- Open Access
Development of real time PCR for detection and quantitation of Dengue Viruses
© Gurukumar et al; licensee BioMed Central Ltd. 2009
- Received: 12 September 2008
- Accepted: 23 January 2009
- Published: 23 January 2009
Dengue virus (DENV), a mosquito borne flavivirus is an important pathogen causing more than 50 million infections every year around the world. Dengue diagnosis depends on serology, which is not useful in the early phase of the disease and virus isolation, which is laborious and time consuming. There is need for a rapid, sensitive and high throughput method for detection of DENV in the early stages of the disease. Several real-time PCR assays have been described for dengue viruses, but there is scope for improvement. The new generation TaqMan Minor Groove Binding (MGB) probe approach was used to develop an improved real time RT-PCR (qRT-PCR) for DENV in this study.
The 3'UTR of thirteen Indian strains of DENV was sequenced and aligned with 41 representative sequences from GenBank. A region conserved in all four serotypes was used to target primers and probes for the qRT-PCR. A single MGB probe and a single primer pair for all the four serotypes of DENV were designed. The sensitivity of the two step qRT-PCR assay was10 copies of RNA molecules per reaction. The specificity and sensitivity of the assay was 100% when tested with a panel of 39 known positive and negative samples. Viral RNA could be detected and quantitated in infected mouse brain, cell cultures, mosquitoes and clinical samples. Viral RNA could be detected in patients even after seroconversion till 10 days post onset of infection. There was no signal with Japanese Encephalitis (JE), West Nile (WN), Chikungunya (CHK) viruses or with Leptospira, Plasmodium vivax, Plasmodium falciparum and Rickettsia positive clinical samples.
We have developed a highly sensitive and specific qRT-PCR for detection and quantitation of dengue viruses. The assay will be a useful tool for differential diagnosis of dengue fever in a situation where a number of other clinically indistinguishable infectious diseases like malaria, Chikungunya, rickettsia and leptospira occur. The ability of the assay to detect DENV-2 in inoculated mosquitoes makes it a potential tool for detecting DENV in field-caught mosquitoes.
- Dengue Virus
- Japanese Encephalitis
- Japanese Encephalitis
- Minor Groove Binder
Dengue virus (DENV) is a mosquito borne flavivirus with four serotypes DENV-1 to 4. The global prevalence of dengue has grown dramatically in the recent decades. The disease is now endemic in more than 100 countries around the globe and it is estimated that DENV causes 50 to 100 million cases of acute febrile disease every year . Since 1945, outbreaks of dengue caused by all 4 serotypes have been reported regularly from different regions of India .
Dengue is diagnosed by either detecting virus or antibody to the virus in blood. Isolation of virus in cell culture or in infant mouse brain remains the gold standard for diagnosis, but it takes more than a week for the test to be completed making it impractical in most situations. Detection of anti-dengue IgM and IgG in the serum by ELISA is the most commonly used criteria for presumptive diagnosis of DENV infections. These serological methods are unable to detect the infection during the early phase of the disease. Thus there is a need for rapid and sensitive methods for detection of DENV early in the course of infection for better patient management. Several PCR based methods for detecting DENV nucleic acid in the serum have been reported, the most widely used test is the nested RT-PCR developed by Lanciotti et al.,  and a modification of the same method to single tube format by Harris et al., . More recently real-time PCR based methods have been reported for detection and serotyping of DENV which use fluorescent based reporter chemistries [5–8]. Real-time PCR has many advantages over conventional RT-PCR, in that it is more sensitive, can be automated thereby enabling high throughput screening, and the hands on time including sample handling is less than four hours. The real-time PCR is also used to quantitate the viral load in blood samples, making it a useful tool to investigate the role of viremia in pathogenesis of dengue. Another important aspect of dengue disease is the surveillance of vector population and detection of DENV in field caught mosquitoes. Real-time PCR because of its high sensitivity could be useful in such surveillance and provide early warning of a possible outbreak of the disease. Recent reports on DENV group specific real-time PCR used SYBR green based method [9, 10] where as an earlier report based on TaqMan probes used multiple probe primer sets for detection of all four serotypes of DENV . In the present study, we describe the development of a DENV-specific TaqMan based real-time PCR for detection and quantitation of all four serotypes using a single probe primer set targeted against the 3'UTR of DENV.
Viruses used in the study
Year and location
1962, Vellore, India
1963, Vellore, India
1963, Vellore, India
2005, Pune, India
1960, Vellore, India
1980, Kolkata, India
1993, Pune, India
2005, Pune, India
Rockefeller lab, USA
2005, Kolkata, India
2005, Kolkata, India
1961, Vellore, India
1962, Vellore, India
1963, Vellore, India
1964, Vellore, India
1958, Vellore, India
1950, Cairo, Egypt
2006, Andhra Pradesh, India
Infection and maintenance of cell cultures
Vero cells were maintained in MEM supplemented with 10% FCS. One strain of each serotype was amplified in Vero cells to make virus stocks for determining standard curves. Cell cultures were infected at 0.1 multiplicity of infection (MOI) with DENV-1 (16007), DENV-2 (803347), DENV-3 (059826) or DENV-4 (642069). The virus was harvested after 4 to 5 days post infection after appearance of cytopathic effect. The infected cell cultures were subjected to a single freeze thaw. The cell lysates were clarified by centrifugation at 2000 rpm and stored in suitable aliquots at -80°C until used.
Infection of mosquitoes
Female Aedes aegypti mosquitoes were inoculated with 103 PFU/mL of DENV-2 (803347) by intrathoracic route. Infected mosquitoes were kept for 14 days at 28°C and 80–90% humidity. Surviving mosquitoes were frozen at -80°C. The mosquito heads were used for Immuno Fluorescence Assay (IFA) as described earlier  and the bodies were used for RNA extraction. A total of 34 mosquitoes including 10 control mosquitoes and 24 mosquitoes inoculated with DENV-2, were tested by IFA and qRT-PCR.
Three hundred and eight blood samples were used to evaluate the usefulness of the qRT-PCR assay for its diagnostic potential. The blood samples were obtained within 10 days post onset of symptoms. The sera were separated, aliquoted and stored at -80°C. All samples were tested for presence of DENV RNA by nested RT-PCR as described earlier  and dengue specific IgM antibodies using the NIV MAC-ELISA kit . Serum samples positive for other febrile illnesses i.e. Leptospira (n = 2), Rickettsia (n = 1), and Malaria (n = 2) were provided by Dr. RR. Gadia of King Edward Memorial Hospital, Pune India. Paired serum samples (n = 4), which were confirmed to be dengue by a four fold rise in HI titre  were also included to analyse the sensitivity of the qRT-PCR. Serum samples from acute cases of CHK and JE were not available therefore sera from healthy individuals were spiked with 105 PFU of CHKV or 103 PFU of JEV and used to evaluate the specificity of the qRT-PCR.
Primer and probe design
The 3'UTR of four strains of DENV-1 (nt 10230–10700), four strains of DENV-2 (nt 10301–10708), two strains of DENV-3 (nt 10243–10689), and three strains of DENV-4 (nt 10315–10635) isolated in India (Table 1) were sequenced using the big dye terminator kit (Applied Biosystems, Foster city, CA, USA). These 13 sequences were aligned with sequences of 11 DENV-1 strains, 11 DENV-2 strains, 13 DENV-3 strains and 6 DENV-4 strains from GenBank using the Clustal W programme. A stretch of nucleotides conserved in the four serotypes was identified and the primers and probe sequences were designed using the Primer Express software.
Generation of RNA standard for the qRT-PCR
The target region from the 3'UTR was amplified from DENV-3 (633798) and cloned in to the TEasy cloning vector (Promega Corporation, Madison, WI, USA). The presence and orientation of the insert DNA was confirmed by sequencing. The plasmid was linearized by digestion with ApaI and the target sequence was amplified using the in vitro RNA transcription kit (Roche Diagnostics, IN, USA). The transcribed RNA was treated with DNAse to digest the plasmid and purified using the QIAamp RNA purification kit (Qiagen Sciences, Valencia, CA, USA). The RNA was quantified by spectrophotometry. The copy numbers of the RNA was calculated based on the concentration and its molecular weight and 10 fold serial dilutions of this RNA from 102 to 108 copies per reaction was used as standard in all qRT-PCRs.
RNA from 140 μl of 10% mouse brain suspension, cell culture lysate or human serum samples was extracted using the QIAamp viral RNA extraction kit (Qiagen Sciences, Valencia, CA, USA) as per the manufacturer's protocol. The RNA was eluted in 60 μl of elution buffer and stored at -80°C. 10 μL of extracted RNA was used in all qRT-PCR. In case of mosquitoes, the entire mosquito was homogenized and the RNA was extracted from the homogenate using TRIZOL reagent (Invitrogen, Carlsbad, CA, USA) as per the manufacturer's protocol. The RNA was resuspended in 20 μl of RNAse free distilled water and stored at -80°C. 0.5 μL of this RNA was used in the qRT-PCR. The transcribed or viral RNA was reverse transcribed using the reverse primer with AMV reverse transcriptase (Promega Corporation, Madison, WI, USA). The reverse transcription reaction was carried out at 42°C for 1 h. The cDNA thus obtained was used as the template in the qPCR. The TaqMan universal PCR master mix (Applied Biosystems, Foster city, USA) was used in all qPCRs. Each reaction had 200 nM of forward primer, 250 nM of probe and 300 nM of reverse primer in a 25 μL final reaction volume. The PCR mixtures were pre incubated at 50°C for 2 min followed by denaturation at 95°C for 10 min and 45 cycles of 95°C for 15 sec and 60°C for 1 min using the Applied Biosystems 7500 real-time PCR system. The real-time data was analyzed using the SDS software provided by Applied Biosystems.
Design and evaluation of primers and probes
Nucleotide sequence of primers and probe used in the qRT-PCR assay
TaqMan MGB Probe
Sensitivity and Specificity of the qRT-PCR assay
RNA extracted from fourteen DENV strains (Table 1) obtained from the Institute's virus repository and fifteen serum samples which were confirmed to be positive for DENV by virus isolation (n = 11) or by a four fold rise in HI titre in paired serum samples (n = 4) were used to evaluate the sensitivity of the qRT-PCR assay. The assay could detect DENV RNA in all the DENV positive samples indicating that the test was 100% sensitive in detecting DENV RNA [(number of positive specimens/(number of positive specimens + number of false negative specimens)].
The specificity of the assay was evaluated using RNA extracted from virus stocks of two related Flaviviruses (JE and WN), and CHK virus, which causes a dengue like disease and is transmitted by the same vector. In addition RNA extracted from serum samples spiked with JE or CHK viruses and from sera of confirmed leptospira (n = 2), rickettsia (n = 1) and malaria (n = 2) cases were also used for determining the specificity of the assay. None of the 10 samples tested by the qRT-PCR assay showed positive amplification, suggesting that the test was 100% specific for the detection of DENV [number of negative specimens/(number of negative specimens + number of false positive specimens)].
Detection limit of the qRT-PCR assay for the four serotypes of DENV
Detection limit (PFU)
1.3 × 103
4.2 × 106
1.4 × 104
Detection and Quantitation of DENV in human samples
Several real-time PCR based methods for detection of DENV have been reported in the last decade. These assays have targeted the 3'UTR [9, 15], NS5 [11, 16], core  and the envelope  gene sequences. Majority of the recent reports describe development of the serotype specific real-time PCR for dengue using TaqMan probes [8, 18] or FRET probes . Though these methods are useful for serotyping of DENV they may not be cost-effective for routine diagnosis as only a small percentage of samples are positive for DENV RNA during the non endemic season and during the active transmission season only about 50% of samples may be positive for DENV RNA . A group specific PCR would be a useful tool for initial screening of the samples and only those samples positive for DENV can then be subjected to serotyping. Lai et al.,  initially screened samples by a SYBR green based group specific real-time PCR and then serotyped the positive samples by a duplex or a fourplex TaqMan based assay, thereby reducing the operation cost on diagnosis by half. The SYBR green-based detection systems have the major disadvantage of false positives due to dye binding to primer dimers or to DNA amplified non-specifically. Melt curve analyses is often used to confirm the fidelity of the reaction. Other group specific real time PCR assays reported are also SYBR green based [9, 10, 19]. The first DENV group specific assay using TaqMan probes was described by Callahan et al., . In the assay the authors have used multiple probe primer sets for establishing a group specific RT-PCR assay. This could have been because of lack of a suitable stretch of conserved nucleotides for designing the conventional TaqMan probes, earlier observed by Laue et al., . The new generation MGB probes can be much shorter in length because of the Minor groove binder tagged to the probe, which increases the thermal stability of the probes. We used the TaqMan MGB probe to develop the group specific assay in this report which made it possible to use the short conserved region in the 3'UTR of the dengue genome to design the assay using a single probe and primer pair combination. The fluorescent probes in TaqMan assay are known to be target specific and sensitive to mismatches . To avoid false negative results due to mismatches in the probe binding sites, the 3'UTR region of thirteen Indian strains, representing all four serotypes, isolated from different regions in India were sequenced. These sequences were aligned with representative sequences of all four serotypes from the GenBank and a conserved region of 100 nucleotides was used to design the primers and probe.
All four serotypes of DENV from infected mouse brain or cell cultures or mosquitoes or clinical samples could be detected by the qRT-PCR. No amplification with the related Flaviviruses or with samples from other febrile illness was observed. The sensitivity of the qRT-PCR in terms of PFU was highest for DENV-2 at 0.001 PFU followed by DENV-1 and DENV-3 at 0.01 PFU and DENV-4 had the lowest sensitivity of 1 PFU. The difference in sensitivity could have been due to a difference in the proportion of non-infectious RNA transcripts to infectious particles. The mismatches between the DENV-4 virus sequence and the reverse primer could have contributed to the lowest sensitivity of 1 PFU. An alternate reverse primer with DENV-4 sequence can be included to bring the sensitivities at par. The qRT-PCR assay was able to detect viral RNA in a significantly larger number of clinical samples (30.7%) than RT-PCR (16.9%) (p < 0.001, Mc Nemar's test). The qRT-PCR also detected DENV RNA in a larger number of IgM positive samples when compared to the nested RT-PCR. There is increasing recognition of the importance of viral burden in the pathogenesis of DHF and a direct association between viral load and severity of the disease has been reported [21–23] In earlier studies on viral load the method used to quantitate the virus in the blood samples were either mosquito infectious dose (MID50) or PFU. It has been difficult to compare any two studies as the measures vary depending upon the strain of the virus and the substrate used i.e. mosquito or cell line. None of the group specific real-time PCRs reported [6, 9–11, 24] developed a quantitative RNA standard. We developed an RNA standard for quantitating DENV RNA in this study. The RNA copy number, which offers minimum variability, was used to quantitate the viral load in human clinical samples. Viremia was found to be higher in the initial days of the illness decreasing gradually during the later stages of the infection. This result is consistent with earlier findings based on mosquito inoculation and serotype specific real-time PCR [15, 25].
An important aspect of dengue disease control is vector surveillance. Detection of DENV positive mosquitoes will be useful to monitor the infection rates within vector mosquito population and provide an early warning signal to predict an impending epidemic . Currently IFA, insect bioassay, ELISA and RT-PCR are the methods available for detection of DENV in mosquitoes . The qRT-PCR assay could detect DENV-2 RNA in all the inoculated mosquitoes. The sensitivity of detection was such that a 40th fraction of the infected mosquito body lysate was sufficient to give a positive signal in the qRT-PCR, thus even low levels of viral RNA in infected mosquitoes should be detected with the assay. No false positives were observed with control mosquitoes. However testing of field caught mosquitoes by the qRT-PCR will prove its usefulness for vector surveillance.
The group specific real-time PCR developed in this study will be a useful tool for differential diagnosis of dengue fever in a situation where a number of other diseases like malaria, Chikungunya, rickettsia and leptospira co-exist and are clinically indistinguishable. In areas where other flaviviruses are circulating, IgM detection is not conclusive because of cross reactivity that exists between DENV, JEV and WNV. The ability of the qRT-PCR assay to detect DENV RNA in seropositive individuals up to 10 days post onset of fever is an advantage in such situations. The ability of the assay to detect DENV-2 in inoculated mosquitoes makes it a potential tool for detecting DENV in field caught mosquitoes.
Gurukumar KR acknowledges Council of Scientific and Industrial Research, Government of India, for providing Senior Research Fellowship. Priyadarshini D acknowledges Indian council of Medical Research Government of India, for providing Senior Research Fellowship. We also Acknowledge Mr Walimbe AM for his help in statistical analysis of the data
- Halstead ScottB: Dengue. Lancet 2007, 370: 1644-1652. 10.1016/S0140-6736(07)61687-0View ArticlePubMedGoogle Scholar
- Cecilia D: Dengue: The Re-emerging Disease. In NIV commemorative compendium. Edited by: Mishra AC. Pune: National Institute of Virology; 2004:278-307.Google Scholar
- Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV: Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol 1992,30(3):545-551.PubMed CentralPubMedGoogle Scholar
- Harris E, Roberts TG, Smith L, Selle J, Kramer LD, Valle S, Sandoval E, Balmaseda A: Typing of dengue viruses in clinical specimens and mosquitoes by single-tube multiplex reverse transcriptase PCR. J Clin Microbiol 1998,36(9):2634-2639.PubMed CentralPubMedGoogle Scholar
- Chao DY, Davis BS, Chang GJ: Development of multiplex real-time reverse transcriptase PCR assays for detecting eight medically important flaviviruses in mosquitoes. J Clin Microbiol 2007,45(2):584-5899. 10.1128/JCM.00842-06PubMed CentralView ArticlePubMedGoogle Scholar
- Lai YL, Chung YK, Tan HC, Yap HF, Yap G, Ooi EE, Ng LC: Cost-effective real-time reverse transcriptase PCR (RT-PCR) to screen for Dengue virus followed by rapid single-tube multiplex RT-PCR for serotyping of the virus. J Clin Microbiol 2007,45(3):935-41. 10.1128/JCM.01258-06PubMed CentralView ArticlePubMedGoogle Scholar
- Kong YY, Thay CH, Tin TC, Devi S: Rapid detection, serotyping and quantitation of dengue viruses by TaqMan real-time one-step RT-PCR. J Virol Methods 2006,138(1–2):123-30. 10.1016/j.jviromet.2006.08.003View ArticlePubMedGoogle Scholar
- Johnson BW, Russell BJ, Lanciotti RS: Serotype-specific detection of dengue viruses in a fourplex real-time reverse transcriptase PCR assay. J Clin Microbiol 2005,43(10):4977-83. 10.1128/JCM.43.10.4977-4983.2005PubMed CentralView ArticlePubMedGoogle Scholar
- Chien LJ, Liao TL, Shu PY, Huang JH, Gubler DJ, Chang GJ: Development of real-time reverse transcriptase PCR assays to detect and serotype dengue viruses. J Clin Microbiol 2006,44(4):1295-1304. 10.1128/JCM.44.4.1295-1304.2006PubMed CentralView ArticlePubMedGoogle Scholar
- Shu PY, Chang SF, Kuo YC, Yueh YY, Chien LJ, Sue CL, Lin TH, Huang JH: Development of group- and serotype-specific one-step SYBR green I-based real-time reverse transcription-PCR assay for dengue virus. J Clin Microbiol 2003,41(6):2408-2416. 10.1128/JCM.41.6.2408-2416.2003PubMed CentralView ArticlePubMedGoogle Scholar
- Callahan JD, Wu SJ, Dion-Schultz A, Mangold BE, Peruski LF, Watts DM, Porter KR, Murphy GR, Suharyono W, King CC, Hayes CG, Temenak JJ: Development and evaluation of serotype- and group-specific fluorogenic reverse transcriptase PCR (TaqMan) assays for dengue virus. J Clin Microbiol 2001,39(11):4119-4124. 10.1128/JCM.39.11.4119-4124.2001PubMed CentralView ArticlePubMedGoogle Scholar
- Rosen L, Gubler D: The use of mosquitoes to detect and propagate dengue viruses. Am J Trop Med Hyg 1974,23(6):1153-1160.PubMedGoogle Scholar
- Sathish N, Manayani DJ, Shankar V, Abraham M, Nithyanandam G, Sridharan G: Comparison of IgM capture ELISA with a commercial rapid immunochromatographic card test and microwell ELISA for the detection of antibodies to dengue viruses. Indian J Med Res 2002, 115: 31-36.PubMedGoogle Scholar
- Clarke DH, Casals J: Techniques for hemagglutination and hemagglutination-inhibition with arthropod-borne viruses. Am J Trop Med Hyg 1958,7(5):561-73.PubMedGoogle Scholar
- Houng HS, Chung-Ming Chen R, Vaughn DW, Kanesa-thasan N: Development of a fluorogenic RT-PCR system for quantitative identification of dengue virus serotypes 1–4 using conserved and serotype-specific 3' noncoding sequences. J Virol Methods 2001,95(1–2):19-32. 10.1016/S0166-0934(01)00280-4View ArticlePubMedGoogle Scholar
- Laue T, Emmerich P, Schmitz H: Detection of dengue virus RNA in patients after primary or secondary dengue infection by using the TaqMan automated amplification system. J Clin Microbiol 1999,37(8):2543-2547.PubMed CentralPubMedGoogle Scholar
- Ito M, Takasaki T, Yamada K, Nerome R, Tajima S, Kurane I: Development and evaluation of fluorogenic TaqMan reverse transcriptase PCR assays for detection of dengue virus types 1 to 4. J Clin Microbiol 2004,42(12):5935-5937. 10.1128/JCM.42.12.5935-5937.2004PubMed CentralView ArticlePubMedGoogle Scholar
- Sadon N, Delers A, Jarman RG, Klungthong C, Nisalak A, Gibbons RV, Vassilev V: A new quantitative RT-PCR method for sensitive detection of dengue virus in serum samples. J Virol Methods 2008,153(1):1-6. 10.1016/j.jviromet.2008.06.023View ArticlePubMedGoogle Scholar
- Chutinimitkul S, Payungporn S, Theamboonlers A, Poovorawan Y: Dengue typing assay based on real-time PCR using SYBR Green I. J Virol Methods 2005,129(1):8-15. 10.1016/j.jviromet.2005.05.006View ArticlePubMedGoogle Scholar
- Heid CA, Stevens J, Livak KJ, Williams PM: Real time quantitative PCR. Genome Res 1996,6(10):986-994. 10.1101/gr.6.10.986View ArticlePubMedGoogle Scholar
- Libraty DH, Endy TP, Houng HS, Green S, Kalayanarooj S, Suntayakorn S, Chansiriwongs W, Vaughn DW, Nisalak A, Ennis FA, Rothman AL: Differing influences of virus burden and immune activation on disease severity in secondary dengue-3 virus infections. J Infect Dis 2002,185(9):1213-1221. 10.1086/340365View ArticlePubMedGoogle Scholar
- Brown JL, Wilkinson R, Davidson RN, Wall R, Lloyd G, Howells J, Pasvol G: Rapid diagnosis and determination of duration of viraemia in dengue fever using a reverse transcriptase polymerase chain reaction. Trans R Soc Trop Med Hyg 1996,90(2):140-1433. 10.1016/S0035-9203(96)90115-7View ArticlePubMedGoogle Scholar
- Vaughn DW, Green S, Kalayanarooj S, Innis BL, Nimmannitya S, Suntayakorn S, Rothman AL, Ennis FA, Nisalak A: Dengue in the early febrile phase: viremia and antibody responses. J Infect Dis 1997,176(2):322-330.View ArticlePubMedGoogle Scholar
- Warrilow D, Northill JA, Pyke A, Smith GA: Single rapid TaqMan fluorogenic probe based PCR assay that detects all four dengue serotypes. J Med Virol 2002,66(4):524-8. 10.1002/jmv.2176View ArticlePubMedGoogle Scholar
- Vaughn DW, Green S, Kalayanarooj S, Innis BL, Nimmannitya S, Suntayakorn S, Endy TP, Raengsakulrach B, Rothman AL, Ennis FA, Nisalak A: Dengue viremia titer, antibody response pattern, and virus serotype correlate with disease severity. J Infect Dis 2000,181(1):2-9. 10.1086/315215View ArticlePubMedGoogle Scholar
- Loroño-Pino MA, Cropp CB, Farfán JA, Vorndam AV, Rodríguez-Angulo EM, Rosado-Paredes EP, Flores-Flores LF, Beaty BJ, Gubler DJ: Common occurrence of concurrent infections by multiple dengue virus serotypes. Am J Trop Med Hyg 1999,61(5):725-730.PubMedGoogle Scholar
- Philip Samuel P, Tyagi BK: Diagnostic methods for detection & isolation of dengue viruses from vector mosquitoes. Indian J Med Res 2006,123(5):615-628.PubMedGoogle 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.