MONOCLONAL ANTIBODIES IN IMMUNODIAGNOSIS AND IMMUNOTHERAPY Volume 32, Number 4, 2013 ª Mary Ann Liebert, Inc. DOI: 10.1089/mab.2012.0108
Preparation of Anti-hECSM2 Mouse Monoclonal Antibodies and Their Application in the Analysis of hECSM2 Expression Liangyin Chen,* Cong Ma,* Jing He, Yi He, Jiangman Wang, Lantu Gou, and Jinliang Yang
Human endothelial cell-specific molecule 2 (hECSM2) is a novel, recently identified gene, the biological functions of which are still unclear. The aim of this study was to prepare anti-hECSM2 mouse monoclonal antibodies and investigate the endogenous expression of hECSM2 in cell lines and human tissues. Mouse monoclonal antibodies (MAbs) specifically against hECSM2 were prepared using the hybridoma method. Western blot and flow cytometry were used to detect the specificity of the antibodies. Immunofluorescence and immunohistochemistry were used to investigate the endogenous expression of hECSM2 in different kinds of cell lines and human tissues, respectively. Two anti-hECSM2 MAbs secreting hybridomas were selected. Experiments showed that these two antibodies were highly specific to hECSM2 and endogenous hECSM2 was located on the endothelial cell membrane. Our anti-hECSM2 mouse antibodies can be used for Western blot, flow cytometry, and immunohistochemistry study, and can be a valuable tool for investigating the function and distribution of hECSM2.
uman endothelial cell-specific molecule 2 (hECSM2) is a novel endothelial gene, which was discovered by Lukasz Huminiecki in 2000 using the silico cloning technique.(1) The hECSM2 gene was mapped to human chromosome 5 at position 5q31 and spanned 10.3kb. This gene was predicted to encode 205 amino acids, and the molecule weight for the constituted protein was 21 kDa. The protein of hECSM2 was mainly located on the cell membrane. Bioinformatics analysis and heterologous expression of GFP-, myc, or FLAG-tagged ECSM2 proteins in several mammalian cell systems further suggested that ECSM2 is a cell membrane protein consisting of an N-terminal extracellular domain (ECD), a single transmembrane domain (TM), and a small highly conserved C-terminal intracellular domain (ICD).(2) Analysis of homologous protein demonstrated that the transmembrane domain and C-terminal domain of hECSM2 was highly conserved in many species. At present a few studies on the function of ECSM2 can be found. The previous study demonstrated that the ECSM2 gene was preferentially expressed in vascular endothelial cells (ECs) largely by means of quantitative RT-PCR and in situ hybridization.(3) SiRNA was used to knock down ECSM2 expression, which resulted in reduced chemotaxis and impaired tube formation, suggesting a role for ECSM2 in angiogenesis.(3) A yeast 2 hybrid screen was performed and identified filamin A as a binding partner for the ECSM2 intracellular domain.(3) A reconstitution mammalian cell system demonstrated that
ECSM2 can cross-talk with epidermal growth factor receptor (EGFR) to attenuate EGF-induced cell migration, possibly via inhibiting the Shc-Ras-ERK(MAP kinase) pathway.(4) Cell aggregation and transwell assay showed that ECSM2 promoted cell-cell adhesion and attenuated basic fibroblast growth factor (bFGF)-driven EC migration.(2) Gain or loss of function assay by overexpression or knockdown of ECSM2 in ECs demonstrated that ECSM2 modulated bFGF-directed EC motility via the FGF receptor (FGFR)-extracellular regulated kinase (ERK)-focal adhesion kinase (FAK) pathway.(2) At one time considered merely a monolayer of cells lining the inner face of vascular vessels, endothelia cells have emerged recently as an organ with functions as complex as any in the body. It is not only fundamental for normal organism development, ovulation, embryo formation, and wound healing, but also critically involved in the pathogenesis of many diseases including atherosclerosis and cancer. Thus, there has been a great deal of interest in identifying new genes specifically expressed in vascular endothelial cells and studying their biological functions, which will help to provide new targets for therapies.(5–8) Until now, there has been no high quality commercial antihECSM2 monoclonal antibody in the market. Shi and Ma generated rabbit anti-ECSM2 monoclonal antibodies (RabMAb) using a GST fusion protein containing the entire conserved Cterminal intracellular domain (ICD) of human ECSM2 as the immunogen.(2) In the current study, we generated mouse anti-hECSM2 monoclonal antibodies (MAbs) using CHOECSM2 cells (CHO cells were transfected with the hECSM2
State Key Laboratory of Biotherapy, West China Hospital, West China Medical School, Sichuan University, Chengdu, China. *These authors contributed equally to this work.
CHEN ET AL.
gene) that express the N-terminal extracellular domain (ECD) of human ECSM2 on the cell surface as the immunogen. By the hybridoma method, we successfully selected two hybridomas secreting anti-hECSM2 antibodies that were highly specific to hECSM2 and these antibodies can be used for Western blot, flow cytometry, and immunohistochemistry. This may provide excellent tools for further research on ECSM2’s biological functions. Materials and Methods Bioinformatics of hECSM2 The amino acid sequences of ECSM2 were retrieved from the ExPASy system (www.expasy.org) in Swiss-Prot/TrEMBL database. The transmembrane domain of hECSM2 was predicted by SOSUI (http://bp.nuap.nagoya-u.ac.jp/sosui). Cell culture The whole hECSM2 gene was cloned to pcDNA3.1 vector, and transformed to Chinese hamster ovarian cancer cells (CHO) by eukaryotic expression technology before this experiment. This cell line was named CHO-ECSM2 and preserved in our laboratory. The murine myeloma cell line SP2/0 was grown in RPMI1640 medium with 10% fetal calf serum
(FBS, Hyclone, Beijing, China) and 1 mL 100x penicillin/ streptomycin (PAA Laboratories, Pasching, Austria). The human umbilical vein endothelial cell line (HUVEC), CHO, a human colon cancer cell line HT29, a human breast cancer cell line MCF-7, a human ovarian cancer cell SK-OV-3, and a human renal epithelial cell line 293T were maintained in DMEM medium supplemented with 10% FBS (Hyclone) and 1 mL 100x penicillin/streptomycin (PAA Laboratories). Hybridoma cells were grown in RPMI 1640 medium with 20% FBS supplemented with HAT (Sigma, St. Louis, MO) according to the manufacturer’s instructions. Hybridoma preparation The MAbs against hECSM2 were produced by hybridoma method as described by Ko¨hler and Milstein.(9) Briefly, 8–12 week BALB/c mice were immunized intraperitoneally with CHO-ECSM2 cells (5 · 106 cells per injection) three times. The blood was collected through the mouse tail vein, and then serum titer was tested. The mice with the highest antibody titer were chosen for cell fusion. Screening of hybridomas The screening of positive hybridomas with HUVEC (as a positive control), CHO-ECSM2 cells (experimental group),
Characterization of hECSM2. (A) Amino acid sequences of hECSM2. (B) Schematic structure of hECSM2.
MAbs AGAINST HECSM2 and CHO (as a negative control) fixed with paraformaldehyde was determined by cell-binding ELISA. Cells were seeded in 96-well culture plates (105 cells/well) overnight, then plates were washed three times with a phosphate buffered saline (PBS, pH 7.0) and the cells were fixed with 4% paraformaldehyde for 20 min at room temperature. After three washes with PBS containing 0.05% Tween-20, 200 mL PBS containing 5% skim milk were added to each well and the plates were incubated for 2 h at 37C. Hybridoma supernatant (100 mL) was added to each well and allowed to react for 1 h at 37C. After three washes, 50 mL of peroxidase-conjugated goat antimouse IgG antibody diluted 1:10,000 in PBS were added to each well and the plates were incubated for 1 h at 37C. After three washes, 100 mL enzyme substrate was added to each well and the enzyme reaction was terminated with HCL after 5–20 min. The absorbance at 450 nm was measured by a microplate reader. As a cutoff value for positive hybridomas, we used 2 · OD caused by negative control. The positive clones of hybridomas were subcloned according to the limiting dilution method at least twice to obtain monoclonal cells. Western blot analysis The HUVEC, CHO-ECSM2, and CHO cell lines were collected, then treated with the cell lysis buffer RIPA (Beyotime, Shanghai, China) and ultrasound; the supernatants were separated by SDS-PAGE and then electrophoretically transferred to PVDF membranes. The membranes were blocked
303 with 5% skimmed milk in TBST buffer (20 mM Tris, 150 mM NaCl, and 0.1% Tween-20 [pH 7.5]) overnight at 4C. After blocking, the membrane was incubated with positive hybridoma supernatant that we prepared. The peroxidaseconjugated goat anti-mouse IgG antibody was used as secondary antibody. The blot was developed with ECL Western blot detection reagents (Millipore, Shanghai, China). Flow cytometry analysis Two-step labeling of cell surface molecules was performed by incubation of HUVEC, CHO-ECSM2, and CHO cell lines (1 · 106) with the positive hybridoma supernatants, then the binding of ECSM2 specific MAbs was detected with TIFC– labeled goat anti-mouse Ig after three washes. Ig binding and cell washing steps were conducted at 4C. Cell surface fluorescence was assayed by FACS analysis (Becton Dickerson, Franklin Lakes, NJ). Immunofluoroscence The HUVEC, CHO-ECSM2, and CHO were grown in 6well culture plates overnight. The next day cells were fixed with 4% paraformaldehyde for 20 min, then blocked with 10% normal goat serum and subsequently incubated with positive hybridoma supernatant. After three washes with PBS buffer, cells were incubated with FITC-labeled goat anti-mouse antibody. The cells were washed and visualized under a fluorescent microscope (Leica, Wetzlar, Germany).
FIG. 2. Immunized serum was highly specific to hECSM2. (a) CHO cells were divided into equal parts, then incubated with immunized serum (green line) and equivalent PBS (peak), respectively. The binding to hECSM2 was measured by TIFClabeled goat anti-mouse IgG antibody subsequently. No fluorescent shift was observed. (b) Unimmunized serum reacted to CHO-ECSM2 cells (as in a). No fluorescent shift was observed. (c) Immunized serum reacted to CHO-ECSM2 cells (as in a and b). Fluorescent shift was observed.
304 Immunohistology The paraffin sections of human umbilical cord tissue and placental tissue were made available courtesy of Prof. Ziqiang Wang and Prof. YangMei Shen at West China Medical School (Chengdu). The colorectal cancer tissues were from West China Hospital of Sichuan University. All patients assented to participating in the study with informed consent. The study was approved by the institutional ethics committee of Sichuan University. These tissues were firstly fixed in 4% formaldehyde, dehydrated in ethanol series, treated with xylene, and mounted in paraffin. The 4 mm thick sections of the two tissues were cut and mounted on slides. For immunohistology, after deparaffinization and epitope retrieval in citrate buffer, the nonspecific binding sites were blocked by incubating the section in 10% normal goat serum in PBS (pH 7.0) for 2 h at room temperature. The sections were incubated with the positive hybridoma supernatant as primary antibody overnight at 4C, then washed in TBS and incubated with biotinylated goat anti-mouse IgG diluted in blocking buffer, followed by avidin-biotin-peroxidase complex for 40 min. Finally, the slides were incubated in DAB solution(Boster,Wuhan,China)for2–5 min.Thepositivecontrol
CHEN ET AL. was performed with goat anti-human CD31 antibody. The slides were visualized under a microscope and photographed. Antibody affinity The positive monoclonal antibody affinity was determined by cell-binding ELISA. Three groups of serial diluted HUVEC cell were seeded in 96-well culture plates. The cell density of each plate was decreased by one-half. Cell was cultured for 3 days, then plates were washed three times with PBS (pH 7.0) and the cells were fixed with 4% paraformaldehyde for 20 min at room temperature. After three washes with PBS containing 0.05% Tween-20 (PBST), 200 mL PBS containing 5% skim milk were added to each well and the plates were incubated for 2 h at 37C. After three washes with PBST, serum-free hybridoma supernatant, the exact antibody concentration that was determined by SDS-PAGE, serial diluted, and added to each well. Reaction was allowed for 1 h at 37C. After three washes, 50 mL of peroxidase-conjugated goat anti-mouse IgG antibody diluted 1:5000 in PBS were added to each well and the plates were incubated for 40 min at 37C. After three washes, 100 mL enzyme substrate was added to each well and the enzyme
FIG. 3. Antibodies can be used both for Western blot and FACS. (A) Reactivity of MAb to hECSM2 expressed in eukaryocyte cell. The lysates of CHO, CHO-ECSM2, and HUVEC were separated by SDS-PAGE, and electrophoretically transferred to PVDF membranes. The membrane was incubated with 6C24 and B11 hybridoma supernatant separately. Peroxidase-conjugated goat anti-mouse IgG antibody was incubated as a secondary antibody. The blot was detected with ECL Western blot detecting reagents. Molecular mass markers (kD) are indicated at left. (B) B11 hybridoma supernatant highly specific to ECSM2. (a) CHO cells were divided into equal parts, then incubated with B11 hybridoma supernatant (red line) and equivalent PBS (peak), respectively. The binding to hECSM2 was measured by TIFC-labeled goat anti-mouse IgG antibody. No fluorescent shift was observed. (b) CHO-ECSM2 cells reacted to the same supernatant (as in a). Fluorescent shift was observed. (c) HUVEC cells reacted to the same supernatant (as in a and b). Fluorescent shift was observed. The 6C24 antibody is not shown.
MAbs AGAINST HECSM2
305 were not true antigen concentrations, but were a measure of antigen density on the plate. Results Characterization of hECSM2 The entire amino sequence of hECSM2 (Swiss Prot: Q19T08) is shown in Figure 1A. The hECSM2 is a single transmembrane protein consisting of an N-terminal extracellular domain containing 100 amino acids, a single transmembrane domain involving aa residues 119 to 147 containing 29 amino acids, and a small highly conserved C-terminal intracellular domain containing 58 amino acids. The hECSM2 has a signal peptide containing 20 amino acids (Fig. 1B). Generation of hECSM2-specific mouse MAbs
FIG. 4. Detection of expression of ECSM2 in different cell lines. The lysates of CHO-ECSM2, HUVEC, and unendothelial cell line including CHO, HT29, MCF-7, SK-OV-3, and 293T were separated by SDS-PAGE and electrophoretically transferred to PVDF membranes. The membrane was incubated with B11 hybridoma supernatant, then incubated with peroxidase-conjugated goat anti-mouse IgG antibody as a secondary antibody. The blot was detected with ECL Western blot detecting reagents. Molecular mass markers (kD) are indicated at left. The 6C24 antibodies are not shown.
We use CHO-ECSM2 cells that express the N-terminal extracellular domain of human ECSM2 on the cell surface as the immunogen and immunized four BALB/c mice as described in the Materials and Methods section. The serum titer was tested by cell ELISA using CHO-ECSM2 cell line. The mouse serum with highest titer was allowed to bind to the surface of CHO-ECSM2 cell line by FACS analyses, as indicated by the fluorescence intensity shift (Fig. 2). These results suggested we immunized the mice successfully. With B lymphoma hybridoma technique, the immunized splenocytes were fused with murine myeloma cells (SP2/0) and through continuous screening, two hybridomas that were able to secrete MAbs with high titers were obtained and named 6C24 and B11. Detection of specificity of 6C24 and B11 antibodies
reaction was terminated with HCL after 5–20 min. The absorbance at 450 nm was measured by a microplate reader. Data analysis and sigmoid curve of OD versus logarithm of total antibody concentration graphing were performed by Microsoft Office Excel (Redmond, WA). The affinity constant (Kaff) was calculated for this equation: Kaff = 1/ 2(2[Ab’]t – [Ab]t). [Ab’]t and [Ab]t were the measurable total antibody concentration in the wells at OD-50’ and OD-50 for plates coated with [Ag’] and [Ag], respectively. [Ag’] and [Ag]
Western blot analysis was used to detect the specificity of the two antibodies to the lysate of HUVEC cell line, CHOECSM2 cell line, and CHO cell line. The results showed that all these monoclonal antibodies were highly specific to a 43 kDa cell surface antigen of HUVEC cell line and CHO-ECSM2 cell line but not to CHO cell line (Fig. 3A). FACS analyses were carried out to confirm specificity of the two antibodies. The fluorescence intensity shift was observed in HUVEC cell line and CHO-ECSM2 cell line that express ECSM2 but not CHO
FIG. 5. Subcellular localization of ECSM2. (A) CHO cell line was incubated with B11 hybridoma supernatant, followed by a second incubation with TIFC-labeled goat anti-mouse IgG antibody. (B) CHO-ECSM2 cell line was incubated with the same supernatant (as in A). (C) HUVEC cell line was incubated with the same supernatant (as in A and B). Cells were visualized under a fluorescent microscope. Original magnification, · 400. The hECSM2 was localized at cell membrane. The 6C24 antibody is not shown.
CHEN ET AL.
FIG. 6. Blood vessels of two normal euangiotic human tissues are stained by anti-hECSM2 antibodies. (A) The vein of human umbilical cord tissue was stained by B11 antibody. The specific staining in vascular endothelial cell of vein was shown in the middle with a corresponding positive control and negative control shown at left and right, respectively. (B) The artery of human umbilical cord tissue was stained by B11 antibody. The specific staining in vascular endothelial cell of artery was shown in the middle with a corresponding positive control and negative control shown at left and right, respectively. (C) The blood vessels of human placenta were stained by B11 antibody. Weak staining for hECSM2 was observed lining the endothelial cell of blood vessels within villi (middle); the positive control and negative control are shown at left and right, respectively. Original magnification, · 200. The 6C24 antibody is not shown. cell line (Fig. 3B). Thus, these two antibodies can be applied to both Western blot and FACS. Detection of expression of hECSM2 in different cell lines using 6C24 and B11 antibodies We added another three tumor cell lines, HT29, MCF-7, and SK-OV-3, and one normal cell line, 293T, to assay the expression of ECSM2 using the two antibodies as primary antibodies through Western blot analysis. The results showed that the two antibodies recognized a 43 kDa cell surface antigen of HUVEC cell line and CHO-ECSM2 cell line but not others (Fig. 4). This demonstrated that these two antibodies were highly specific to the cell lines that expressed hECSM2. Subcellular localization of hECSM2 The subcellular localization of endogenous hECSM2 using corresponding antibody has not been demonstrated thus far. In the present study, the 6C24 and B11 antibodies were applied for immunofluorescence. The specific staining was de-
tected in the HUVEC cell line and CHO-ECSM2 cell line but not in the CHO cell line. Moreover, the shining distinct green color showed that the endogenous hECSM2 was localized on the plasma membrane (Fig. 5). Blood vessel of two normal euangiotic human tissues and tumor tissues are specifically stained by 6C24 and B11 antibodies We chose human umbilical cord tissue, human placenta sample, and colorectal cancer tissue, all of which were euangiotic, to evaluate these identified antibodies by immunohistology. Since CD31 was an acknowledged vascular endothelial marker, it was used as positive control to detect hECSM2 expression. The experiment showed strong staining of hECSM2 in the vein and artery of human umbilical cord tissue (Fig. 6A, B) but weak staining in the blood vessel of placenta (Fig. 6C). Meanwhile, endothelial restricted expression of ECSM2 was observed in human colorectal cancer (Fig. 7A, B). These results demonstrated that these anti-hECSM2
MAbs AGAINST HECSM2
FIG. 7. Blood vessel of human colorectal cancer tissues are stained by anti-hECSM2 antibodies. (A) The blood vessels of human colorectal cancer were stained by B11 antibody. Weak staining for blood vessels of colorectal cancer (middle); the positive control and negative control are shown at left and right, respectively. (B) The blood vessels of the adjacent normal tissue were stained by B11 antibody. Weak staining for blood vessels of the adjacent normal tissue (middle); the positive control and negative control are shown at left and right, respectively. Original magnification, · 200. The 6C24 antibody is not shown.
antibodies can be used in immunohistochemistry and are a valuable tool for investigating hECSM2 expression.
was examined in the same fashion and Kaff for 6C24 was 2.2 · 1010 mol/L.
Measurement of antibody affinity
The antibody concentration of B11 and 6C24 (both serumfree hybridoma supernatant) was determined by SDS-PAGE and the exact concentration was 7.5 ng/mL and 5.0 ng/mL, respectively. Based upon the sigmoid curve of OD versus logarithm of total antibody concentration (Fig. 8), the Kaff for B11 antibody was 5.9 · 1010 mol/L. The 6C24 antibody
In the current study, we generated mouse anti-hECSM2 monoclonal antibodies (MAbs). Experiments demonstrated that these antibodies were highly specific to hECSM2 and could be applied in Western blot, FCAS, and immunohistochemistry. Using these antibodies, we initially demonstrated that endogenous hECSM2 mainly localized at the cell
FIG. 8. Sigmoid curves for B11 antibody at different HUVEC cell density. The seeding density of HUVEC was: 1, 10,000 cell/well; 2, 5000 cell/well; 3, 2500 cell/well. · denotes the OD-50 point of each curve. OD-50 value was one-half of the OD100 value (upper plateau). The calculated antibody concentrations (mol/L) at OD-50 were: 1, 7.9 · 10 - 12; 2, 6.3 · 10 - 12; 3, 5.2 · 10 - 12. The graph of 6C24 is not shown.
308 membrane. More importantly, it was the first instance of researching endogenous hECSM2 using anti-hECSM2 mouse antibodies. We found ECSM2 located at cell membrane. The finding was consistent with earlier reports. Previous studies on the distribution of ECSM2 mainly relied on the heterologous expression of GFP-, myc, or FLAG-tagged ECSM2 proteins in mammalian cell systems. It should be noticed that the abundance of a novel gene was usually low.(10–12) So heterologous expression may be causing the wrong distribution of the target protein. We used the corresponding antibodies to detect the endogenous expression of ECSM2. Since there was no exogenous expression of ECSM2, we believe that our approach is valuable in reflecting real subcellular localization. Acknowledgments Special thanks to Prof. Ziqiang Wang for providing paraffin sections of human umbilical cord tissues and Prof. YangMei Shen for providing paraffin sections of placenta tissues. Author Disclosure Statement The authors have no financial interests to disclose. References 1. Huminiecki L, and Bicknell R: In silico cloning of novel endothelial-specific genes. Genome Res 2000;10:1796–1806. 2. Shi C, Lu J, Wu W, Ma F, Georges J, Huang H, Balducci J, Chang Y, and Huang Y: Endothelial cell-specific molecule 2 (ECSM2) localizes to cell-cell junctions and modulates bFGFdirected cell migration via the ERK-FAK pathway. Plos One 2011;6:1–15. 3. Armstrong LJ, Heath VL, Sanderson S, Kaur S, Beesley JFJ, Herbert JMJ, Legg JA, Poulsom R, and Bicknell R: ECSM2, an endothelial specific filamin A binding protein that mediates chemotaxis. Arterioscl Throm Vas 2008;28:1640–1646. 4. Ma F, Zhang D, Yang H, Sun H, Wu W, Gan Y, Balducci J, Wei Y, Zhao X, and Huang Y: Endothelial cell-specific molecule 2 (ECSM2) modulates actin remodeling and epidermal growth factor receptor signaling. Genes Cells 2009;14:281–293. 5. Fogal B, and Pober JS: Vascular endothelial cells as immunological targets in atheroscleroisis. Inflamm Atheroscler 2012;87–114.
CHEN ET AL. 6. Cho SW, Yang F, Son SM, Park HJ, Green JJ, Bogatyrev S, Mei Y, Park S, Langer R, and Anderson DG: Therapeutic angiogenesis using genetically engineered human endothelial cells. J Control Release 2012;160:515–524. 7. Uneda S, Toi H, Tsujie T, Tsujie M, Harada N, Tsai H, and Seon BK: Anti-endoglin monoclonal antibodies are effective for suppressing metastasis and the primary tumors by targeting tumor vasculature. Int J Cancer 2009;125:1446– 1453. 8. Wu W, Shi C, Ma F, Balducci J, Huang H, Ji HL, Chang Y, and Huang Y: Structural and functional characterization of two alternative splicing variants of mouse endothelial cellspecific chemotaxis regulator (ECSCR). Int J Mol Sci 2012; 13:4920–4936. 9. Ko¨hler G, and Milstein C: Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 1975;256: 495–497. 10. Kawaguchi Y, Okamoto T, Taniwaki M, Aizawa M, Inoue M, Katayama S, Kawakami H, Nakamura S, Nishimura M, and Akiguchi I: CAG expansions in a novel gene for Machado-Joseph disease at chromosome 14q32. 1. Nat Genet 1994;8:221–228. 11. Birnbaum MJ: Identification of a novel gene encoding an insulin-responsive glucose transporter protein. Cell 1989; 57:305–315. 12. Long M: Evolution of novel genes. Curr Opin Genet Dev 2001;11:673–680.
Address correspondence to: Prof. Jinliang Yang State Key Laboratory of Biotherapy West China Hospital West China Medical School Sichuan University Keyuan Road 4, Gaopeng Street Chengdu 610041 China E-mail: [email protected]
Received: November 21, 2012 Accepted: March 26, 2013