Development of a direct competitive ELISA for the detection of Mycoplasma bovisinfection based on a monoclonal antibody of P48 protein
© Fu et al.; licensee BioMed Central Ltd. 2014
Received: 3 November 2013
Accepted: 6 February 2014
Published: 18 February 2014
Mycoplasma bovis (M. bovis) is a major, but often overlooked, pathogen documented to cause respiratory disease, mastitis, and arthritis in cattle throughout China since 2008. Here, we report the development of a direct competitive enzyme-linked immunosorbent assay (Dc-ELISA) to detect M. bovis antibody.
We used a recombinant P48 protein and monoclonal antibody (mAb) 10E. MAb 10E, prepared against the recombinant P48 protein of M. bovis, identified all M. bovis strains with no cross-reactivity with other related pathogens. Coating micro plates with P48 protein instead of whole M. bovis cells as well as the use of mAb 10E produced a specific and sensitive Dc-ELISA for M. bovis antibody detection with a cut-off percent inhibition (PI) value of 32%. Compared with two commercial indirect ELISA (i-ELISA) kits, our Dc-ELISA offered a higher positive detection rate in 165 clinical bovine serum samples.
A rapid, sensitive, and reliable serological diagnosis method was developed for M. bovis, which can facilitate M. bovis surveillance, assisting researchers in understanding the ecology and epidemiology of M. bovis.
KeywordsDirect competitive ELISA Mycoplasma bovis Monoclonal antibody
M. bovis is a major, but often overlooked, pathogen causing respiratory disease, mastitis, and arthritis in cattle. First isolated from the milk of mastitic cow in 1961 in US, this bacteria appears to have spread across Ireland, South America, and Europe . In China, the first M. bovis pneumonia case was reported in Hubei province in 2008, and the pathogen genome was sequenced in 2011 [2, 3]. Until now, M. bovis pneumonia cases, mostly in cattle, have been reported in more than ten Chinese provinces. And it often outbreaks in calves after transport. The mAbs used for M. bovis detection were reported previously, and most of them were selected against whole M. bovis cells [4, 5] or M. bovis variable surface lipoproteins (Vsps) that undergo high-frequency phase and size variations .
P48 protein is an immunodominant invariable lipoprotein exposed on the membrane of M. agalactiae. The M. bovis P48 protein is homologous to that of M. agalactiae, which was first studied in 2005. A single protein band was observed in rabbit antiserum raised against P48 protein of M. agalactiae in a M. bovis reference strain PG45 and 10 field isolates. An Antibody response against M. bovis recombinant P48 protein has been detected both in experimentally and naturally infected animals, suggesting a stable expression of the corresponding genes. Over all, P48 protein is a useful marker for M. bovis infection and an alternative candidate for the development of specific serological test for M. bovis.
Here, we developed a direct competitive ELISA (Dc-ELISA) to detect M. bovis specific antibody in serum based on a P48 protein mAb 10E. This Dc-ELISA showed a higher positive rate than two commercial i-ELISA kits in detection of 165 Chinese clinical bovine serum samples.
Expression of P48 protein in E. coli
Selection and characterization of mAb against P48 protein
To select mAbs with the desired affinity, hybridomas were screened by i-ELISA, and positive cells were selected and subcloned at least three times using a limiting dilution method. MAbs isotypes, detected using the commercial ELISA kit protocol, were all lgG1 (data not shown). One monoclone (from 6 monoclones) designated 10E, which could produce an antibody that specifically recognized P48 protein, was selected to produce and purify mAb from mouse ascetic fluid of mice for further characterization.
Determination of cut-off value and specificity of Dc-ELISA
To confirm Dc-ELISA specificity, Infectious bovine rhinotracheitis virus (IBRV), bovine viral diarrhea virus (BVDV), bovine parainfluenza virus type 3 (BPIV3), and bovine respiratory syncytial virus (BRSV) positive control serum, rabbit hyperimmune antiserum against Mycoplasma bovirhinis (M. bovirhinis), Mycoplasma arginine (M. arginine), and M. agalactiae were used in this study. As shown in Figure 4B, a PI level lower than 32% was considered as negative, suggesting that the Dc-ELISA was specific for detecting antibody against M. bovis.
Comparison among Dc-ELISA, two commercial i-ELISA kits, and Dot blot
Results of 165 bovine serum samples for M. bovis antibody detection by Dc-ELISA, Biovet, Bio-X and Dot blot
Positive rates of serum samples from different provinces in China detected by four methods
A simple, rapid, specific and sensitive diagnostic technique is preferred for intensive surveillance of animal infectious diseases such as M. bovis. Fortunately M. bovis grows well in a variety of media and produces “centered” colonies within 3-5 days, enabling the culture of these pathogens for study and assay development . Ball reported that a sandwich ELISA using monoclonal antibodies for M. bovis antigen has similar sensitivity to that of conventional culture diagnosis , and he also reported real-time PCR was better for detecting M. bovis than sandwich ELISA, which suggested that real-time PCR could replace the sandwich ELISA for M. bovis detection . Conventional PCR assays are developed based the 16S rRNA gene [12, 13], but some researchers suggested that the PCR-based uvrC gene is more specific [14, 15]. In situ hybridization and immunohistochemistry [16, 17] also have been used to identify M. bovis.
Many serological tests including indirect hemagglutination and film inhibition have been reported, but these methods were inferior to i-ELISA using whole cell or chemically treated antigens [1, 18, 19]. And, i-ELISAs based on recombinant VspA, membrane lipoprotein P48 protein [8, 20] or glyceraldehyde-3-phosphate dehydrogenase (GAPDH)  have been developed. In a competitive/blocking ELISA, immunoglobulins from positive serum samples compete to inhibit the mAb from binding to its specific binding site on antigens and subsequently prevent color development. In contrast, non-reactive sera produce a strong colored reaction . Competitive/blocking ELISAs have been confirmed to be significantly sensitivity and specific for detecting antibody to Bluetongsue Virus, West Nile virus, and Francisella tularensis, and even a higher sensitivity than i-ELISA for detection to Mycoplasma meleagridis and PRRSV . A blocking ELISA with a mAb to a certain antigen and the whole-cell lysate of M. bovis as the target antigen is a high specific with a good sensitivity, low background, and no cross-reactivity .
Here, we report the development of a Dc-ELISA with a mAb 10E against P48 protein and P48 protein as the coated-antigen. MAb 10E was quite specific for all eight M. bovis strains and did not cross-react with other related pathogens as evidenced by Western blot (Figure 3). Thus, we concluded that a Dc-ELISA with micro plates coated with purified recombinant P48 protein was more specific than those coated with whole-cell lysates of M. bovis. Antisera of other pathogens had PI values less than 32%, indicating that our Dc-ELISA was more specific for the detection of antibodies against M. bovis (Figure 4B). Furthermore, the mAb 10E was conjugated with HRP, which replaced the use of HRP-conjugated goat-anti-bovine IgG and the procedure steps were simplified.
A total of 165 clinical bovine serum samples (Table 2) were assayed, and Dc-ELISA (93/165) detected more samples than a Biovet commercial i-ELISA kit (79/165) or a Bio-X kit (65/165). Data from the Dc-ELISA suggested a relatively wide serological population of M. bovis in the studied provinces (more than 43.9% by Dc-ELISA), especially in Shandong province (75.8%). Simultaneously, PCR detections of M. bovis has been done with the nasal swab samples from Shandong province using MB-F/R primers. In these 33 samples from Shandong, 25 samples were serologically positive by Dc-ELISA whereas 27 samples were etiologically positive, including all the 25 serologically positive ones (data not shown). These results showed that PCR was more sensitive than Dc-ELISA, a finding that agreed with Ali Ghadersohi’s report . Generally, during early stages of microbiological infection, less antibodies are produced, so PCR pathogen detection is better than serological detection for the evaluation of infection. In contrast Dc-ELISA is reliable when antibiotics have been used to treat an infection and Dc-ELISA is simpler and more suitable for high-throughput detection. The Dc-ELISA method we we developed holds promise for facilitating surveillance of M. bovis, ultimately leading to a better understanding of the ecology and epidemiology of M. bovis.
A Dc-ELISA based on mAb 10E, which was prepared against P48 protein and capable of identifying all M. bovis strains with no cross-reactivity to related pathogens, was developed to detect M. bovis infection. With a PI cut-off value of 32%, the Dc-ELISA had higher positive detection rates than two commercial i-ELISA kits when they were used to detect 165 clinical bovine serum samples, which suggested that Dc-ELISA represented an improved and simplified method for surveillance of M. bovis.
Mycoplasma and virus strains
Various Mycoplasma and Virus species used in this study
Field isolates (GY-2, PD-2, HRB-1, HY-1, HG-1)c
Twenty negative bovine serum samples were collected from cattle without a history of M. bovis infection. Twenty positive bovine serum samples were collected from naturally infected M. bovis cattle, from which M. bovis had been isolated nasal swabs using PPLO medium. Bovine BADV3, IBRV, BVDV, and BPIV3 positive control sera were purchased from Real Bio-technology (Qingdao, China). Rabbit hyperimmune antisera against M. bovirhinis, M. arginine and M. agalactiae were prepared as per standard techniques . Here 165 clinical bovine serum samples were collected from five provinces in China (Beijing, Hebei, Tianjin, Shandong and Heilongjiang) and stored at -20°C.
Preparation of recombinant P48 protein
Primer sets for identification and mutation of P48 gene in M. bovis strain Hubei-1
Identify P48 gene
Mutate point 192
Mutate point 648
Mutate point 1275
Mutate point 1317
Mutate point 699
The mutated gene was cloned into the pET28a (+) expression vector. Transetta (DE3) Chemically Competent Cell (Transgen, Beijing, China) was used to express P48 protein. Total cell lysates were obtained by sonicating, and P48 protein was recovered from the soluble fraction and purified using His · Bind kits (Novagen, Darmstadt, Germany). Purity and yield of the recombinant P48 protein was evaluated by SDS-PAGE and the Bradford method .
Preparation of MAb against P48 protein
Female BALB/c mice (8 weeks-of-age) were purchased from the Meiliyaweitong Experimental Animal Technology (Beijing, China). All animal research was approved by the Beijing Association for Science and Technology (approval ID SYXK (Beijing) 2007-0023) and complied with the guidelines of Beijing Laboratory Animal Welfare and the Ethics of the Beijing Administration Committee of Laboratory Animals. All animal studies were also performed in accordance with the China Agricultural University Institutional Animal Care and Use Committee guidelines (ID: SKLAB-B-2010-003) and approved by the Animal Welfare Committee of China Agricultural University. The mice were first inoculated subcutaneously(sc) with 100 μg of purified P48 protein particles in complete Freund’s adjuvant and three weeks later, animals were immunologically boosted with the same dose of P48 protein (sc injection) in incomplete Freund’s adjuvant. Another boost was administered with 25 μg (sc) of P48 protein in incomplete adjuvant after three weeks. After two weeks, the last boost was injected intraperitoneally with 25 μg P48 protein in PBS buffer. Three days later, spleen cells of the immunized mice were fused with SP2/0 cells using 50% (w/v) polyethylene glycol 1000 (PEG1000) (Sigma Chemical Co., Beijing, China). Hybridoma cells were selected by HAT media (Sigma Chemical Co., Beijing, China) supplemented with 20% fetal bovine serum (FBS) (Hyclone Co., Beijing, China). The culture supernatants were screened for antibodies against P48 protein by i-ELISA as described below. Single-cell cloning was carried out three times in positive hybridomas using a limiting dilution method until monoclones were obtained.
The isotypes of the mAb were also measured by testing the culture supernatants using a commercial ELISA kit (SouthernBiotech, Birmingham, AL). Capture ELISA was performed according to the manufacturer’s protocol.
MAb 10E lgG1 was purified from mouse ascetic fluid, cultured as hybridomas in vivo and conjugated with horseradish peroxidase (HRP) (Sigma Chemical Co., Beijing, China) performed according to standard protocol [30, 31]. Conjugated mAb 10E (HRP-10E) was stored with 50% glycerol at -20°C.
I-ELISA for screening of mAb
For screening of mAb 10E, 96-well micro plates were coated with 100 ng/well of P48 protein in 0.05 M carbonate-bicarbonate buffer (pH 9.6) at 4°C overnight. After washing three times with PBS containing 0.05% Tween20 (PBST) and blocking with 200 μL 5% non-fat dried milk (Sigma Chemical Co., Beijing, China) in PBST at 37°C for 2 h, the plates were then incubated with culture supernatant at 37°C for 1 h, and washed three times with PBST. HRP-conjugated goat-anti-mouse lgG (Kpl, Gaithersburg, TN), diluted 5,000 times, were incubated at 37°C for 1 h and then washed with PBST four times. Substrate 3, 3′, 5, 5′-tetramethylbenzidine (TMB) was added to each well for visualization for at 37°C 10 min, and then 50 μL/well of 2 M sulfuric acid was added to stop the reaction was stopped. The optical density was read at 450 nm  in a micro plate reader (Bio-Rad, Hercules, USA).
Cross-reactivity and specificity of mAb 10E
Western blot was used to measure the cross-reactivity of mAb 10E. Twenty micrograms of whole cell lysates of M. bovis, M. bovirhinis, M. arginine, M. agalactiae, IBRV, BADV3, BPIV3, and BVDV were analyzed by 12% SDS-PAGE, transferred to PVDF membranes, and blocked with 5% non-fat dried milk at room temperature for 4 h on a rocker. After incubation with 1:5,000 purified mAb 10E at 4°C overnight, the membranes were washed three times with PBST and incubated with 1:5,000 HRP-conjugated goat-anti-mouse lgG at room temperature for 30 min. Then washed the membranes four times with PBST, added ELC plus substrate (Applygen, Beijing, China), placed x-ray film on the top of the membranes, and exposed in dark room with safe light.
Establishment of Dc-ELISA
Micro plates (96-well) were coated with approximately 400 ng/well purified P48 protein in 0.05 M carbonate-bicarbonate buffer (pH 9.6) at 4°C overnight and washed with PBST and blocked with 200 μl 10% horse serum at 37°C for 2 h. Fifty microlitre non-diluted bovine serum (two wells were added 50 μL PBST as mAb control) and 50 μL 1:250 mAb HRP-10E were added. After incubation at 37°C for 1 h, all unbound materials were removed by washing with PBST four times. The colorimetric reaction was achieved as described above. The percent inhibition (PI) values were determined using the formula: PI (%) = (1- OD450nm of test serum/OD450nm of monoclonal control) × 100%. Twenty negative bovine sera were adopted to determine the PI cut-off value, which was designed as the mean PI of 20 negative sera + 2 standard deviations (SD), which would ensure that 95% of PI values for the negative sera fall within this range.
Dot blot assay
Dot blot assay was carried out in this study to detect the antibody to M. bovis and compare these data with Dc-ELISA. Briefly, M. bovis whole cell lysates (20 μg) and PBS (negative control) were spotted onto nitrocellulose membranes and allowed to air dry. Then membranes were blocked with 10% goat sera in DB buffer (20 mM Tris, 0.5 M NaCl, 0.055% Tween80, pH 7.5) at room temperature for 4 h and incubated with 1:400 bovine sera at 4°C overnight. After washing three times with DB buffer, the membranes were incubated with 1:5,000 HRP-conjugated goat-anti-bovine IgG at room temperature for 30 min, and then washed the membrane four times with DB buffer, added ELC plus substrate, placed x-ray film on top of membranes, and exposed in dark room with safe light.
Comparisons of Dc-ELISA, two commercial i-ELISA kits, and Dot blot
A total of 165 bovine serum samples from five farms in different provinces of China were detected using Dc-ELISA, two commercial i-ELISA kits from Bio-X (Jemelle, Belgium) and Biovet (Saint-Hyacinthe, Canada), and a Dot blot assay. The i-ELISAs were performed as recommended by the manufacturer.
SPSS v19.0 software was used to analyze the normality test of of the negative bovine sera PI values using a Shapiro-Wilk test and the degrees of agreement between Dc-ELISA and two commercial kits, and Dot blot by kappa statistics.
We acknowledged the participation and assistance of Huiling Chen, Wenqiang Gan, Binbin Cao for providing the bovine serums. This work was supported by Agricultural Finance Program, Ministry of Agriculture of China and Program for New Century Excellent Talents in University of Ministry of Education of China.
- Nicholas R, Ayling R: Mycoplasma bovis: disease, diagnosis, and control. Res Vet Sci. 2003, 74 (2): 105-10.1016/S0034-5288(02)00155-8.View ArticlePubMed
- Xin J, Li Y, Guo D, Song N, Hu S, Chen C, Pei J, Cao P: First isolation of Mycoplasma bovis from calf lung with pneumoniae in China. Chin J Prev Vet Med. 2008, 30 (9): 661-664.
- Qi J, Guo A, Cui P, Chen Y, Mustafa R, Ba X, Hu C, Bai Z, Chen X, Shi L: Comparative geno-plasticity analysis of Mycoplasma bovis HB0801 (Chinese isolate). PLoS ONE. 2012, 7: 8239.
- Dénes B, Tenk M, Tekes L, Varga I, Ferenczné IP, Stipkovits L: Recognition of multiple Mycoplasma bovis antigens by monoclonal antibodies. Hybrid Hybridomics. 2003, 22 (1): 11-16. 10.1089/153685903321538035.View ArticlePubMed
- Ghadersohi A, Fayazi Z, Hirst R: Development of a monoclonal blocking ELISA for the detection of antibody to Mycoplasma bovis in dairy cattle and comparison to detection by PCR. Vet Immunol Immunopathol. 2005, 104 (3–4): 183.View ArticlePubMed
- Sachse K, Helbig J, Lysnyansky I, Grajetzki C, Müller W, Jacobs E, Yogev D: Epitope mapping of immunogenic and adhesive structures in repetitive domains of Mycoplasma bovis variable surface lipoproteins. Infect Immun. 2000, 68 (2): 680-687. 10.1128/IAI.68.2.680-687.2000.PubMed CentralView ArticlePubMed
- Rosati S, Pozzi S, Robino P, Montinaro B, Conti A, Fadda M, Pittau M: P48 major surface antigen of Mycoplasma agalactiae is homologous to a malp product of Mycoplasma fermentans and belongs to a selected family of bacterial lipoproteins. Infect Immun. 1999, 67 (11): 6213-6216.PubMed CentralPubMed
- Robino P, Alberti A, Pittau M, Chessa B, Miciletta M, Nebbia P, Le Grand D, Rosati S: Genetic and antigenic characterization of the surface lipoprotein P48 of Mycoplasma bovis. Vet Microbiol. 2005, 109 (3–4): 201.View ArticlePubMed
- Nicholas R, Baker S: Recovery of mycoplasmas from animals. In Mycoplasma Protocols Methods. Springer; 1998:37-43.View Article
- Ball H, Finlay D, Reilly G: Sandwich ELISA detection of Mycoplasma bovis in pneumonic calf lungs and nasal swabs. Vet Rec. 1994, 135 (22): 531-532. 10.1136/vr.135.22.531.View ArticlePubMed
- Bell C, Blackburn P, Patterson I, Ellison S, Ball H: Real-time PCR demonstrates a higher prevalence of Mycoplasma bovis in Northern Ireland compared with sandwich ELISA. Vet Rec. 2012, 171 (16): 402.View ArticlePubMed
- Ayling R, Nicholas R, Johansson K: Application of the polymerase chain reaction for the routine identification of Mycoplasma bovis. Vet Rec. 1997, 141 (12): 307-308. 10.1136/vr.141.12.307.View ArticlePubMed
- Cremonesi P, Vimercati C, Pisoni G, Perez G, Ribera AM, Castiglioni B, Luzzana M, Ruffo G, Moroni P: Development of DNA extraction and PCR amplification protocols for detection of Mycoplasma bovis directly from milk samples. Vet Res Commun. 2007, 31: 225-227. 10.1007/s11259-007-0011-x.View ArticlePubMed
- Hotzel H, Sachase K, Pfützner H: Rapid detection of Mycoplasma bovis in milk samples and nasal swabs using the polymerase chain reaction. J Appl Microbiol. 1996, 80 (5): 505-510.
- Pinnow C, Butler J, Sachse K, Hotzel H, Timms L, Rosenbusch R: Detection of Mycoplasma bovis in preservative-treated field milk samples. J Dairy Sci. 2001, 84 (7): 1640-10.3168/jds.S0022-0302(01)74599-7.View ArticlePubMed
- Björn J, Hermeyer K, Jechlinger W, Zimmermann M, Spergser J, Rosengarten R, Marion H-T: In situ hybridization for the detection of Mycoplasma bovis in paraffin-embedded lung tissue from experimentally infected calves. J Vet Diagn Investig. 2010, 22 (1): 90-93. 10.1177/104063871002200117.View Article
- Hermeyer K, Peters M, Brügmann M, Jacobsen B, Hewicker-Trautwein M: Demonstration of Mycoplasma bovis by immunohistochemistry and in situ hybridization in an aborted bovine fetus and neonatal calf. J Vet Diagn Investig. 2012, 24 (2): 364-369. 10.1177/1040638711435145.View Article
- Byrne W, McCormack R, Ball H, Brice N, Baker S, Ayling R, Nicholas R: Application of an indirect ELISA to milk samples to identify cows with Mycoplasma bovis mastitis. Vet Rec. 2000, 146 (13): 368-369. 10.1136/vr.146.13.368.View ArticlePubMed
- Le Grand D, Bézille P, Calavas D, Poumarat F, Brank M, Citti C, Rosengarten R: Serological prevalence of Mycoplasma bovis infection in suckling beef cattle in France. Vet Rec. 2002, 150 (9): 268-273. 10.1136/vr.150.9.268.View ArticlePubMed
- Caswell JL, Archambault M: Mycoplasma bovis pneumonia in cattle. Anim Health Res Rev. 2008, 8 (2): 161.View Article
- Perez Casal J, Prysliak T: Detection of antibodies against the Mycoplasma bovis glyceraldehyde-3-phosphate dehydrogenase protein in beef cattle. Microb Pathog. 2007, 43 (5–6): 189.View ArticlePubMed
- Reddington J, Reddington G, MacLachlan N: A competitive ELISA for detection of antibodies to the group antigen of bluetongue virus. J Vet Diagn Investig. 1991, 3 (2): 144-147. 10.1177/104063879100300207.View Article
- Hirota J, Shimizu S: A new competitive ELISA detects West Nile virus infection using monoclonal antibodies against the precursor-membrane protein of West Nile virus. J Virol Methods. 2013, 188 (1-2): 132-138. 10.1016/j.jviromet.2012.12.002.View ArticlePubMed
- Sharma N, Hotta A, Yamamoto Y, Fujita O, Uda A, Morikawa S, Yamada A, Tanabayashi K: Detection of Francisella tularensis-specific antibodies in patients with tularemia by a novel competitive enzyme-linked immunosorbent assay. Clin Vaccine Immunol. 2013, 20 (1): 9-16. 10.1128/CVI.00516-12.PubMed CentralView ArticlePubMed
- Dufour-Gesbert F, Kempf I, Kobisch M: Development of a blocking enzyme-linked immunosorbent assay for detection of turkey antibodies to Mycoplasma meleagridis. Vet Microbiol. 2001, 78 (3): 275-284. 10.1016/S0378-1135(00)00280-7.View ArticlePubMed
- Ferrin NH, Fang Y, Johnson CR, Murtaugh MP, Polson DD, Torremorell M, Gramer ML, Nelson EA: Validation of a blocking enzyme-linked immunosorbent assay for detection of antibodies against porcine reproductive and respiratory syndrome virus. Clin Diagn Lab Immunol. 2004, 11 (3): 503-514.PubMed CentralPubMed
- Schunk MK, Macallum GE: Applications and optimization of immunization procedures. ILAR J. 2005, 46 (3): 241-257. 10.1093/ilar.46.3.241.View ArticlePubMed
- Villarroel A, Regalado M: A fast and simple method to introduce multiple distant point mutations. Tech Tips Online. 1997, 2 (1): 24-26. 10.1016/S1366-2120(08)70021-7.View Article
- Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976, 72 (1): 248-254.View ArticlePubMed
- Wilson M: Recent development in the periodate method of conjugating horseradish peroxidase (HRPO) to antibodies. Immunofluorescence and Related Staining Techniques. 1978, 215-224.
- Hendriksen C, De Leeuw W: Production of monoclonal antibodies by the ascites method in laboratory animals. Res Immunol. 1998, 149 (6): 535-542. 10.1016/S0923-2494(98)80002-3.View ArticlePubMed
- Chusri M, Wongphanit P, Palaga T, Puthong S, Sooksai S, Komolpis K: Production and characterization of a monoclonal antibody against enrofloxacin. J Microbiol Biotechnol. 2013, 23 (1): 69-75. 10.4014/jmb.1201.01017.View ArticlePubMed
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 credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.