Yoshiaki Tomiyama, Department of Hematology and Oncology, Graduate School of Medicine C9, Osaka University, 2-2 Yamadaoka, Suita Osaka 565-0871, Japan. Tel.: +81 6 6879 3821; fax: +81 6 6879 3879; e-mail: email@example.com
Summary. In this study, we have identified a patient (OSP-1) with a congenital P2Y12 deficiency showing a mild bleeding tendency from her childhood and examined the role of P2Y12 in platelet function. At low concentrations of agonists OSP-1 platelets showed an impaired aggregation to several kinds of stimuli, whereas at high concentrations they showed a specifically impaired platelet aggregation to adenosine diphosphate (ADP). ADP normally induced platelet shape change and failed to inhibit PGE1-stimulated cAMP accumulation in OSP-1 platelets. Molecular genetic analysis revealed that OSP-1 was a homozygous for a mutation in the translation initiation codon (ATG to AGG) in the P2Y12 gene. Heterologous cell expression of wild-type or mutant P2Y12 confirmed that the mutation was responsible for the deficiency in P2Y12. OSP-1 platelets showed a markedly impaired platelet spreading onto immobilized fibrinogen. Real-time observations of thrombogenesis under a high shear rate (2000 s−1) revealed that thrombi over collagen were small and loosely packed and most of the aggregates were unable to resist against high shear stress in OSP-1. Our data suggest that secretion of endogenous ADP and subsequent P2Y12-mediated signaling are critical for platelet aggregation, platelet spreading, and as a consequence, for stabilization of thrombus.
Platelets play a crucial role not only in a hemostatic plug formation, but also in a pathologic thrombus formation, particularly within atherosclerotic arteries subjected to high shear stress [1,2]. As an initial step in thrombogenesis, platelets adhere to exposed subendothelial matrices such as von Willebrand factor (VWF) and collagen, then become activated and aggregate to each other. These processes are primarily mediated by platelet surface glycoproteins such as GPIb-IX-V, α2β1, GPVI, and αIIbβ3 (GPIIb-IIIa) [3,4]. In addition, several mediators such as adenosine diphosphate (ADP), thromboxane A2, and thrombin cause further platelet activation and recruitment of circulating platelets to the injury sites through activation of αIIbβ3 and subsequent binding of VWF and fibrinogen.
Recent studies have demonstrated a critical role for ADP in arterial thrombogenesis [5–7]. ADP is actively secreted from platelet dense granules on platelet activation and is passively released from damaged erythrocytes and endothelial cells. Platelets possess at least two major G protein-coupled ADP receptors that are largely responsible for platelet responses to ADP: P2Y1 and P2Y12 . P2Y1 is the Gq-coupled receptor responsible for mediating platelet shape change and reversible platelet aggregation through intracellular calcium mobilization [8,9], whereas P2Y12 is the Gi-coupled receptor responsible for mediating inhibition of adenylyl cyclase and sustained platelet aggregation [10–12]. P2Y12 is the therapeutic target of efficacious antithrombotic agents, such as ticlopidine, clopidogrel, and AR-C compounds [5,6], and its congenital deficiency results in a bleeding disorder [13,14]. The analyses of patients with P2Y12 deficiency as well as P2Y12-null mice would provide more precise information about the role of P2Y12 in platelet function than those using P2Y12 inhibitors. To date, four different families with a defect in the expression or the function of P2Y12 have been characterized [10,13–16]. In this study, we have described a patient with the congenital P2Y12 deficiency due to a homozygous mutation in the translation initiation codon and analyzed the role of P2Y12 in platelet aggregation, platelet spreading onto immobilized fibrinogen, and thrombogenesis on a type I collagen-coated surface under a high shear rate. Our present data have demonstrated a crucial role of P2Y12 in various platelet functions.
Materials and methods
The proband (OSP-1) is a 67-year-old Japanese female with a lifelong history of easy bruising. She (OSP-1) was born from non-consanguineous parents who had no hemorrhagic diathesis. Although she showed massive bleeding during delivery of her son, she had no history of transfusions. Patient OSP-1 showed normal platelet count, normal coagulation tests (prothrombin time and activated partial thromboplastin time) and slightly elevated plasma fibrinogen (398 mg dL−1). Ivy bleeding time of the patient was consistently prolonged (>15 min). Clot retraction by MacFarlane's method was normal (50%; normal values 40%–70%). Her son never suffered from a bleeding tendency. Informed consent for analyzing their platelet function and molecular genetic abnormalities was obtained from OSP-1, her husband and their son.
Preparation of platelet-rich plasma and washed platelet suspension
Platelet-rich plasma (PRP) for aggregation studies was prepared by a centrifugation of whole blood anticoagulated with citrate at 250 g for 10 min and then the platelet count was adjusted at 300 × 106 mL−1 by platelet-poor plasma. Washed platelets were prepared as previously described . In brief, 6 volumes of freshly drawn venous blood from the patient, her husband, son or healthy volunteers were mixed with 1 volume of acid–citrate–dextrose (ACD; National Institutes of Health Formula A, NIH, Bethesda, MD, USA) and centrifuged at 250 g for 10 min to obtain PRP. After incubation with 20 ng mL−1 prostaglandin E1 (PGE1; Sigma-Aldrich, St Louis, MO, USA) for 15 min, the PRP was centrifuged at 750 g for 10 min, washed three times with 0.05 mol L−1 isotonic citrate buffer containing 20 ng mL−1 PGE1 and resuspended in an appropriate buffer.
Platelet aggregation using PRP was monitored by a model PAM-6C platelet aggregometer (Mebanix, Tokyo, Japan) at 37 °C with a stirring rate of 1000 r.p.m. as previously described . Protease-activated receptor 1-activating peptide (PAR1 TRAP, SFLLRNPNDKYEPF) and adenosine 3′,5′-diphosphate (A3P5P) were purchased from Sigma-Aldrich Corp. P2Y12 antagonist, AR-C6993MX (2-propylthio-d-fluoromethylene adenosine 5-triphosphate) was a kind gift from AstraZeneca (Loughborough, UK).
Flow cytometry and measurement of intracellular cAMP
Flow cytometric analysis using various monoclonal antibodies (mAbs) specific for platelet membrane glycoproteins was performed as previously described .
For measuring intracellular cAMP levels, samples of 200 μL of washed platelets (60 × 106) in Walsh buffer (137 mm of NaCl, 2.7 mm of KCl, 1.0 mm of MgCl2, 3.3 mm of NaH2PO4, 3.8 mm of HEPES, 0.1% of glucose, 0.1% of BSA, pH 7.4) were incubated with 1 μmol L−1 PGE1 for 15 min, and then platelets were stimulated with ADP or epinephrine. After incubation for 15 min, total cellular cAMP levels were measured using the Biotrak cAMP enzyme immunoassay system from Amersham Pharmacia Biotech (Piscataway, NJ, USA).
Platelet adhesion assay
Adhesion study was performed as previously described . In brief, non-treated polystyrene 10 cm dishes were coated with 100 μg mL−1 human fibrinogen in 5 mL of phosphate-buffered saline (PBS) at 4 °C overnight. After washing with PBS, dishes were blocked with PBS containing 1% of bovine serum albumin (BSA) for 90 min at 37 °C. Aliquots (1 mL) of washed platelets (25 × 106 mL−1) were added to the fibrinogen-coated dishes and incubated at 37 °C. After incubation for 40 min, adherent platelets were fixed with 3.7% formaldehyde, permeabilized with 0.1% Triton X-100 and stained with TRITC-conjugated phalloidin. Platelet morphology and degrees of spreading were determined by fluorescence microscopy (Olympus, Tokyo, Japan).
Platelet thrombus formation under flow conditions
The real-time observation of mural thrombogenesis on a type I collagen-coated surface under a high shear rate (2000 s−1) was performed as previously described . In brief, type I collagen-coated glass coverslips were placed in a parallel plate flow chamber (rectangular type; flow path of 1.9-mm width, 31-mm length, and 0.1-mm height). The chamber was assembled and mounted on a microscope (BX60; Olympus, Tokyo, Japan) equipped with epifluorescent illumination (BX-FLA; Olympus) and a charge-coupled device (CCD) camera system (U-VPT-N; Olympus). Whole blood containing mepacrine-labeled platelets obtained from OSP-1 or control subjects was aspirated through the chamber by a syringe pump (Model CFV-3200, Nihon Kohden, Tokyo, Japan) at a constant flow rate of 0.285 mL min−1, producing a wall shear rate of 2000 s−1 at 37 °C.
Amplification and analysis of platelet RNA
Total cellular RNA of platelets was isolated from 20 mL of whole blood, and P2Y1 or P2Y12 mRNA was specifically amplified by reverse transcription-polymerase chain reaction (RT-PCR), as previously described . The following primers were constructed based on the published sequence of P2Y12 cDNA and used for the first round PCR for P2Y12 cDNA: Y12F1, 5′-GGCTGCAATAACTACTACTTACTGG-3′ [sense, nucleotide(nt) −74 to −50]; Y12R4, 5′-CAGGACAGTGTAGAGCAGTGG-3′ (antisense, nt 85 to 105) .
Genomic DNA was isolated from mononuclear cells using SepaGene kit (Sanko Junyaku Co Ltd, Tokyo, Japan). Amplification of the region around the initiation codon of the P2Y12 gene was performed by using primers BsrDI-GF, 5′-CTTTTGTTCTCTAGGTAACCAACAAGCAA-3′ (sense, the mismatched base is underlined), and Y12R4 (antisense described above) using 250 ng of DNA as a template. These primers can be found in GenBank accession no. AC024886.20 and the sense primer corresponds to 127558–127585. PCR products were then digested with restriction enzyme BsrDI. The resulting fragments were electrophoresed in a 6% polyacrylamide gel.
Construction of P2Y12 expression vectors and cell transfection
The full-length cDNA of wild-type (WT) and mutant P2Y12 was amplified by RT-PCR using primers Y12-HindIII-F, 5′-GAATTCAAGCTTCAAGAAATGCAAGCCGTCGACAACCTC-3′ (sense, nt −6 -21 for WT, EcoRI and HindIII sites introduced at the 5′ end were underlined) or Y12-HindIII-F2, 5′-GAATTCAAGCTTCAAGAAAGGCAAGCCGTCGACAACCTC-3′ (sense, nt −6 −21 for mutant), and Y12H-Not-R, 5′-TCTAGAGCGGCCGCTCAATGGTGATGGTGATGATGCATTGGAGTCTCTTCATT-3′ (antisense, nt 1012–1029, His × 6 were introduced before stop codon, NotI and XbaI sites introduced at the 5′ end were underlined). The amplified fragments were digested with HindIII and NotI, and the resulting 1059-bp fragments (nt −9 −1050) were extracted using QIAquick gel extraction kit (Qiagen, GmbH, Germany). These fragments were inserted into the pcDNA3 (Invitrogen, San Diego, CA, USA) digested with HindIII and NotI. The fragments inserted were characterized by sequence analysis to verify the absence of any other substitutions and the proper insertion of the PCR cartridge into the vector.
A total of 10 μg of WT or mutant P2Y12 construct was transfected into human embryonic kidney 293 cells (HEK293 cells, 106 cells) using the calcium phosphate method as previously described . Transfectants were lyzed by 1% Triton X-100 PBS containing protease inhibitors 2 days after transfection, and proteins were separated by 7.5% SDS-PAGE. After transferred onto a PVDF membrane, expressed proteins were detected by rabbit anti-His tag antibody.
Platelet aggregation studies
We first examined the expression of platelet membrane glycoproteins in OSP-1 by flow cytometry. The patient's platelets (OSP-1 platelets) normally express GPIb-IX, αIIbβ3 (GPIIb-IIIa), α2β1, and CD36 (data not shown). Fig. 1 shows platelet aggregation of PRP in response to various agonists. The aggregation of OSP-1 platelets induced by 20 μm of ADP was markedly impaired with only a small and transient aggregation (Fig. 1A), and the aggregation was still impaired even at 100 μm of ADP (Fig. 1B). As compared with control platelets, the aggregation of OSP-1 platelets was also impaired with a transient aggregation in response to low concentrations of collagen (0.5 μg mL−1, Fig. 1E), U46619 (0.63 μm, Fig. 1F), or PAR1 TRAP (25 μm, Fig. 1H). In response to 1.3 mg mL−1 ristocetin (not shown) or 10 μm of epinephrine (Fig. 1G), OSP-1 platelets aggregated normally. When OSP-1 platelets were stimulated with 20 μm of ADP in the presence of 5 mm of EDTA, the light transmission decreased equivalent to control platelets suggesting that OSP-1 platelets changed shape normally (Fig. 1D). We then examined effects of ADP receptor antagonists on the aggregation of OSP-1 platelets induced by 20 μm of ADP. A total of 1 mm of A3P5P, a specific P2Y1 antagonist, abolished the residual response of OSP-1 platelets to ADP, whereas 1 μm of AR-C69931MX, a specific P2Y12 antagonist, did not induce an additional inhibition on the platelet aggregation (Fig. 1C). These data suggest that the impaired response of the patient's platelets may be due to an abnormality in signaling evoked by ADP and that P2Y12-mediated signaling rather than P2Y1-mediated signaling may be completely defective in patient OSP-1.
We also examined the aggregation of OSP-1 platelets induced by higher concentrations of agonists. As shown in Fig. 2, the aggregation response of OSP-1 platelets improved as the concentrations of agonists increased, and they aggregated almost normally in response to high concentrations of collagen (5 μg mL−1), U46619 (5 μm), or PAR1 TRAP (100 μm) (not shown). In addition, we confirmed that 1 μm of AR-C69931MX conferred essentially the same defect on the aggregation of control platelets in response to U46619 as that of OSP-1 platelets and did not further inhibit OSP-1 platelet aggregation induced by 5 μg mL−1 of collagen, 5 μm of U46619, or 100 μm of PAR1 TRAP (data not shown). These data indicated that at high concentrations of agonists OSP-1 platelets showed the specifically impaired aggregation to ADP.
Effect of ADP on PGE1-stimulated cAMP accumulation in platelets
To determine whether P2Y12-mediated signaling is specifically impaired, we examined an inhibitory effect of ADP on 1 μm of PGE1-stimulated cAMP accumulation in platelets from the patient, her husband, their son, and healthy unrelated controls. ADP inhibited intracellular cAMP levels in platelets from the patient's husband, son and healthy unrelated controls (not shown) by approximately 80%, whereas the inhibition was only 15% in the patient's platelets (Fig. 3). In contrast to ADP, epinephrine normally inhibited cAMP accumulation in platelets from the patient as well as her husband and son. These results strongly suggest that the defect could be due to an abnormality in Gi coupling ADP receptor, P2Y12.
Nucleotide sequence analysis of cDNA and genomic DNA of P2Y12
To reveal a molecular genetic defect in OSP-1, we analyzed the entire coding regions of both P2Y1 and P2Y12 cDNAs amplified from platelet mRNA by RT-PCR. A single nucleotide substitution (T → G) was identified within the translation initiation codon (ATG → AGG) in the patient's P2Y12 cDNA (Fig. 4A). This substitution was also confirmed by reverse sequencing. No other nucleotide substitutions were detected within the coding region of either P2Y12 or P2Y1 cDNA from the patient. OSP-1 appeared homozygous for the substitution, and the substitution was not detected in 20 control subjects.
Nucleotide sequence analysis of PCR fragments from the patient's genomic DNA also suggested the homozygosity of the substitution (data not shown). To further confirm the homozygosity, allele-specific restriction enzyme analysis (ASRA) was performed. The region around the initiation codon of the P2Y12 gene was amplified by PCR using primers BsrDI-GF and Y12R4. A restriction site for BsrDI would be abolished by the T → G substitution. As shown in Fig. 4B, ASRA clearly indicated that the patient and her son were homozygous and heterozygous for the substitution, respectively. These results also confirm that the substitution is inheritable.
Heterologous cell expression of WT and mutant P2Y12
As the substitution at the translation initiation codon might induce an alternative translation starting at downstream ATGs leading to an expression of shorter form of P2Y12, we decided to investigate effects of the substitution found in the patient on the expression of P2Y12. Expression vectors encoding WT and mutant P2Y12 in which His-tag was attached at the C-terminal portion of P2Y12 were constructed as described in the Materials and methods. Wild-type or mutant P2Y12 construct was transfected into HEK 293 cells, and then expressed proteins were analyzed 48 h after transfection in an immunoblot assay employing anti-His antibodies. As shown in Fig. 5, WT P2Y12 protein with an apparent molecular weight of ∼60 KDa was expressed in 293 cells as a His-tag-positive protein. In sharp contrast, the mutant P2Y12-expression vector failed to express any His-tag-positive protein. These results provide strong evidence that the T → G substitution at the translation initiation codon of P2Y12 cDNA is responsible for the P2Y12 deficiency.
Platelet spreading on immobilized fibrinogen
As it has been well documented that release of endogenous ADP is required for full platelet spreading onto immobilized fibrinogen , we next analyzed the patient's platelet spreading in order to evaluate the role of P2Y12. Control platelets adhered to fibrinogen underwent morphological changes ranging from filopodia protrusion to complete spreading, and 50.5% ± 21.3% of the adherent platelets spread (n = 3) (Fig. 6A). In sharp contrast, the patient's platelets showed an impaired spreading and only 2.3% ± 1.4% of the adherent platelets spread (n = 3, P < 0.001, Fig. 6C). Similar results were obtained with control platelets in the presence of 1 μm of AR-C69931MX (6.2% ± 2.2%, n = 3, P < 0.001, Fig. 6B). In addition, 1 mm of A3P5P also markedly inhibited platelet spreading (n = 3, 10.1% ± 2.2%, P < 0.001, not shown). These results suggest that both P2Y12 and P2Y1 are necessary for platelet spreading.
Platelet-thrombus formation on immobilized collagen under flow conditions
To investigate the role of P2Y12 in thrombus formation, we observed the real-time process of mural thrombogenesis on a type I collagen-coated surface under flow conditions with high shear rate (2000 s−1) using the whole blood from OSP-1. Real-time observation revealed that thrombi formed on type I collagen were unstable. As platelet aggregates of the patient were loosely packed each other and unable to resist against high shear stress, most of the aggregates at the apex of the thrombi came off the thrombi constantly. On the other hand, most of thrombi formed by control platelets were densely packed with higher fluorescent intensity and were stable with constant growth during observation (Video 1 and 2).
As shown in Fig. 7A, the area covered with patient platelets after 7 min of flow was greater than that of control platelets (91.8% ± 0.3% vs. 82.2% ± 1.4%, n = 3, P < 0.01). However, thrombi formed by OSP-1 platelets were loosely packed, whereas thrombi were large and densely packed in controls. The overall fluorescent intensity of thrombi of OSP-1 platelets was lower than that of control platelets. Three-dimensional analysis revealed the striking difference in size and shape of individual thrombus formed after 10 min between the patient and control platelets (Fig. 7B). Thrombi formed by control platelets were large in size, clearly edged and surrounded by thrombus-free areas. On the other hand, individual thrombus formed by patient platelets was mostly small and appeared to be a thin layer of platelet aggregates. Thrombus height at the plateau phase was 10.2 ± 0.4 μm, which was less than half of controls (21.2 ± 0.4 μm).
P2Y12 coupled with Gαi, primarily with Gαi2, consists of 342 amino acid residues with seven transmembrane domains (TM), and its deficiency is responsible for congenital bleeding diathesis [10–16]. To date, five mutations responsible for a defect in the expression or the function of P2Y12 in four different families have been demonstrated [10,15,16]. Patient ML possessed a mutation consisting of a two nucleotide deletion at amino acid 240 (near the N-terminal end of TM6), which would lead to a premature termination of P2Y12 [10,14]. A two nucleotide deletions at amino acid 98 (next to the N-terminal end of TM3) and a single nucleotide deletion occurring just beyond TM3 were identified in other two families, both of which would lead to a premature termination of P2Y12 [13,15]. However, in these reports expression studies had not been performed to show the direct association between these mutations and the P2Y12 deficiency [10,15]. Patient AC, whose platelets expressed dysfunctional P2Y12 with normal binding capacity of 2-methylthioadenosine 5′-diphosphate (2MeS-ADP), was compound heterozygous for Arg256 →Gln in TM6 and for Arg265 → Trp in the third extracellular loop of P2Y12. Platelets from patient AC showed an increased platelet aggregation at high dose ADP compared with low dose ADP, suggesting the presence of residual receptor function . In this study, we described a patient (OSP-1) with congenital bleeding diathesis bearing a novel homozygous mutation within the translation initiation codon (ATG → AGG) of the P2Y12 gene. Consistent with previous studies, the aggregation of OSP-1 platelets with P2Y12 deficiency was impaired to various agonists such as collagen, U46619, and PAR1 TRAP at low concentrations, but almost normal at high concentrations [11–14]. These findings confirmed the critical role of P2Y12-mediated signaling evoked by endogenous ADP in platelet aggregation induced by low concentrations of agonists. In contrast to patient AC with residual P2Y12-mediated signaling, the impaired platelet aggregation in OSP-1 in response to ADP was neither improved even at 100 μm of ADP stimulation nor reduced by adding 1 μm of AR-C69931MX, suggesting a complete loss of the P2Y12 function. Family study confirmed that patient OSP-1 was homozygous for the mutation, and our expression study demonstrated that the mutation is responsible for the P2Y12 deficiency.
A number of examples of mutations in the translation initiation codons have been demonstrated in various diseases . Although some cases having mutations in the initiation codons did not express any related abnormal protein, Pattern et al. showed an abnormal Gαs protein possibly synthesized as a result of initiation at downstream ATGs due to a mutation at an initiation codon (ATG →GTG) in patients with Albright's hereditary osteodystrophy [24,25]. In OSP-1, we detected the T→ G mutation at position +2, and our expression study denied the possibility that the substitution might induce an alternative translation at downstream ATGs leading to an expression of shorter form of P2Y12.
As to platelet spreading onto immobilized fibrinogen, OSP-1 platelets showed the impaired platelet spreading. Similarly, A3P5P inhibited the platelet spreading. It has been well documented that release of endogenous ADP is required for full platelet spreading onto immobilized fibrinogen , and Obergfell et al.  have demonstrated that the platelet spreading requires sequential activation of Src and Syk in proximately to αIIbβ3. In contrast to the ADP-induced platelet shape change shown in OSP-1 platelets in the platelet aggregometer, our data indicated that both P2Y12 and P2Y1 were necessary for the full spreading onto immobilized fibrinogen.
Employing clopidogrel or AR-C69931 MX as an inhibitor for P2Y12, several studies examined the role of P2Y12 in thrombogenesis under flow conditions [27–30]. However, some discrepancy between the studies was pointed out and non-specific effects of these inhibitors were not completely ruled out [28–30]. As patient OSP-1 was deficient in P2Y12 as shown in this study, it would be informative to examine the real-time process of thrombogenesis on a type I collagen-coated surface under a high shear rate (2000 s−1) employing whole blood obtained from OSP-1. Our data demonstrated that P2Y12-deficiency led to the loosely packed thrombus and the impaired thrombus growth with enhancing adhesion to collagen, which was consistent with the study by Remijn et al.  employing patient ML's platelets. The increase in platelet adhesion to collagen was probably due to the impaired platelet consumption by the growing thrombi [27,30]. Moreover, our real-time observation indicated that the loosely packed aggregates were unable to resist against high shear stress, and then most of the aggregates at the apex of the thrombi came off the thrombi. In contrast, Andre et al.  did not detect significant differences in ex vivo thrombus volume formed over human type III collagen under a shear rate of 871 s−1 between P2Y and WT mice. Although Andre et al. used non-anticoagulated mouse blood instead of anticoagulated blood, Roald et al.  demonstrated that clopidogrel significantly reduced the thrombus volume over type III collagen employing non-anticoagulated human blood. Loosely packed platelet thrombi with swollen non-degranulated platelets were detected following clopidogrel intake, whereas densely packed thrombi with partly fused platelets were detected before clopidogrel intake by electron microscopy . Thus, it is likely that differences between human and mouse, rather than those between non-anticoagulated and anticoagulated blood, may account for the discrepancy. Nevertheless, they showed that ex vivo thrombi were loosely packed and that only small and unstable thrombi were formed in P2Y mice without reaching occlusive size in mesenteric artery injury model in vivo .
Our present study has demonstrated the novel mutation responsible for the P2Y12 deficiency and suggested that secretion of endogenous ADP and subsequent P2Y12-mediated signaling is critical for platelet aggregation, platelet spreading, and as a consequence, for stabilization of thrombus. Mild bleeding tendency observed in patient OSP-1 further emphasizes the efficacy of P2Y12 receptor as a therapeutic target for thrombosis.
We thank Dr Mitsuhiko Sugimoto (Nara Medical University) for his valuable advice to perform the real-time observation of thrombogenesis under flow conditions. This study was supported in part by Grant-in Aid for Scientific Research from the Ministry of Education, Science and Culture in Japan, Grant-in Aid from the Ministry of Health, Labor and Welfare in Japan, Astellas Foundation for Research on Metabolic Disorder, Tukuba, Japan, and Mitsubishi Pharma Research Foundation, Osaka, Japan.