Rapid screening and identification of dominant B cell epitopes of HBV surface antigen by quantum dot-based fluorescence polarization assay
- Zhongji Meng†2,
- Ruihua Song†1,
- Yue Chen2,
- Yang Zhu1,
- Yanhui Tian1,
- Ding Li1Email author and
- Daxiang Cui3Email author
© Meng et al.; licensee Springer. 2013
Received: 20 December 2012
Accepted: 20 February 2013
Published: 2 March 2013
A method for quickly screening and identifying dominant B cell epitopes was developed using hepatitis B virus (HBV) surface antigen as a target. Eleven amino acid fragments from HBV surface antigen were synthesized by 9-fluorenylmethoxy carbonyl solid-phase peptide synthesis strategy, and then CdTe quantum dots were used to label the N-terminals of all peptides. After optimizing the factors for fluorescence polarization (FP) immunoassay, the antigenicities of synthetic peptides were determined by analyzing the recognition and combination of peptides and standard antibody samples. The results of FP assays confirmed that 10 of 11 synthetic peptides have distinct antigenicities. In order to screen dominant antigenic peptides, the FP assays were carried out to investigate the antibodies against the 10 synthetic peptides of HBV surface antigen respectively in 159 samples of anti-HBV surface antigen-positive antiserum. The results showed that 3 of the 10 antigenic peptides may be immunodominant because the antibodies against them existed more widely among the samples and their antibody titers were higher than those of other peptides. Using three dominant antigenic peptides, 293 serum samples were detected for HBV infection by FP assays; the results showed that the antibody-positive ratio was 51.9% and the sensitivity and specificity were 84.3% and 98.2%, respectively. In conclusion, a quantum dot-based FP assay is a very simple, rapid, and convenient method for determining immunodominant antigenic peptides and has great potential in applications such as epitope mapping, vaccine designing, or clinical disease diagnosis in the future.
An epitope or antigenic determinant is the core part of antigen involved in the recognition with an antibody. After antigens, containing numerous epitopes, are recognized by the human immunological system, B lymphocytes will synthesize and secrete miscellaneous antibodies targeting different epitopes to mediate further immunological process. Nowadays, there are many methods that are used to investigate and confirm antigen epitopes, for example, proteolytic cleavage of antigen-monoclonal antibody complexes, proteolytic or chemical cleavage fragment method, Western blotting, PEPSCAN method, chemical modification or mutation analyses, chemosynthesis of peptide, surface display of peptide libraries and random fragment expression libraries, X-ray crystallographic assay and nuclear magnetic resonance spectroscopy assay, and so on . However, most of these methods are complicated, difficult to perform, or of low efficiency. With the development of computer technology and bioinformatics, a set of methods for epitope prediction were developed based on the structural property of amino acid sequences ; the accuracy of prediction of antigenic determinants is about 75% . The methods for epitope prediction can reduce the range of possible epitope and bring us much less workloads for epitope screening. However, it is possible that some of the predicted epitopes exhibit no strong antigenicity. So, developing a novel method to analyze the antigenicity of predicted peptides has become an urgent requirement for epitope determination.
Fluorescence polarization (FP) is a unique and powerful technique for the rapid analysis of interactions of small molecular weight molecules (labeled with fluorophore) and macromolecules. The theory of fluorescence polarization was for the first time described in 1926 by Perrin . When fluorescent molecules in solution are excited by a plane-polarized light beam with an appropriate wavelength, they emit fluorescent signals back into the same polarized plane, provided that the molecules remain stationary. However, if the excited molecules rotate or tumble while in the excited state, then fluorescence is emitted into a plane different from the plane used for excitation. Therefore, if the viscosity and temperature of a solution are kept constant, the degree of fluorescence polarization is dependent on the molecular volume, that is, the size of a fluorescent molecule. FP assay is based on the rotational differences between a small soluble molecule in solution (labeled with a fluorochrome) and the small molecule combined with its ligand. A small molecule can rotate randomly at a rapid rate, resulting in rapid depolarization of light, while a larger complex molecule can rotate slower and depolarize light at a reduced rate. The rate change in depolarization can be measured. High polarization values indicate that the small molecule is reacting with its ligand or target molecule, and low values mean that there is no small molecule ligand or small molecule to react with target molecule. Nowadays, homogeneous FP assays have been successfully applied to many research fields, including DNA-protein, protein-protein, and antigen-antibody binding [5–11]. However, the current FP assay is based on organic dye labeling, having some problems such as intrinsic photobleaching and low-emission efficiency, and how to solve these questions has become a great challenge.
Quantum dots have been subject to intensive investigations due to their unique properties and potential application prospect [12–14]. So far, several methods have been developed to synthesize water-soluble quantum dots (QDs) for use in biologically relevant studies [15–18]. QDs exhibit high quantum yield, high photostability, and size-dependent tunable emission, being attractive alternative luminescent labels for ultrasensitive detection and molecular imaging. For example, QDs have been used successfully in cellular imaging [19, 20], immunoassays , DNA hybridization , and optical barcoding . Quantum dots provide a new functional platform for bioanalytical sciences and biomedical engineering. Therefore, it is feasible to use QD labeling to improve the FP technique for detection of tumor biomarkers in patient sera [24, 25].
If micromolecular antigens are adopted, FP assays can also be used to analyze the interaction of the antigen and its antibody. Herein, we reported a CdTe quantum dot-based method to screen rapidly antigenic epitopes. All possible antigenic epitopes from hepatitis B virus (HBV) surface antigen protein were predicted, and the antigenicity of peptide was determined by analyzing the recognition and combination of peptide and standard antibody samples using the FP technique. Subsequently, the immunodominant epitopes of HBV surface antigen in Chinese people with positive anti-HBV surface antigen were screened using the same method. Besides, the application of the obtained dominant antigenic peptides in detecting anti-HBV surface antibody was also investigated by FP assay.
Peptide sequence design
Designed antigenic peptide sequences from HBV surface antigen protein
No. of peptides
Amino acid sequences
Location in HBV surface antigen protein
14 to 32
56 to 62
70 to 77
80 to 93
105 to 114
132 to 143
147 to 164
217 to 224
229 to 248
291 to 300
307 to 314
Synthesis of antigenic peptides
All peptides were synthesized on 2-chlorotrityl chloride resin (1.6 mmol/g) using the standard solid-phase method of 9-fluorenylmethoxy carbonyl (Fmoc) chemistry . Peptides were produced on a 0.2-mmol scale, and Fmoc-preactivated amino acids as pentafluorophenyl esters were used for the coupling reactions in the presence of hydroxybenzotriazole (Sigma Chemical Co., St. Louis, MO, USA) in dimethylformamide (DMF). Excess amino acids were used throughout the synthesis. Chain elongation reaction was performed followed by Fmoc deprotection in 20% piperidine in DMF.
When the chain elongation reaction was finished, the Fmoc protecting groups were removed from the N terminus of the peptides by 25% piperidine in DMF followed by washing with DMF for four times. Following washing for four times with DMF and dichloromethane, the resin was dried under vacuum. Subsequently, the as-prepared peptides were cleaved from the resin using standard trifluoroacetic acid (TFA) cleavage procedures in TFA with 5% H2O followed by multiple ether extractions. All synthetic peptides were purified to >95% by reverse-phase high-pressure liquid chromatography performed with a liquid chromatograph (Waters, Milford, MA, USA). Peptides were analyzed by mass spectrometry to confirm that the desired product was obtained.
Preparation of QDs and QD-peptide conjugates
The CdTe QDs were prepared according to our previous report . Briefly, 5 mmol of CdCl2·5H2O was dissolved in 110 mL of water, and 12 mM of thioglycolic acid (TGA) was added under stirring. NaOH solution was used to adjust the pH of the resultant solution to 11. The solution was cleared by N2 bubbling for 30 min. Under stirring, 2.5 mM of oxygen-free NaHTe solution was injected into the solution. After reflux at 100°C for 4 h, the TGA-capped CdTe QDs were synthesized. The obtained QDs were purified by precipitation in ethanol and redispersed in phosphate-buffered saline (PBS; 0.2 mg/mL KCl, 1.44 mg/mL Na2HPO4, 0.24 mg/mL KH2PO4, 8 mg/mL NaCl; pH 7.4). Absorbance spectrum and photoluminescence spectrum were analyzed to characterize the fluorescent properties of QDs with a PerkinElmer LS 55 spectrofluorimeter (Waltham, MA, USA).
Afterwards, 0.5 mL of 3 mg/mL QDs and 0.5 mL of 0.8 mg/mL antigenic peptides were mixed, and then 50 μL of 1 mg/mL EDC was added. The resulting solution reacted at room temperature for 3 h with continuous mixing and then stayed at 4°C for 24 h. Bovine serum albumin (BSA) was added into the solution at a concentration of 1 mg/mL and incubated at room temperature for 3 h. The QD-labeled SPAs were then centrifuged at 15,000×g for 30 min, and the supernatant was discarded. A volume of 1.05 mL PBS with 0.5% Tween-20 (PBST;, v/v) was used to resuspend and wash QD-labeled antigenic peptides by centrifugation at 15,000×g for three times. Finally, the QD-labeled conjugates were dispersed in 1.05 mL PBST and kept at 4°C for usage. Then, 1% agarose gel electrophoresis was performed to analyze the QD-peptide conjugates.
Standard serum samples
HBV-positive sera were collected from patients who were confirmed by enzyme-linked immunosorbent assay (ELISA) test. The negative sera were collected from healthy volunteers. One hundred anti-HBV surface antigen antibody-positive sera and 100 negative sera were mixed separately at equal volume ratio. The mixtures were used as standard antibody-positive and antibody-negative serum samples.
Fluorescence polarization assay of QD-labeled HBV antigenic peptide binding to standard antibody
Fluorescence polarization experiments were performed in a black 384-well plate (MJ Research, Waltham, MA, USA) using a Wallac Victor2 (1420 multilabel counter) fluorescence polarization analyzer (PerkinElmer Life Sciences). Assays were done at room temperature using filters for fluorescein excitation (480 nm) and emission (595 nm). To obtain optimal concentration for fluorescence polarization assay, QD-labeled antigenic peptides were diluted to different concentrations (from 0 to 2.5 nM, at intervals of 0.25 nM) in PBS, each of the samples was added to three wells of the 384-well plate (25 μL/well), and then the fluorescence polarization of the samples was measured. The results of the FP assay were expressed as millipolarization (mP) values, and the experiment was repeated three times.
To reduce the interference to FP values caused by impurities existing in serum samples, different dilutions (1:5, 1:10, 1:15 to 1:55) of standard serum samples were tested for FP assay. Serum samples were diluted with 2.5 nM QD-labeled peptide/PBS buffer (containing 0.2 mg/mL BSA). After thorough mixing, the mixture was added to three wells of the 384-well plate (25 μL/well) and incubated for 30 min before reading. This assay was repeated to obtain the reaction time needed for binding saturation with changed incubation time (0, 2, 5, 10, 15, 20, 25, and 30 min). The positive standard serum, negative standard serum, and diluent buffer blank control were included in the test.
According to optimal reaction factors, the antigenicity of all synthetic peptides was identified by analyzing the recognition and combination between peptides and standard antibody samples using the FP method. When the peptides bind to specific antibodies, the FP values will increase, and the increment can express the antigenicity indirectly.
Screening for immunodominant antigenic peptides
One hundred fifty-nine samples of anti-HBV surface antigen-positive antisera were identified by the standard ELISA method with commercial ELISA kits. Specific antibodies against each peptide of HBV surface antigen with distinct antigenicity were detected using the FP method in all the antiserum samples. The distribution and levels of specific antibody against each peptide were analyzed according to the results of the FP assay.
Detecting for HBV infection by FP assay
Using the immunodominant antigenic peptides, 293 serum samples were detected for HBV infection based on the FP assay. In order to evaluate the FP method for detection of HBV infection, ELISA experiment was carried out using a commercial ELISA kit for detection of IgG of anti-HBV. The ELISA results were used as real results; then, receiver operating characteristic (ROC) curve analysis (MedCalc Software, Ostend, Belgium) was performed on the FP assay results to determine the optimal cutoff point (at which the sum of the sensitivity and specificity values is maximal) to distinguish between positive and negative FP assay results. The ROC curve (a plot of the true-positive rate (sensitivity) against the false-positive rate (100-specificity) that is obtained at each cutoff point) was constructed, and the area under the curve (AUC) value was calculated as a measure of the accuracy of the test.
Results and discussion
QD conjugates and their fluorescence polarization property
Dilution of serum for FP assay
Incubation time for FP assay
Antigenicity of synthetic peptides
Immunodominant peptides of HBV surface antigen
The results of FP assay detecting antibodies against 10 antigenic peptides in 159 serum samples
No. of peptides
Numbers of samples (n= 159)
ΔmP ≤ 25
25 < ΔmP ≤ 50
50 < ΔmP ≤ 100
ΔmP > 100
Detection of HBV infection using immunodominant peptides based on FP assay
FP assay can be used in the study of antigen-antibody interaction, and the attachment of fluorescein to antigen does not affect its ability to bind with an antibody. The only limitation of the FP assay is the size of the antigen, and the principle of FP assay restricts that only the micromolecular antigen is suitable for interaction analysis. The more differences of molecular weight between antigen and antibody exist, the more differences of FP values between free antigen and antigen-antibody complex can be measured. The molecular mass of synthetic antigenic peptides is far smaller than general antigens, so they are suitable for the screening of antigenic epitopes by the FP method. By investigating the interaction between peptide and standard antibody sample, the antigenicity of this peptide can be easily determined. For instance, when the QD-labeled peptides are mixed with the standard antibody in solution, if the peptides have antigenicity, they can bind with antibodies rapidly. The formations of antigen-antibody complex slow the rotation of the fluorescent tracer and thereby increase the polarization of the emitted light compared with only peptides existing in solution. On the other hand, the polarization has no change if the peptides have no antigenicity. In other words, a high FP value represents a strong antigenicity of peptides, and a low value represents a weak antigenicity of peptides after FP measure. When the peptides reacted with standard antibody-positive serum, in this report, the measured FP values of the 10 of 11 HBV synthetic peptides were between 200 and 250 mP, far higher than the FP values of the peptides that reacted with the standard antibody-negative serum, which were only about 150 to 170 mP, and these peptides may have antigenicity.
In order to optimize the FP assay used in detecting the interaction of antigenic peptide and antibody, we investigated the effects of different concentrations of fluorophore-labeled peptides, different dilution times of serum samples, and different incubation times of antigen-antibody mixture on FP assay. The obtained optimal factors are as follows: 1 nmol/L of fluorophore-labeled peptides, 1:25 of dilution times of serum samples, and 15 min of incubation time of antigen-antibody mixture.
The established FP assay not only can be used to identify the antigenicity of peptides with standard antibody, but can also be used to detect antibodies in serum samples with known antigenic peptides because the usages of FP assay in the two aspects share the same principle and procedures. By analyzing the antibody levels against 10 antigenic peptides in 159 anti-HBV surface antigen-positive antiserum using FP assay, we found that the antibody levels against nos. 1, 10, and 11 antigenic peptides were higher than those against other peptides or the antibodies against these peptides widely existed in 159 samples. This indicates that the three peptides may be immunodominant among 159 samples.
Based on the three dominant antigenic peptides, we also study the application of FP assay in detecting HBV infection. The FP assay data were subjected to ROC curve analysis which estimates the sensitivity and specificity of a test at every possible cutoff point and provides a measure of test accuracy. The ROC curve that was obtained from the analysis of the FP assay results indicated that the three dominant antigenic peptides are accurate indicators of HBV infection. The antibody-positive ratio was 51.9%, analyzed using the three antigenic peptides; the sensitivity and specificity estimates at the cutoff point 77 mP were 85.4% and 98.6%, respectively.
In conclusion, homogeneous QD-based FP assay offers several advantages in analyzing the interaction of peptide antigen and antibody. This assay is a single-step primary binding assay using a single reagent - the QD-labeled antigenic peptides. The assay can be completed in a few minutes. Secondly, FP assay requires no repetitive washing procedures to remove unbound reactants. This also decreases the assay time considerably. In addition, the outstanding optical quality of QDs in photostability makes them an excellent fluorescent reporter. Due to the simple and rapid manipulations and high sensitivity and specificity, FP assay is very suitable for high-throughput screening of antigenic peptides and screening of immunodominant epitopes. The technical simplicity, rapid speed, and low cost of this assay make it very attractive in specific antibody detection and clinic serological tests of infectious diseases. In one word, FP assay has great applied potential in epitope mapping, vaccine designing, or clinical disease diagnosis in the future.
This work was supported by the National Nature Science Foundation of China (no. 30972608), Beijing Medicine Research and Development Fund (no. 2009–2048), and Important National Science & Technology Specific Projects (2009ZX10004-311).
- Morris Glenn E: Overview. In Epitope Mapping Protocols. Methods in Molecular Biology. Volume 66. Edited by: Morris Glenn E. Totowa: Humana; 1996:1–9.View ArticleGoogle Scholar
- Carter JM: Epitope prediction methods. Methods Mol Biol 1994, 36: 193–206.Google Scholar
- Kolaskar AS, Tongaonkar PC: A semi-empirical method for prediction of antigenic determinants on protein antigens. FEBS Lett 1990, 276: 172–174. 10.1016/0014-5793(90)80535-QView ArticleGoogle Scholar
- Perrin F: Polarization of light of fluorescence, average life of molecules in the excited state. J Phys Radium 1926, 7: 390–401. 10.1051/jphysrad:01926007012039000View ArticleGoogle Scholar
- Radek JT, Jeong JM, Wilson J, Lorand L: Association of the A subunits of recombinant placental factor XIII with the native carrier B subunits from human plasma. Biochemistry 1993, 32: 3527–3534. 10.1021/bi00065a002View ArticleGoogle Scholar
- Wei AP, Herron JN: Use of synthetic peptides as tracer antigens in fluorescence polarization immunoassays of high molecular weight analytes. Anal Chem 1993, 65: 3372–3377. 10.1021/ac00071a007View ArticleGoogle Scholar
- Lin M, Nielsen K: Binding of the Brucella abortus lipopolysaccharide O-chain fragment to a monoclonal antibody. Quantitative analysis by fluorescence quenching and polarization. J Biol Chem 1997, 272: 2821–2827. 10.1074/jbc.272.5.2821View ArticleGoogle Scholar
- Onnerfjord P, Eremin S, Emneus J, Marko-Varga G: Fluorescence polarisation for immunoreagent characterization. J Immunol Methods 1998, 213: 31–39. 10.1016/S0022-1759(98)00019-2View ArticleGoogle Scholar
- Nasir MS, Jolley ME: The use of fluorescence polarization assays for the detection of infectious diseases. Comb Chem High Throughput Screen 2003, 6: 235–244. 10.2174/138620703106298419View ArticleGoogle Scholar
- Surujballia OP, Romanowskaa A, Sugdena EA, Turcottea C, Jolley ME: A fluorescence polarization assay for the detection of antibodies to Mycobacterium bovis in cattle sera. Vet Microbiol 2002, 87: 149–157. 10.1016/S0378-1135(02)00044-5View ArticleGoogle Scholar
- Nielsen K, Lin M, Gall D, Jolley M: Fluorescence polarization immunoassay: detection of antibody to Brucella abortus . Methods 2000, 22: 71–76. 10.1006/meth.2000.1038View ArticleGoogle Scholar
- Jeong H, Chang AM, Melloch MR: The Kondo effect in an artificial quantum dot molecule. Science 2001, 293: 2221–2223. 10.1126/science.1063182View ArticleGoogle Scholar
- Li XQ, Wu YW, Steel D, Gammon D, Stievater TH, Katzer DS, Park D, Piermarocchi C, Sham LJ: An all-optical quantum gate in a semiconductor quantum dot. Science 2003, 301: 809–811. 10.1126/science.1083800View ArticleGoogle Scholar
- Daxiang C, Hong Z, Jie S, Zheng W, Asahi T, Rong H, Osaka T, Feng G, Hoon Sung C, Chris H, Hengyao H, Pauletti GM, Donglu S: Effects of CdSe/ZnS quantum dots covered multi-walled carbon nanotubes on murine embryonic stem cells. Nano Biomed Eng 2010, 2: 236–244.Google Scholar
- Cui D, Li Q, Huang P, Wang K, Kong Y, Zhang H, You X, He R, Song H, Wang J, Bao C, Asahi T, Gao F, Osaka T: Real time PCR based on fluorescent quenching of mercaptoacetic acid-CdTe quantum dots for ultrasensitive specific detection of nucleic acids. Nano Biomed Eng 2010, 2: 45–55.Google Scholar
- Kaul Z, Yaguchi T, Kaul SC, Hirano T, Wadhwa R, Taira K: Mortalin imaging in normal and cancer cells with quantum dot immuno-conjugates. Cell Research 2003, 13: 503–507. 10.1038/sj.cr.7290194View ArticleGoogle Scholar
- Rizwan W, You Bing Y, Ahmad U, Surya S, Hwang IH, Hyung-Shik S, Young-Soon K: Platinum quantum dots and their cytotoxic effect towards myoblast cancer cells (C2C12). J Biomed Nanotechnol 2012, 8: 424–431. 10.1166/jbn.2012.1448View ArticleGoogle Scholar
- Zhang X, Li D, Wang C, Zhi X, Zhang C, Wang K, Cui D: A CCD-based reader combined quantum dots-labeled lateral flow strips for ultrasensitive quantitative detection of anti-Hbs antibody. J Biomed Nanotechnol 2012, 8: 372–379. 10.1166/jbn.2012.1401View ArticleGoogle Scholar
- Medintz IL, Uyeda HT, Goldman ER, Mattoussi H: Quantum dot bioconjugates for imaging, labelling and sensing. Nat Mater 2005, 4: 435–446. 10.1038/nmat1390View ArticleGoogle Scholar
- Aslan K: Rapid and sensitive detection of troponin I-T-C complex from human serum using microwave-accelerated metal-enhanced fluorescence. Nano Biomed Eng 2011, 3: 179–183.Google Scholar
- Hoshino A, Fujioka K, Manabe N, Yamaya S, Goto Y, Yasuhara M, Yamamoto K: Simultaneous multicolor detection system of the single-molecular microbial antigen with total internal reflection fluorescence microscopy. Microbiol Immunol 2005, 49: 461–470.View ArticleGoogle Scholar
- Edgar R, McKinstry M, Hwang J, Oppenheim AB, Fekete RA, Giulian G, Merril C, Nagashima K, Adhya S: High-sensitivity bacterial detection using biotin-tagged phage and quantum-dot nanocomplexes. PNAS 2006, 103: 4841–4845. 10.1073/pnas.0601211103View ArticleGoogle Scholar
- Ruan J, Shen J, Song H, Ji J, Wang K, Cui D, Wang Z: Viability and pluripotency studying of human embryo stem cells labeled with quantum dots. Nano Biomed Eng 2010, 2: 245–251.View ArticleGoogle Scholar
- Tian J, Zhou L, Zhao Y, Wang Y, Peng Y, Zhao S: Multiplexed detection of tumor markers with multicolor quantum dots based on fluorescence polarization immunoassay. Talanta 2012, 92: 72–77.View ArticleGoogle Scholar
- Tian J, Zhou L, Zhao Y, Wang Y, Peng Y, Hong X, Zhao S: The application of CdTe/CdS in the detection of carcinoembryonic antigen by fluorescence polarization immunoassay. J Fluoresc 2012, 22: 1571–1579. 10.1007/s10895-012-1097-2View ArticleGoogle Scholar
- Chou PY, Fasman GD: Prediction of the secondary structure of proteins from their amino acid sequence. Adv Enzymol Relat Areas Mol Biol 1978, 47: 45–148.Google Scholar
- Karplus PA, Schulz GE: Prediction of chain flexibility in proteins - a tool for the selection of peptide antigens. Naturwissenschafren 1985, 72: 212–213. 10.1007/BF01195768View ArticleGoogle Scholar
- Kyte J, Doolittle RF: A simple method for displaying the hydropathic character of a protein. J Mol Biol 1982, 157: 105–132. 10.1016/0022-2836(82)90515-0View ArticleGoogle Scholar
- Emini EA, Hughes JV, Perlow DS, Boger J: Induction of hepatitis A virus-neutralizing antibody by a virus-specific synthetic peptide. J Virol 1985, 55: 836–839.Google Scholar
- Jameson BA, Wolf H: The antigenic index: a novel algorithm for predicting antigenic determinants. Comput Appl Biosci 1988, 4: 181–186.Google Scholar
- Weng CC, Peter DW: Fmoc Solid Phase Peptide Synthesis: A Practical Approach. Oxford: Oxford University Press; 2000.Google Scholar
- Yang H, Li D, He R, Guo Q, Wang K, Zhang X, Huang P, Cui D: A novel quantum dots-based point of care test for syphilis. Nanoscale Res Lett 2010, 5: 875–881. 10.1007/s11671-010-9578-1View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.