGTCTCCATGGTGGATTATGC-3′ (sense) and 5′-TGGCAAAGACCGCCACAT (antisense); osteocalcin (OCN; 63-bp product; GenBank accession no. X04141), 5′-GAGCTAGCGGACCACATTGG-3′ (sense) and 5′-CCTAAACGGTGGTGCCATAGA-3′ (antisense); and β-actin (67-bp product; GenBank accession no. NM031144), 5′-TTCAACACCCCAGCCATGT-3′ (sense) and 5′-GTGGTACGACCAGAGGCATACA-3′ (antisense). Samples were examined in triplicate. The reaction volume was 50 μl, and samples were subjected to 45 cycles of amplification at 95°C for 15 s, followed by 52°C for 60 s using 3 μl diluted cDNA (1:30), 10 μl SYBR-green buffer, and 10 pmol of each primer. The concentration of each amplified cDNA in each sample tested was calculated relative to that of β-actin cDNA. After RT-PCR amplification, a dissociation analysis was performed on the products to ensure that only one product was produced in each PCR reaction. Products were also run on a 2% agarose gel to check for single, correctly sized products.
Effect of ghrelin on BMD in vivo
Six-week-old male SD rats and SDRs were used for in vivo studies. Rats were infused intraperitoneally with rat ghrelin (SD; n = 5, SDR; n = 3) or saline (SD; n = 5, SDR; n = 5) using osmotic minipumps (Alzet 2004; Alza Corp., Palo Alto, CA, USA). So that there was no significant difference in body weight and food intake in rats with or without ghrelin, we infused the appropriate ghrelin concentration (50 μg/kg/day).(37) We measured the body weight and food intake of rats once a week. After 4 weeks of treatment, the BMD of the femur was measured by DXA (model DCS-600; Aloka, Tokyo, Japan).
All experiments were repeated three or four times. Data are presented as the mean ± SD. The statistical significance of the difference in mean values was assessed by a two-factor ANOVA. Statistical significance was assessed as p < 0.05.
Expression of ghrelin and GHS-R1a
First, we examined the expression of ghrelin and GHS-R1a in osteoblast-like cells. Ghrelin and GHS-R1a transcripts corresponding to predicted sizes of 254 and 314 bp, respectively, were found in osteoblast-like cells (Fig. 1A). The identity of these PCR products was also confirmed by direct sequencing (data not shown).
Next, the pattern of cellular distribution of ghrelin and GHS-R1a protein in primary osteoblast-like cells was analyzed by IHC, using well-characterized specific anti-ghrelin and anti-GHS-R1a polyclonal antibodies. Positive IHC staining for ghrelin and GHS-R1a was detected in osteoblast-like cells, providing evidence that these cells synthesize ghrelin and GHS-R1a protein (Fig. 1B). Absorption tests for ghrelin and GHS-R1a confirmed the specificity of these results (data not shown).
We also examined the in vivo IHC localization of ghrelin and GHS-R1a and observed positive IHC staining for ghrelin and GHS-R1a in osteoblasts (Fig. 1C).
Effect of ghrelin on osteoblast proliferation
To assess the effect of ghrelin on primary osteoblast-like cells, we measured viable cell number and DNA synthesis in response to ghrelin treatment. Ghrelin increased cell proliferation in a time-dependent manner (Fig. 2A). Measurement of WST-8 in cultured primary osteoblast-like cells revealed that ghrelin administration induced a dose-dependent, significant increase in viable cell number up to 1.33-fold compared with vehicle (Fig. 2B). Ghrelin also induced a dose-dependent increase in DNA synthesis, as measured by BrdUrd incorporation into cells (Fig. 2C). Furthermore, treatment with GHRP-6 also showed dose-dependent proliferative effects similar to those seen with ghrelin (Fig. 2D). To verify whether ghrelin promotes cell proliferation through binding and activation of GHS-R1a, we examined the effect of a GHS-R1a antagonist, [D-Lys3]-GHRP-6. The proliferative effects of ghrelin were completely abolished by co-treatment with [D-Lys3]-GHRP-6 in a dose-dependent manner (Fig. 2E). This result indicates that ghrelin directly stimulates osteoblast-like cell proliferation through binding to GHS-R1a.
Effect of ghrelin on osteoblast differentiation
To examine further the effect of ghrelin on osteoblast differentiation, we analyzed the gene expression of several markers of osteoblast differentiation, including Runx2, COLI, ALP, and OCN by quantitative RT-PCR. Ghrelin did not affect the expression of Runx2, a transcriptional factor necessary for osteoblast differentiation (Fig. 3A), but did significantly increase the expression of two early osteoblast differentiation markers, COLI and ALP, at day 5, by up to 1.95- and 2.42-fold (p < 0.01), respectively, compared with vehicle-treated cultured cells (Figs. 3B and 3C). Furthermore, the expression of OCN, a marker for the late stage of differentiation, began to increase at day 7 by up to 2.14-fold (p < 0.01; Fig. 3D).
We next evaluated the ALP activity of osteoblast-like cells. Ghrelin increased the ALP activity in a time-dependent manner (Fig. 4Aa). As shown in Fig. 4Ab, ghrelin significantly increased the ALP activity of osteoblast-like cells by up to 1.65-fold compared with vehicle-treated cells after 5 days in culture. These results indicate that ghrelin also stimulates osteoblast differentiation.
To elucidate further the role of ghrelin in osteoblast differentiation, we studied calcium accumulated by osteoblast-like cells. As expected, ghrelin treatment increased the calcified area by up to 1.8-fold compared with the controls (Fig. 4B).
Effects of ghrelin on BMD in vivo
To explore the in vivo effects of ghrelin in bone, we measured BMD by DXA in normal SD rats, as well as SDRs, which are deficient in GH. Six-week-old male rats were infused intraperitoneally with ghrelin for 4 weeks. Compared with vehicle, ghrelin significantly increased BMD both in SD rats and SDRs (Fig. 5).
In this study, we identified ghrelin and GHS-R1a in osteoblasts and showed that ghrelin directly promotes osteoblast proliferation and differentiation in vitro and increases BMD in vivo.
First, ghrelin and GHS-R1a were detected in rat osteoblasts by RT-PCR and IHC. This finding implies that osteoblasts respond to ghrelin signaling, and that ghrelin itself may have a physiological function in these cells. Indeed, ghrelin significantly stimulated osteoblast proliferation and differentiation. Furthermore, we showed that ghrelin increased ALP activity and calcified accumulation. These data confirm that ghrelin stimulates osteoblast differentiation.
In this study, we also found that ghrelin enhanced the transcription of genes encoding the osteoblast differentiation markers COLI, ALP, and OCN. However, we showed that ghrelin did not change the transcription marker Runx2. Each of these markers signifies a distinct mechanism of inducing differentiation. BMP2 and TGF-β are representative bone-forming factors that activate the transcription factors Smad and Runx2, leading to osteoblast differentiation.(38–40) Although ghrelin treatment led to the activation of many differentiation markers, it did not change Runx2 mRNA expression, suggesting that ghrelin affects osteoblast differentiation in a Runx2-independent manner. However, we have already committed that the cell culture method used in this study involved mature osteoblastic cells as opposed to more primitive stromal cells; this may account for ghrelin's failure to affect Runx2 expression in this study. The detailed mechanisms of ghrelin's modulation of osteoblast differentiation remain unclear and require further study.
It has been reported that growth hormone secretagogues accelerate skeletal growth(41) and increase BMC.(42) These findings suggest that ghrelin upregulates bone formation in vivo. In support of this hypothesis, we found that ghrelin treatment of normal SD rats significantly increased BMD compared with saline treatment. In normal SD rats, ghrelin stimulates GH secretion(1) and can promote bone formation through activation of the GH-IGF-I axis.(20–22) To assess the extend of ghrelin's effects on bone formation through this axis, we also examined the effect of ghrelin on BMD in GH-deficient SDRs. These rats lack GH(43) and thus the GH-IGF-I axis,(41,43) allowing us to assess the direct effect of ghrelin on bone formation. As expected, ghrelin augments BMD, even in SDRs. In general, dynamic loading such as body weight promotes bone formation and increases BMD.(44,45) To eliminate the confounding effects of these factors, we performed an in vivo study using appropriate ghrelin's concentrations,(37) such that body weight and food intake were not increased (data not shown). Together with the findings of the in vitro studies, these data clearly show that ghrelin directly promotes bone formation and increases BMD.
It is interesting to note that ghrelin-null mice do not show any change in BMD and fat deposition patterns compared with wildtype mice,(46) suggesting that ghrelin is not critically required for bone formation. However, bone metabolism is regulated by a number of pathways, implying that a complementary pathway such as that involving PTH, vitamin D, calcitonin, and BMP may maintain bone homeostasis in ghrelin-null mice. The absolute extent to which ghrelin contributes to the promotion of bone formation under normal circumstances remains to be determined.
Considerable attention has been focused on the effect of energy-regulating peptides such as leptin on bone metabolism.(23–29) Leptin, secreted by adipocytes, is known to be a crucial element of the body weight regulatory system.(47) Leptin reduces food intake and body weight, increases energy expenditure, and regulates bone formation.(23–29) Recently, it was reported that leptin inhibited bone formation through the sympathetic nervous system.(48) However, the mechanism of leptin's effect on bone metabolism is complicated and has been reported to be involved in both peripheral and central regulatory mechanisms.(23–29,48,49) Ghrelin, another important peptide in the regulation of energy metabolism, stimulates GH secretion, food intake, and body weight. In addition, it has been reported that ghrelin suppresses sympathetic nerve activity.(50) These findings suggest that ghrelin may also regulate bone metabolism through a neuronal pathway.
In conclusion, this study indicates that ghrelin is a positive regulator that acts directly on osteoblasts.
The authors thank Michihisa Zenmyo, Tetsuya Hamada, and Koji Hiraoka (Department of Orthopedic Surgery, Kurume University School of Medicine) for helpful discussion. This work was supported by a Program for Promotion of Basic Research Activities for Innovative Biosciences (PROBRAIN; to MK), a Grant-in-Aid for Scientific Research (B) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to MK and KN), and the 21st Century COE Program for Medical Science (Kurume University, Research Center of Innovative Cancer Therapy, Molecular Surgery).