The Journal of Molecular Diagnostics, Vol. 14, No. 4, July 2012 Copyright © 2012 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2012.02.006 Detection and Typing of Viruses Using Broadly Sensitive Cocktail-PCR and Mass Spectrometric Cataloging Demonstration with Dengue Virus Charul Gijavanekar,* Rafal Drabek,† Mithil Soni,† George W. Jackson,† Ulrich Strych,* George E. Fox,*‡ Yuriy Fofanov,*§¶ and Richard C. Willson*‡¶ From the Departments of Biology and Biochemistry,* Chemical and Biomolecular Engineering,‡ and Computer Science,§ University of Houston, Houston; BioTex, Inc.,† Houston; and the Methodist Hospital Research Institute,¶ Houston, Texas Virus detection and taxonomic identification of serotypes, strains, or genotypes provide important information relevant for diagnosis, and for the epidemiological characterization and tracking of new strains in an endemic region. In the specific case of dengue virus, rapid serotype identification can also be useful in the treatment of secondary infections that may cause the more severe dengue hemorrhagic fever and dengue shock syndrome. In this work, dengue virus was used as a model to test a new approach of combining broadly sensitive RT-PCR amplification of nearly any virus strain with subsequent serotype- and finer-level identification by mass spectrometry. PCR primers were appended with promoter sequences, such that the resulting PCR products could be transcribed into RNA. RNA fragments generated by guanosine-specific RNase T1 digestion were analyzed by matrix-assisted laser desorption/ionization-time of flight mass spectrometry. Viral serotypes were identified by comparing the pattern of observed fragment masses to a mass database. The database was created by computationally fragmenting 2517 dengue strains after each guanosine residue using the same primers. Computationally, all 2517 strains in the mass database were correctly identified at the serotype level from the predicted PCR product. The methodology was successfully demonstrated experimentally by identifying the serotypes of eight test strains using mosquito cell cultures infected with strains of all four serotypes and with full-length cDNA clones. ( J Mol Diagn 2012, 14:402407; http://dx.doi.org/10.1016/j.jmoldx.2012.02.006) 402 It is often desirable that a pathogen diagnostic method provide some degree of taxonomic information about the particular genotype that has been detected, both for surveillance and in some cases for prognostic uses. Taxonomic classification of a particular isolate without recourse to sequencing, culture, or other time-consuming methods is especially desirable. Using dengue virus as a model system, we demonstrate the utility of combining broadly sensitive RT-PCR amplification of sequences from nearly any strain of the target virus with rapid mass spectrometric characterization of the individual isolates. Dengue virus is the causative agent of dengue fever, dengue hemorrhagic fever, and dengue shock syndrome with an estimated 50 million infections and 12,500 fatalities worldwide each year.1,2 The virus is reported in an increasing number of countries, including, most recently, locally acquired infections in the United States of America.3 The virus occurs in four serotypes (DENV-1 to -4) with more than 2700 different strains currently reported [National Center for Biotechnology Information (NCBI), http://www.ncbi.nlm.nih.gov/genome/10308, last accessed April 30, 2011]. Since a second dengue virus infection by a different serotype is more likely to lead to hemorrhagic fever or dengue shock syndrome, the rapid identification of the serotype is therefore of great interest in these cases. Current methods of detection and typing are primarily serological or rely on time-consuming virus cul- This work was supported in part by the Department of Homeland Security under contract number HSHQDC-08-C-00183 (G.E.F., Y.F., and R.C.W.), by Welch Foundation grants E-1264 (R.C.W.) and E-1451 (G.E.F.), and National Institutes of Health SBIR grant 2R44AI066425 (G.W.J.). Accepted for publication February 3, 2012. Disclosures: R.D., M.S., and G.W.J. receive financial support from BioTex, Inc., Houston, TX. Supplemental material for this article can be found at http://jmd. amjpathol.org or at http://dx.doi.org/10.1016/j.jmoldx.2012.02.006. Current address of C.G., Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX. Address reprint requests to Richard C. Willson, Ph.D., Department of Chemical and Biomolecular Engineering, University of Houston, 4800 Calhoun Rd., Houston, Texas 77204-4004. E-mail: willson@uh.edu. Mass Typing of Dengue Virus 403 JMD July 2012, Vol. 14, No. 4 ture, antigen and antibody detection, and nucleic acid– based technologies.4 –12 We have recently developed a new approach to rapid identification of pathogens by combining PCR amplification and matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry of base-specific ribonuclease digestion fragments of RNA transcribed from the PCR products.13–16 In this approach, PCR primers are appended with an RNA polymerase promoter sequence so that the amplification products can be readily transcribed into RNA. The RNA is then digested using RNase T1, a highly guanosine-specific endoribonuclease, and the resulting fragments are analyzed by MALDI-TOF. The mass spectrum of the digest is then compared to an existing database of digestion patterns for pathogen genes, eg, bacterial genomic 16S rRNA13–15 or viral genomic sequences.16 We also recently reported a single set of 10 PCR primers designed for the specific and sensitive detection of nearly all (95%) of 1688 then-sequenced dengue virus strains of all four serotypes.17 These new primers are highly specific for dengue virus when tested in the presence of a vast excess of human DNA or DNA of the carrier mosquito, Aedes aegypti. Moreover, they can be used directly for reverse transcription of dengue viral RNA, bypassing the need for initial synthesis of a cDNA template with a different set of primers. Computational tests further predict that the primers are highly specific for dengue virus when tested against other non-dengue flaviviruses. Additionally, sequence variations in dengue virus due to wide geographical distribution did not significantly degrade the predicted performance of the primers, supporting the potential global use of these primers. In the present work, we combine this near-universal amplification capability with mass spectrometric identification to enable rapid detection, serotyping, and strain identification of nearly all known dengue strains. We report the results of simulation of PCR and subsequent typing of dengue strains. We also demonstrate the methodology experimentally using mosquito cell cultures infected with DENV-1 (Piura, Peru), DENV-2 (New Guinea C), DENV-3 (Asuncion, Paraguay), and DENV-4 (Dominica, West Indies) strains, and full-length cDNA clones. Materials and Methods Predictive Mass Spectral Database We created a database using publicly available sequence information on dengue virus strains. On April 14, 2011, a search of the NCBI nucleotide database (GenBank) for “dengue[title] and viruses[organism]” resulted in 9566 hits. Among the dengue sequences, we identified 2718 as serotype 1, 3119 as serotype 2, 2671 as serotype 3, and 855 as serotype 4, using sequence name identifiers. The serotypes of the remaining 203 sequences could not be determined because of ambiguity in sequence names and sequence information. Of the 9566 sequences, 2736 sequences were identified as complete genomes and were recognized as individual strains. To create virtual PCR amplicons using our primer cocktail (Table 1), all possible combinations of forward and reverse primers were queried against genomic sense and antisense strands in these 2736 strains. Additionally, mismatch-tolerant searches were conducted using custom-coded fast Fourier transform– based correlation functions. This approach is similar to, but distinct from that described previously by other groups18,19 in that an initial transform of genetic sequence information to complex numbers (eg, [A,T,G,C] ↔ [1,⫺1,i,⫺i]) allows the use of well-established Fourier transform– based correlation functions for quickly searching for primer sequences while allowing for any specified number of mismatches. By allowing for up to four mismatches in each primer (mismatches positioned randomly in the primer sequence), we simulated less-stringent experimental conditions. We cataloged all potential amplicons in a database for subsequent in silico base-specific cleavage by RNase T1 (simulated cleavage after guanosine residues and prediction of mass fragment patterns). The entire data set is organized and curated using a MySQL database running under Linux. Spectral matching and database updating procedures were implemented in the Perl scripting language. Results are formatted into HTML and presented to the user in tabular format in a standard web browser. This database structure allows seamless Table 1. Primer Sequences Used for Dengue Virus Detection Primer pair Dengue serotype 1G1P1 DENV-1 1G1P1 DENV-2 2G2P5 DENV-2 1G3P6 DENV-1 1G4P217 DENV-3 1G5P30 DENV-4 Primer sequences F: R: F: R: F: R: F: R: F: R: F: R: 5=-TAATACGACTCACTATAGGCAAACCATGGAAGCTGTACG-3= 5=-CTATATACTATATATATACTTCTGTGCCTGGAATGAT-3= 5=-TAATACGACTCACTATAGGCAAACCATGGAAGCTGTACG-3= 5=-CTATATACTATATATATACTTCTGTGCCTGGAATGATGCT-3= 5=-TAATACGACTCACTATAGGGAGTGGAGTGGAAGGAGAAGGG-3= 5=-CTATATACTATATATATACCCTCTTGGTGTTGGTCTTTGC-3= 5=-TAATACGACTCACTATAGGCAGACTAGTGGTTAGAGGAGA-3= 5=-CTATATACTATATATATACGGAATGATGCTGTAGAGACA-3= 5=-TAATACGACTCACTATAGGATATGCTGAAACGCGTGAG-3= 5=-CTATATACTATATATATACCATCATGAGACAGAGCGAT-3= 5=-TAATACGACTCACTATAGGTTCCAACAAGCAGAACAACAT-3= 5=-CTATATACTATATATATACGCTACAGGCAGCACGGTTT-3= All forward (F) and reverse (R) primer sequences were appended (bold) to either a T7 promoter sequence or an internal standard mass calibration sequence, respectively. For antisense strand mass patterns, reverse primers were appended to the T7 promoter, and forward primers to a mass calibration sequence (not shown). 404 Gijavanekar et al JMD July 2012, Vol. 14, No. 4 updating at any desired frequency as additional sequence information becomes available. PCR Amplification and Mass Spectrometry of PCR Products In a typical experiment, each dengue strain was tested with the primer set with a T7 promoter sequence appended to either forward or reverse primers (Table 1). The PCR products were purified using QIAquick PCR Purification Kit (Qiagen, Valencia, CA). DNA was then transcribed into RNA and enzymatically cleaved with RNaseT1, followed by acquisition of the corresponding MALDI spectra. We purified PCR products for use as templates for in vitro transcription using the T7-Flash kit (Epicenter, Madison, WI). Amino-allyl UTP was substituted for UTP in the reaction to give improved distinction of U/C-containing mass fragments.15 The transcripts were completely digested by RNase T1 at 37°C for 5 minutes and then placed on ice for MALDI preparation using ZipTip reverse-phase pipette tips (Millipore, Billerica, MA). Mass spectra of the digests were acquired in linear, negative ion mode on a Voyager DE-STR MALDITOF instrument (Applied Biosystems, Carlsbad, CA). We then compared the observed masses to the database above using a spectral scoring function based on inner products as described previously.13–16 Primers (Table 1) were tested against dengue virus cDNAs and viral RNA. Full-length cDNA clones of DENV-1 West Pacific (GenBank accession: U88535; NCBI GI: 1854036), DENV-2 New Guinea C (M29095; GI: 323447), DENV-3 (FJ639719; GI: 221071269), and DENV-4 (GU289913; GI: 280987259) in the yeast–Escherichia coli shuttle vector pRS424 were used. Each reaction contained 100 pg of cDNA and 100 nmol/L of each primer. Controls omitting the DNA template were included. Brilliant II SYBR Green Q-PCR Master Mix (Agilent Technologies, Valencia, CA) was used to perform real-time PCR. Thermocycling was performed for 35 cycles, with annealing at 60°C (1 minute) and extension at 72°C (1 minute). Total RNA from C6/36 mosquito cell cultures infected with dengue virus; DENV-1 (Piura, Peru), DENV-2 (New Guinea C) (GenBank accession: M29005; NCBI GI: 323447), DENV-3 (Asuncion, Paraguay), and DENV-4 (Dominica, West Indies) was tested. The DENV-1, DENV-3, and DENV-4 strains do not have NCBI identifiers because they are not yet sequenced. We used primer cocktails (100 nmol/L per primer; 10 primers) for both the reverse transcription and amplification steps. Brilliant II SYBR Green Q-RTPCR Master Mix (Agilent Technologies) was used to perform real-time RT-PCR. The reverse transcription step was performed according to the manufacturer’s protocol (50°C, 30 minutes). Thermocycling was performed for 35 cycles, with annealing at 60°C (1 minute) and extension at 72°C (1 minute). Controls omitting RNA template and reverse transcriptase were included to rule out primer– dimer formation and amplification, respectively, due to the presence of any DNA in the total RNA sample. We also tested total RNA from normal Table 2. Recovery of Serotype by Simulated Identification of 2517 Dengue Strains in the Mass Spectrometric Database Number of test strains matching a Serotype of in silico serotype (%) sample (no. of DENV-1 DENV-2 DENV-3 DENV-4 tested strains) DENV-1 (982) DENV-2 (856) DENV-3 (597) DENV-4 (82) 99.98 0.05 0 0 0.02 99.95 0 0 0 0 100 0 0 0 0 100 Numbers are percentage of tested strains matching a serotype with rank of 1. uninfected C6/36 cells as a negative control. The Mx3005P QPCR system (Agilent Technologies) was used for thermocycling and real-time monitoring. For demonstration of serotype identification, only the sense strand was transcribed in case of the virus RNA samples. Simulated Serotyping of Dengue Strains Using the same set of 10 primers (Table 1) that were used to compile the database, we exhaustively simulated the recovered serotype for each entry in our database. Specifically, we predicted the expected set of RNase T1 fragments and compared it with the predicted mass patterns from the dengue mass spectral database. The final calculated mass spectral coincidence was normalized by the overall highest value, generating an “identification rank” value. The highest rank of 1 was therefore the best match or serotype ID of a test strain. The trivial result, that is, the recovered serotype of the test strain itself, was ignored. As the strains of one serotype are more closely related to each other than to other serotypes, the predictive mass database is useful in the identification of the serotype of a test strain.16 The results are expressed as the percentage of tested strains matching their respective serotype with a rank of 1 (Table 2). Results In Silico PCR Detection, Mass-Spectrometric Database Prediction, and Serotyping Using the primer set and considering mass patterns from amplicons of both sense and antisense strands of the dengue virus genome, we predicted the detection of 2517 of 2736 dengue strains. Following simulated RNase T1 digestion, we predicted 14,084 unique mass patterns (4146 DENV-1, 6055 DENV-2, 3573 DENV-3, and 310 DENV-4) to form the mass spectrometric database. The number of unique patterns is higher than the number of unique strains because each strain can have multiple patterns obtained through the use of different primer pairs. The recovered serotype for each entry in our database of 2517 strains was determined by simulated identification where the trivial result (recovered serotype for the test strain) was ignored. All mass patterns obtained were considered for comparison with the data- Mass Typing of Dengue Virus 405 JMD July 2012, Vol. 14, No. 4 1.0 DENV-4 0.5 0.0 1.0 Relative Peak Intensity DENV-3 0.5 0.0 1.0 DENV-2 0.5 0.0 DENV-1 1.0 0.5 0.0 1000 2000 3000 4000 5000 rank number and the total number of strains can differ because more than one strain can have the same rank based on mass pattern. Comparing mass patterns of sense strands for PCR amplicons of DENV RNA, the top hits matched the test serotype in every case (Table 4, detailed in Supplemental Tables S5–S8 available at http://jmd.amjpathol.org). For DENV-1 (Piura, Peru), the 200 highest ranking strains belonged to serotype 1. For DENV-2, all 505 strains matched to serotype 2, and the test strain (GI: 323447) was found among the highest ranked strains. For DENV-3, all 500 strains belonged to serotype 3. For DENV-4, the top 7 identified strains belonged to serotype 4, and the remaining to either DENV-3 or DENV-4. We found that there are similarities in mass patterns between DENV-3 and -4, which are not directly discernible from sequence alignment. (Supplemental Table S9, available at http://jmd.amjpathol.org, gives some of the details of the underlying database in terms of the number of masses and unique patterns for each serotype.) The similarity of DENV-4 strains implies that when one of the DENV-4 mass patterns matches that of DENV-3, most other DENV-4 mass patterns also show similar results. Fortunately, such problems can be predicted by simulation and can be resolved by analysis of the antisense strand or additional sequence regions. Mass (Da) Figure 1. Mass spectra of fragments generated from PCR products of all four serotypes (DENV-1 to -4) using the primer cocktail. Vertical axis is a normalized representation of relative peak intensity of all serotypes. base. For each of the 2517 test strains, the serotype was identified correctly (Table 2). Experimental Demonstration of PCR Detection and Mass-Spectrometric Serotyping In a typical experiment, each dengue strain was amplified with the primer set shown in Table 1. PCR products were purified and DNA was transcribed into RNA and enzymatically cleaved with RNaseT1 followed by acquisition of the corresponding MALDI-TOF spectrum. The observed PCR product fragment masses of all four serotypes were then compared with the predicted masses for 2517 dengue virus strains in the database (Figure 1). The matches are listed in order from highest to lowest match, which was based on the identification rank or the normalized spectral coincidence analysis score. The product of coincidence scores of mass spectra of both sense and antisense strands was used to determine the final outcome and normalized by the highest value. Comparing mass patterns of the sense and antisense strands of the PCR amplicons of DENV cDNA clones against the predicted database (summarized in Tables 3 and 4, detailed in Supplemental Tables S1–S4 available at http://jmd.amjpathol.org), the top hits matched the tested serotype in every case. The test strains of DENV-1, DENV-2, and DENV-4 were found among the highest ranking strains. However, the DENV-3 test strain ranked 21st among 498 DENV-3 strains considered. The lowest Discussion We have shown that the combination of human-blind primer-cocktail PCR and typing by relatively inexpensive MALDI-TOF mass spectrometry enables broadly sensitive detection and serotype identification of dengue virus. We demonstrated this methodology by simulation of typing of dengue strains in silico, and subsequent experiments with both cDNA clones and dengue virus genomic RNA. Having just one set of primers for the identification of all dengue serotypes and strains constitutes a significant advantage as any test sample can be identified for the presence of dengue virus rapidly without any prior knowledge about its identity. Comparative analysis of the mass spectrum profile can particularly aid in the identification of a previously unidentified or new isolate. Standard mass spectra may be used as a reference to identify unknown or ambiguous test samples. As demonstrated by the correct identification of three of the four tested dengue cDNA strains (Tables 3 and 4) using PCR amplicons, strain-level resolution may be achieved using the mass spectra. Strain identification may be informative in tracking virus spread for epidemiological characterization, but may not be necessary in curbing a potential outbreak or even for treatment. Identification of both virus and serotype, however, is necessary and may help control hemorrhagic fever spread and fatality by timely identification. Serotypes of all 2517 strains were predicted to be identified correctly, when each strain was tested against the database and the test-strain was ignored from the recovered matches. The current primer set only examines 2% to 4% of the dengue genome. To further improve identification, a 406 Gijavanekar et al JMD July 2012, Vol. 14, No. 4 Table 3. Mass Spectral Typing of DENV cDNA Clones with Primer Cocktail by Overall Score Overall rank Normalized score DENV-1 WP (GI: 1854036) 1 1 108† 1 1 0.73 Dengue virus 1, isolate DENV-1/PH/BID-V2940/2004 Dengue virus 1, clone WestPac Dengue virus type 1 recombinant clone rDEN1mutF, complete genome 259496234 1854036 22901063 DENV-2 NGC (GI: 323447) 1 1 Dengue virus 2 isolate DENV-2/CO/BID-V1598/1944, complete genome DEN2CGA dengue virus type 2 complete genome/dengue virus 2 Dengue virus type 2 strain ThNH54/93, complete genome/ dengue virus 2 209976573 1 390 1 ‡ DENV-3 (GI: 221071269) 1 0.53 1 21 0.95 498§ 0.5 DENV-4 (GI: 280987259) 1 1 76¶ 1 1 0.27 Identified serotype NCBI GI* 323447 6841593 Dengue virus type 3 strain C0360/94, complete genome/Thailand/plasma from patient DF Dengue virus 3 isolate DENV-3/KH/BID-V2071/2000, complete genome DENCME dengue type 3 virus complete genome RNA, complete cds/dengue virus 3 60326968 221071269 323468 Dengue virus 4 strain 341750, complete genome/Colombia Dengue virus type 4 strain 814669, complete genome Dengue virus type 4 strain ThD4_0476_97, complete genome 280987259 12659201 53653742 *Test strain in bold. † Lowest ranking DENV-1 strain; all 108 DENV-1 strains ranked above all strains of any other serotype. ‡ Lowest ranking DENV-2 strain; all 397 DENV-2 strains ranked above all strains of any other serotype. § Lowest ranking DENV-3 strain; all 500 DENV-3 strains ranked above all strains of any other serotype. ¶ Lowest ranking DENV-4 strain; all 77 DENV-4 strains ranked above all strains of any other serotype. cds, coding sequence. larger fraction of the genome may be analyzed by using primers spanning a larger genomic region using more primer pairs, or using more than one amplicon. Ibis Biosciences’ T-5000 Biosensor (Abbott Laboratories, Abbott Park, IL) also utilizes a “universal” PCR approach coupled with subsequent mass spectrometric characterization of the resulting whole amplicons.20,21 Although our approach using RNase T1 fragmentation requires additional transcription and cleavage reactions, such fragmentation allows for less expensive, lower-resolution MALDI instrumentation to be used. In recent years, several studies have suggested serotype-specific disease behavior.22–24 Specifically, the carrier mosquito A. aegypti has been reported to be more susceptible to infection by DENV-2 serotype,25 which may lead to greater incidence of infection by DENV-2. Rapid and quan- titative typing may help in early detection of DENV-2 for aggressive prevention of a potential outbreak. Co-circulation of two serotypes in a region is also thought to promote occurrence of dengue hemorrhagic fever,26 associated with secondary infections. Serotype identification in such regions may help reduce secondary infection. Although there is no specific drug for treatment of dengue infection, several candidates have been reported.27–30 Recently, one drug candidate was reported to inhibit only DENV-1, -2, and -3 serotypes29 (possibly because DENV-4 is most divergent in E protein sequence31,32 where the inhibitor is thought to bind). If eventually dengue serotype-specific drugs are developed, then serotyping would be of essence to determine the course of treatment. Finally, in the case of a highly mutable RNA virus such as dengue (in which the mutation rate can be as high as Table 4. Experimental Serotype Identification of All Dengue Virus cDNA and RNA Test Strains Tested strain (NCBI Accession/GI) Virus cDNA DENV-1 WP (U88535/1854036) DENV-2 NGC (M29095/323447) DENV-3 (FJ639719/221071269) DENV-4 (GU289913/280987259) Virus RNA DENV-1 (Piura, Peru) DENV-2 NGC (M29095/323447) DENV-3 (Asuncion, Paraguay) DENV-4 (Dominica, West Indies) Normalized score Correct serotype ID Correct strain ID with top score 1 1 1 1 Yes Yes Yes Yes Yes Yes No; Score: 0.95 Yes NA 1 NA NA Yes Yes Yes Yes NA Yes NA NA Mass spectral score normalized for each serotype is listed with observed serotype ID. NA, not applicable, only the serotype can be identified when sequence information is not available. Mass Typing of Dengue Virus 407 JMD July 2012, Vol. 14, No. 4 one per genome (11 kb) per replication33,34), the predictive mass database must be continuously updated using current sequence information. Our method and supporting database can be easily updated, and are thus well suited to address this issue. The combination of cocktailPCR strategy and subsequent mass spectral analysis of the amplicons also may be useful for the diagnosis of other pathogens16 for which PCR detection methods are already well established. Acknowledgments We thank Drs. Barry Falgout, Bangti Zhao, and Robin Levis (US Food and Drug Administration) for providing dengue virus cDNA clones of DENV and to Dr. Robert B. Tesh (World Reference Center for Emerging Viruses and Arboviruses, University of Texas Medical Branch) for providing DENV-infected C6/36 mosquito cell cultures. References 1. WHO: Dengue and dengue haemorrhagic fever. Fact Sheet 2009. Geneva, Switzerland, World Health Organization, 2009 2. Guzman MG, Halstead SB, Artsob H, Buchy P, Farrar J, Gubler DJ, Hunsperger E, Kroeger A, Margolis HS, Martinez E, Nathan MB, Pelegrino JL, Simmons C, Yoksan S, Peeling RW: Dengue: a continuing global threat. Nat Rev Microbiol 2010, 8:S7–S16 3. Centers for Disease Control and Prevention (CDC): Locally acquired Dengue—Key West, Florida, 2009 –2010. MMWR Morb Mortal Wkly Rep 2010, 59:577–581 4. Saxena P, Dash PK, Santhosh SR, Shrivastava A, Parida M, Rao PL: Development and evaluation of one step single tube multiplex RT-PCR for rapid detection and typing of dengue viruses. Virol J 2008, 5:20 5. Parida M, Horioke K, Ishida H, Dash PK, Saxena P, Jana AM, Islam MA, Inoue S, Hosaka N, Morita K: Rapid detection and differentiation of dengue virus serotypes by a real-time reverse transcription-loop-mediated isothermal amplification assay. J Clin Microbiol 2005, 43:2895– 2903 6. Lo CL, Yip SP, Cheng PK, To TS, Lim WW, Leung PH: One-step rapid reverse transcription-PCR assay for detecting and typing dengue viruses with GC tail and induced fluorescence resonance energy transfer techniques for melting temperature and color multiplexing. Clin Chem 2007, 53:594 –599 7. Vazquez S, Hafner G, Ruiz D, Calzada N, Guzman MG: Evaluation of immunoglobulin M and G capture enzyme-linked immunosorbent assay Panbiokits for diagnostic dengue infections. J Clin Virol 2007, 39:194 –198 8. Kao CL, King CC, Chao DY, Wu HL, Chang GJJ: Laboratory diagnosis of dengue virus infection: current and future perspectives in clinical diagnosis and public health. J Microbiol Immunol Infect 2005, 38:5–16 9. Hang VT, Nguyen NM, Trung D, Tricou V, Yoksan S, Dung NM, Van Ngoc T, Hien TT, Farrar J, Wills B, Simmons CP: Diagnostic accuracy of NS1 ELISA and lateral flow rapid tests for dengue sensitivity, specificity and relationship to viraemia and antibody responses. PLoS Negl Trop Dis 2009, 3:e360 10. Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV: Rapid detection of typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol 1992, 37:545–551 11. Innis BL, Nisalak A, Nimmannitya S, Kusalerdchariya S, Chongswasdi V, Suntayakorn S, Puttisri P, Hoke C: An enzyme-linked immunosorbent assay to characterize dengue infections where dengue and Japanese encephalitis co-circulate. Am J Trop Med Hyg 1989, 40: 418 – 427 12. Kong YY, Thay CH, Tin TC, Devi S: Rapid detection, serotyping and quantitation of dengue viruses by TaqMan real-time one-step RTPCR. Jf VirolMethods 2006, 138:123–130 13. Jackson GW, McNichols RJ, Fox GE, Willson RC: Bacterial genotyping by 16S ribosomal RNA mass cataloging. BMC Bioinformatics 2006, 7:321 14. Zhang Z, Jackson GW, Fox GE, Willson RC: Microbial identification by mass cataloging. BMC Bioinformatics 2006, 7:117 15. Jackson GW, McNichols RM, Fox GE, Willson RC: Universal bacterial identification by mass spectrometry of 16S ribosomal RNA cleavage products. Int J Mass Spectrom 2007, 261:218 –226 16. Jackson GW, McNichols RJ, Fox GE, Willson RC: Toward universal flavivirus identification by mass cataloging. J Mol Diagn 2008, 10: 135–141 17. Gijavanekar C, Anez-Lingerfelt M, Feng C, Putonti C, Fox GE, Sabo A, Fofanov Y, Willson RC: PCR detection of nearly any dengue virus strain using a highly-sensitive primer ‘cocktail’. FEBS J 2011, 278: 1676 –1687 18. Rockwood AL, Crockett DK, Oliphant JR, Elenitoba-Johnson KS: Sequence alignment by cross-correlation. J Biomol Tech 2005, 16:453–458 19. Leitao HC, Pessoa LS, Stolfi J: Mutual information content of homologous DNA sequences. Genet Mol Res 2005, 4:553–562 20. Ecker DJ, Sampath R, Massire C, Blyn LB, Hall TA, Eshoo MW, Hofstadler SA: Ibis T5000: a universal biosensor approach for microbiology. Nat Rev Microbiol 2008, 6:553–558 21. Grant-Klein RJ, Baldwin CD, Turell MJ, Rossi CA, Li F, Lovari R, Crowder CD, Matthews HE, Rounds MA, Eshoo MW, Blyn LB, Ecker DJ, Sampath R, Whitehouse CA: Rapid identification of vector-borne flaviviruses by mass spectrometry. Mol Cell Probes 2010, 24:219 –228 22. 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, 181:2–9 23. Fried JR, Gibbons RV, Kalayanarooj S, Thomas SJ, Srikiatkhachorn A, Yoon IK, Jarman RG, Green S, Rothman AL, Cummings DA: Serotype-specific differences in the risk of dengue hemorrhagic fever: an analysis of data collected in Bangkok, Thailand from 1994 to 2006. PLoS Negl Trop Dis 2010, 4:e617 24. Thein S, Aung MM, Shwe TN, Aye M, Zaw A, Aye K, Aye KM, Aaskov J: Risk factors in dengue shock syndrome. Am J Trop Med Hyg 1997, 56:566 –572 25. Armstrong PM, Rico-Hesse R: Efficiency of dengue serotype 2 virus strains to infect and disseminate in Aedes aegypti. Am J TropMed Hyg 2003, 68:539 –544 26. Khan E, Hasan R, Mehraj V, Nasir A, Siddiqui J, Hewson R: Co-circulations of two genotypes of dengue virus in 2006 out-break of dengue hemorrhagic fever in Karachi. Pakistan J Clin Virol 2008, 43:176 –179 27. Yang CC, Hsieh YC, Lee SJ, Wu SH, Liao CL, Tsao CH, Chao YS, Chern JH, Wu CP, Yueh A: Novel dengue virus-specific NS2B/NS3 protease inhibitor, BP2109, discovered by a high-throughput screening assay. Antimicrob Agents Chemothery 2011, 55:229 –238 28. Yin Z, Chen Y-L, Schul W, Wang Q-Y, Gu F, Duraiswamy J, Kondreddi RR, Niyomrattanakit P, Lakshminarayana SB, Goh A, Xu HY, Liu W, Liu B, Lim JY, Ng CY, Qing M, Lim CC, Yip A, Wang G, Chan WL, Tan HP, Lin K, Zhang B, Zou G, Bernard KA, Garrett C, Beltz K, Dong M, Weaver M, He H, Pichota A, Dartois V, Keller TH, Shi PY: An adenosine nucleoside inhibitor of dengue virus. Proc Natl Acad Sci U S A 2010, 106:20435–20439 29. Kaptein SJ, De Burghgraeve T, Froeyen M, Pastorino B, Alen MM, Mondotte JA, Herdewijn P, Jacobs M, de Lamballerie X, Schols D, Gamarnik AV, Sztaricskai F, Neyts J: A derivate of the antibiotic doxorubicin is a selective inhibitor of dengue and yellow fever virus replication in vitro. Antimicrob Agents Chemother 2010, 54:5269 –5280 30. Costin JM, Jenwitheesuk E, Lok SM, Hunsperger E, Conrads KA, Fontaine KA, Rees CR, Rossmann MG, Isern S, Samudrala R, Michael SF: Structural optimization and de novo design of dengue virus entry inhibitory peptides. PLoS Negl Tropis 2010, 4:e721 31. Holmes EC, Twiddy SS: The origin, emergence and evolutionary genetics of dengue virus. Infect, Genet Evol 2003, 3:19 –28 32. Lanciotti RS, Gubler DJ, Trent DW: Molecular evolution and phylogeny of dengue-4 viruses. J Gen Virol 1997, 78:2279 –2286 33. Jenkins GM, Rambaut A, Pybus OG, Holmes EC: Rates of molecular evolution in RNA viruses: a quantitative phylogenetic analysis. J Mol Evol 2002, 54:156 –165 34. Twiddy SS, Holmes EC, Rambaut A: Inferring the rate and time-scale of dengue virus evolution. Mol Biol Evol 2003, 20:122–129