Journal of Medical Virology 66:524±528 (2002) DOI 10.1002/jmv.2176
Single Rapid TaqMan Fluorogenic Probe Based PCR Assay That Detects All Four Dengue Serotypes David Warrilow,* Judith A. Northill, Alyssa Pyke, and Greg A. Smith Queensland Health Scienti®c Services, Queensland, Australia
Public health laboratories require rapid diagnosis of dengue outbreaks for application of measures such as vector control. We have developed a rapid single ¯uorogenic probe-based polymerase chain reaction assay for the detection of all four dengue serotypes (FUDRT-PCR). The method employs primers and probe that are complementary to the evolutionarily conserved 30 untranslated region of the dengue genome. The assay detected viral RNA of strains of all four dengue serotypes but not of the ¯aviviruses Japanese encephalitis virus, Murray Valley encephalitis virus, Kunjin, Stratford, West Nile, Alfuy or Yellow fever. When compared to an existing nested-PCR assay for the detection of dengue on clinical samples, FUDRT-PCR detected dengue 1 (100%, n 14), dengue 2 (85%, n 13), dengue 3 (64%, n 14) and dengue 4 (100%, n 3) with the indicated sensitivities. FUDRT-PCR enables diagnosis of acute dengue infection in four hours from sample receipt. In addition, a single-test procedure should result in a reduction in the number of tests performed with considerable cost savings for diagnostic laboratories. J. Med. Virol. 66:524±528, 2002. ß 2002 Wiley-Liss, Inc. KEY WORDS:
Dengue; ¯avivirus; ¯uorogenic; 30 ; untranslated region
INTRODUCTION Dengue is a viral disease of major health importance affecting up to 80 million people worldwide annually [Pinheiro and Corber, 1997]. The agents causing the disease are enveloped positive strand viruses classi®ed in the family Flaviviridae, genus Flavivirus; there are four distinct serotypes. The disease is transmitted by mosquitoes of the genus Aedes [Rodhain and Rosen, 1997]. Infection results in a febrile illness during which the patient is viremic [Gubler et al., 1981]. Infection can occur in a naõÈve host (primary infection) or in an individual exposed to a dengue virus of a different serotype to that previously (secondary infection). In primary infection, IgM antibodies are detectable in half of patients at defervescence (3±5 days after onset of ß 2002 WILEY-LISS, INC.
symptoms) [Innis et al., 1989] coincident with the beginning of the decline in virus levels (5±12 days from the onset) [Gubler et al., 1981]. Dengue infection can be diagnosed serologically by a four-fold rise in antibody titer of convalescent over acute phase sera. Convalescent sera, however, are not often available. Antibodies resulting from primary infection are crossreactive with other ¯avivirus antigens; a feature exacerbated in secondary ¯avivirus infections [Innis, 1997]. Hence, it is only possible to make a de®nitive diagnosis of dengue infection if exposure to other ¯aviviruses can be excluded. Due to the greater speci®city of IgM antibodies in ¯avivirus infection [Westaway, 1968; Edelman et al., 1973], a presumptive diagnosis of recent infection is often made if IgM antibodies are present. Hence, a serological diagnosis is often not that informative. PCR-based dengue tests, due to the speci®city of ampli®cation, enable a de®nitive diagnosis and serotyping of the virus. In addition, DNA sequencing of the ampli®cation product enables the virus to be genotyped, providing important information on the source of infection. These methods use primers complementary to common sequences present in ¯avivirus RNA [Pierre et al., 1994; Meiyu et al., 1997; Figueiredo et al., 1998] or dengue RNA [Henchal et al., 1991; Sudiro et al., 1997] (so-called universal primer sets) or, alternatively, methods based on speci®c complementary to regions of the dengue viruses [Eldadah et al., 1991; Lanciotti et al., 1992; Tanaka, 1993; Chang et al., 1994; Morita et al., 1994; Seah et al., 1995; Harris et al., 1998; Miagostovich et al., 2000; Wang et al., 2000]. Detection, however, is limited to the viremic period. More recently, tests have incorporated ¯uorogenic probe, so-called TaqMan, technology for the speci®c real-time detection of dengue 2 [Houng et al., 2000] or dengue 1-4 amplicons [Laue et al., 1999; Houng et al., 2001]. Product is detected by a speci®c oligodeoxynucleotide probe that is labeled with 6-carboxy-¯uorescein (FAM). The FAM ¯uorophore is released from the quenching activity of the 6-carboxy-tetramethyl-rhodamine *Correspondence to: David Warrilow, Queensland Health Scienti®c Services, PO Box 594 Archer®eld, Queensland, 4108, Australia. E-mail:
[email protected] Accepted 17 August 2001
525
Rapid Dengue TaqMan Assay
group (TAMRA) on the 30 end by the exonuclease activity of Taq polymerase enabling detection [Holland et al., 1991; Heid et al., 1996]. This technology offers the advantages of being both rapid and potentially quantitative. Dengue virus is not endemic to Australia [Mackenzie et al., 1998]. Outbreaks are initiated by viremic tourists traveling from endemic or epidemic areas to Australia and transmission by the mosquito Aedes aegypti. Hence, to limit infection it is important that vector control measures are implemented as soon as possible. In this study we describe a single-reaction method for the ¯uorogenic probe-based PCR detection of all four dengue serotypes. It is a rapid and speci®c method enabling a reduction in the requirement for individual tests for each of the dengue viruses and, as a result, should enable cost savings for diagnostic laboratories. MATERIALS AND METHODS Viruses and Sera All virus isolates were from the reference collection held at Queensland Health Scienti®c Services. The dengue serotypes 1 and 4 strains used were isolates from viraemic tourists. The dengue 2 strain was the prototype strain New Guinea C and the dengue 3 strain was an isolate from Myanmar (strain PRS225489). Details of clinical samples are given in Table I. Sera from ten non-febrile individuals were used as healthy control material. RNA Extraction Viral RNA was extracted from culture supernatant and serum (140 ml) using the QIAamp viral RNA extraction kit (QIAGEN, Chatsworth, CA) following the instructions provided by the manufacturer. Puri®ed RNA was eluted in a ®nal volume of 80 ml and stored at 808C. Design of Primers and Fluorogenic Probe 0
An alignment of the 3 UTR region of the four dengue serotypes revealed evolutionary conservation that enabled the design of the universal primer and probes TABLE I. Clinical Dengue Samples Tested in This Study Source Australia North Queensland North Queensland South Australia North Queensland South Paci®c East-Timora Palau Vanuatu Tonga Fiji a
Year
Number of samples
2000 1998 1998 1997
3 9 1 1
2000 2000 1999 1998 1997
9 13 5 1 2
Returned Australian army personnel.
(data not shown). DENUNI-F 50 -dAAGGACTAGAGGTTA(G/T)AGGAGACCC-30 and DENUNI-R 50 -dCG(A/T)TCTGTGCCTGGA(A/T)TGATG-30 are nt 10578±10602 and complementary to nt 10665±10685, respectively, of the dengue 2 prototype strain sequence [Irie et al., 1989]. Probe DEN2 4 UTR 50 -d(TCTGGTCTTTCCCAGCGTCAATATGCTGTT)-30 is complementary to nt 10616±10645 and was labeled at its 50 end with a FAM reporter group and at its 30 end with a TAMRA quencher group. Fluorescent Universal Dengue Reverse Transcriptase Polymerase Chain Reaction Dengue RNA was ampli®ed in a standard reaction mixture containing 10 mM Tris HCl, pH 8.3, 50 mM KCl, 5.5 mM MgCl2, 300 mM each of dATP, dCTP, and dGTP, 600 mM dUTP, 30 pmol DENUNI-F, and DENUNI-R primers, 100 pmol DEN 2 4 UTR probe, 60 nM passive reference dye (ROX), 20 U RNase inhibitor (Applied Biosystems, Foster City, CA), 12.5 U Multiscribe reverse transcriptase (Applied Biosystems), 1.25 U AmpliTaq Gold (Applied Biosystems) in a volume of 50 ml. The following cycling conditions were used: 488C for 30 min and 948C for 10 min followed by 50 cycles of 958C for 15 sec and 608C for 1 min. Cycling was conducted on a ABI 7700 Sequence Detector and product was detected by measuring the ¯uorescence signal from the FAM reporter. The threshold for the determination of Ct-value was set at approximately halfway on the logarithmic axis of the maximum value of the DRn of dengue virus positive control. A sample was taken to be positive if it returned a Ct-value < 50 after cycling. Dengue Nested RT-PCR and Serotyping For the determination of serotype, dengue RNA was ampli®ed using the method of a nested RT-PCR assay described previously [Lanciotti et al., 1992]. RESULTS Design of Fluorogenic Universal Dengue Reverse Transcriptase-Polymerase Chain Reaction The 30 non-coding region (30 UTR) has been used previously in the design of primers for the ampli®cation of all four dengue serotypes [Sudiro et al., 1997]. We performed an alignment of this region (data not shown). We identi®ed a region of near-complete evolutionary conservation, equivalent to nt 10616±10645 of the dengue 2 prototype sequence, ¯anked by two regions of lesser evolutionary conservation, nt 10578±10602 and nt 10665±10685, respectively. A ¯uorogenic probe was designed complementary to the highly conserved central sequence, and upstream and downstream primers complementary to the ¯anking sequence. The ¯uorogenic probe was designed such that the predicted thermal stability when bound to its target sequence (Tm 738C) should be greater than that of the predicted
526
Warrilow et al.
thermal stability of the upstream and downstream ampli®cation primers (Tm 60 and 588C, respectively). Test of the Speci®city of Fluorogenic Universal Dengue Reverse Transcriptase-Polymerase Chain Reaction To determine the speci®city before testing on clinical samples, ¯uorogenic universal dengue reverse transcriptase-polymerase chain reaction (FUDRT-PCR) was performed on cell-culture isolates of all four dengue serotypes (Table II). All four dengue serotypes were detected in duplicate tests. No signal was detected in puri®ed RNA samples of the ¯aviviruses Japanese encephalitis virus, Murray Valley encephalitis virus, Kunjin, Stratford, West Nile, Alfuy or Yellow fever. These RNA extracts were routinely used as positive control material for conventional RT-PCR in our laboratory. Hence, FUDRT-PCR was capable of speci®cally detecting all four dengue serotypes. Determination of Sensitivity of FUDRT-PCR on Clinical Samples It was then necessary to determine if FUDRT-PCR was suf®ciently sensitive to detect dengue in clinical samples. The specimens were a mix of sera from acute infections of all four dengue serotypes and included Australian army personnel returning from East Timor, outbreaks in North Queensland from 1997±2000 initiated by tourists traveling from dengue-endemic areas, viraemic tourists, and outbreaks in South Paci®c islands (Table I). The presence of virus was con®rmed by isolation in many of these RT-PCR positive samples or, alternatively, by seroconversion from acute and convalescent sera (Table III). The human sera used were shown to be positive by a conventional nestedPCR [Lanciotti et al., 1992]. This method had proven
TABLE II. FUDRT-PCR On Flavivirus RNAs* Virus
Ct a
Dengue Serotype 1 Serotype 2 Serotype 3 Serotype 4 Other ¯aviviruses Murray Valley encephalitis virus Japanese encephalitis virusb Stratfordc Kunjind West Nilee Alfuy Yellow Feverf *NEG, not detected; Ct > 50. a Dengue strains were tested in duplicate. b Nakayama strain and one isolate. c Two isolates. d Three isolates. e Two isolates. f 17 D vaccine strain.
29, 22, 29, 22,
28 21 30 22
NEG NEG NEG NEG NEG NEG NEG
TABLE III. Sensitivity of FUDRT-PCR On Dengue Positive Clinical Samples* Sample Dengue serotypea 1 2 3 4 Healthy controls
Virus isolationb
Serologyc
Totald (%)
13/13 9/11 3/6 3/3 Ð
1/1 2/2 6/8 Ð Ð
14/14 (100) 11/13 (85) 9/14 (64) 3/3 (100) 0/10 (0)
*Number positive by FUDRT-PCR/number positive samples tested. a Typed by conventional RT-PCR [Lanciotti et al., 1992]. b Con®rmed by virus isolation and serotyping by reactivity with speci®c monoclonal antibodies [Hanna et al., 1998]. c Con®rmed by seroconversion of acute and convalescent sera. d Combined data for viral isolate and serologically con®rmed samples; positives were Ct < 50.
reliable in our hands and enabled serotyping of the virus. FUDRT-PCR detected all four dengue serotypes in clinical samples (Table III). Healthy controls (n 10) were negative. The sensitivity for detection of dengue 1, 2, and 3 was 100%, 85%, and 64%, respectively. Although the number of specimens was limited for dengue 4, the sensitivity of FUDRT-PCR was 100% for this serotype. Hence, FUDRT-PCR had comparable sensitivity to nested RT-PCR for the detection of dengue in clinical samples. DISCUSSION In this study we describe a rapid single assay for the detection of all four dengue serotypes in approximately four hours. The assay detected viral RNA of cultured dengue virus of all four serotypes, but not any of the ¯aviviruses found in Australia, nor some others of medical importance such as yellow fever and West Nile viruses. It was capable of detecting RNA of dengue viruses 1±4 in acute phase human sera obtained from viraemic tourists, Australian domestic outbreaks and outbreaks on islands in the South Paci®c. We developed a ¯uorogenic probe-based assay as it has a number of advantages over standard RT-PCR. First, it avoids the need for electrophoresis of PCR product, saving time. Second, the detection of product by hybridization to ¯uorochrome-labeled probes increases speci®city. Third, as product is detected without the need to open the reaction tube, the risk of contamination by product carry-over is minimized. In our situation the advantages of speed and contamination minimization justify application of this assay over the current nested PCR assay. In our laboratory this assay is used as a screen before a speci®c ¯uorogenicprobe based assay that enables serotyping. Ideally, the sensitivity of a new diagnostic test is determined relative to a so-called ``gold standard'' such as virus isolation on cultured cells. In this work the presence of virus in most of the clinical samples was con®rmed by virus isolation from historical data.
Rapid Dengue TaqMan Assay
Where virus isolation had not been performed, historical seroconversion data were used for con®rmation. The dengue 3 clinical samples were approximately 2 years old and, hence, older on average than most of the material that was positive for the other dengue serotypes. For this reason the conventional nested RT-PCR, performed concurrently with FUDRT-PCR, was used to demonstrate that the viral RNA was intact and, in addition, to determine the serotype of the virus. Good sensitivity was obtained for dengue 1, 2, and 4. The sensitivity of detection for dengue 3 was somewhat lower. All assay data were obtained using freshly extracted RNA. For one of the dengue 3 samples, however, FUDRT-PCR was performed concurrently on freshly extracted material and on an extraction from one year previously. In this case the older material was positive whereas the freshly extracted material was negative, suggesting some degradation of the virus in the serum sample during storage. If the result from the older extraction is incorporated into the data of Table III, then there is a further improvement in sensitivity of the FUDRT-PCR for dengue 3 detection from 64% to 71%. The samples were chosen on the basis that they represent a mixture of dengue strains that were either recently circulating in the South Paci®c or identi®ed in outbreaks in Australia. They had originally been shown to be PCR-positive for one of the four dengue serotypes. The majority of acute phase samples had no antibodies; only two showed low levels of IgM antibody. The paucity of clinical specimens of dengue 4 re¯ects the low frequency of this serotype in the South Paci®c and we were unable to obtain more samples from patients infected with this dengue serotype. The FUDRT-PCR method described is a rapid single assay for the detection of all four dengue serotypes in clinical samples. The sensitivity is comparable to that obtained for a speci®c ¯uorogenic RT-PCR system that requires four individual reactions for each sample [Houng et al., 2001]. It has advantages of reduced costs, signi®cantly reduced turn-around time and much lower risk of contamination over conventional RT-PCR. ACKNOWLEDGMENTS We thank Ina Smith and Michael Lyon for careful reading of the manuscript. REFERENCES Chang GJ, Trent DW, Vorndam AV, Vergne E, Kinney RM, Mitchell CJ. 1994. An integrated target sequence and signal ampli®cation assay, reverse transcriptase-PCR-enzyme-linked immunosorbent assay, to detect and characterize ¯aviviruses. J Clin Microbiol 32:477±483. Edelman R, Nisalak A, Pariyanonda A, Udomsakdi S, Johnsen DO. 1973. Immunoglobulin response and viremia in dengue-vaccinated gibbons repeatedly challenged with Japanese encephalitis virus. Am J Epidemiol 97:207±218. Eldadah ZA, Asher DM, Godec MS, Pomeroy KL, Goldfarb LG, Feinstone SM, Levitan H, Gibbs CJ Jr, Gajdusek DC. 1991. Detection of ¯aviviruses by reverse-transcriptase polymerase chain reaction. J Med Virol 33:260±267.
527 Figueiredo LT, Batista WC, Kashima S, Nassar ES. 1998. Identi®cation of Brazilian ¯aviviruses by a simpli®ed reverse transcription-polymerase chain reaction method using Flavivirus universal primers. Am J Trop Med Hyg 59:357±362. Gubler DJ, Suharyono W, Tan R, Abidin M, Sie A. 1981. Viremia in patients with naturally acquired dengue infection. Bull World Health Organ 59:623±630. Hanna JN, Ritchie SA, Merritt AD, van den Hurk AF, Phillips DA, Sera®n IL, Norton RE, McBride WJH, Glesson FV, Poidinger M. 1998. Two contiguous outbreaks of dengue 2 in north Queensland. Med J Aust 168:221±225. Harris E, Roberts TG, Smith L, Selle J, Kramer LD, Valle S, Sandoval E, Balmaseda A. 1998. Typing of dengue viruses in clinical specimens and mosquitoes by single-tube multiplex reverse transcriptase PCR. J Clin Microbiol 36:2634±2639. Heid CA, Steven J, Livak KJ, Williams PM. 1996. Real time quantitative PCR. Genome Res 6:986±994. Henchal EA, Polo SL, Vorndam V, Yaemsiri C, Innis BL, Hoke CH. 1991. Sensitivity and speci®city of a universal primer set for the rapid diagnosis of dengue virus infections by polymerase chain reaction and nucleic acid hybridization. Am J Trop Med Hyg 199 45:418±428. Holland PM, Abramson RD, Watson R, Gelfund DH. 1991. Detection of speci®c polymerase chain reaction product by utilizing the 50 ! 30 exonuclease activity of Thermus aquaticus DNA polymerase. Proc Natl Acad Sci USA 88:7271±7280. Houng HH, Hritz D, Kanesa-thasan N. 2000. Quantitative detection of dengue 2 virus using ¯uorogenic RT-PCR based on 30 -noncoding sequence. J Virol Methods 86:1±11. Houng HH, Chen RC, Vaughn DW, Kanesa-thasan N. 2001. Development of a ¯uorogenic RT-PCR system for quantitative identi®cation of dengue virus serotypes 1-4 using conserved and serotype-speci®c 30 noncoding sequences. J Virol Methods 95:19± 32. Innis BL. 1997. Antibody responses to dengue virus infection. In: Gubler DJ, Kuno G, editors. Dengue and dengue hemorrhagic fever. New York: CAB International. p 221±243. Innis BL, Nisalak A, Nimmannitya S, Kusalerdchariya S, Chongswasdi V, Suntayakorn S, Puttisiri P, Hoke CH. 1989. An enzymelinked immunoassay to characterize dengue infections where dengue and Japanese encephalitis co-circulate. Am J Trop Med Hyg 40:418±427. Irie K, Mohan PM, Sasaguri Y, Putnak R, Padmanabhan R. 1989. Sequence analysis of the cloned dengue virus type 2 genome (New Guinea-C strain). Gene 75:197±211. Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV. 1992. Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol 30:545±551. Laue T, Emmerich P, Schmitz H. 1999. Detection of dengue virus RNA in patients after primary or secondary dengue infection by using the TaqMan automated ampli®cation system. J Clin Microbiol 37: 2543±2547. Mackenzie JS, Broom AK, Hall RA, Johansen CA, Lindsay MD, Phillips DA, Ritchie SA, Russell RC, Smith DW. 1998. Arboviruses in the Australian region. Comm Dis Intell (Aust) 22:93±100. Meiyu F, Huosheng C, Cuihua C, Xiaodong T, Lianhua J, Yifei P, Weijun C, Huiyu G. 1997. Detection of ¯aviviruses by reverse transcriptase-polymerase chain reaction with the universal primer set. Microbiol Immunol 41:209±213. Miagostovich MP, dos Santos FB, Gutierrez CM, Riley LW, Harris E. 2000. Rapid subtyping of dengue virus serotypes 1 and 4 by restriction site-speci®c PCR. J Clin Microbiol 38:1286±1289. Morita K, Maemoto T, Honda S, Onishi K, Murata M, Tanaka M, Igarashi A. 1994. Rapid detection of virus genome from imported dengue fever and dengue hemorrhagic fever patients by direct polymerase chain reaction. J Med Virol 44:54±58. Pierre V, Drouet MT, Deubel V. 1994. Identi®cation of mosquito-borne ¯avivirus sequences using universal primers and reverse transcription/polymerase chain reaction. Res Virol 145:93±104. Pinheiro FP, Corber SJ. 1997. Global situation of dengue and dengue hemorrhagic fever, and its emergence in the Americas. World Health Stat Q 50:161±168. Rodhain F, Rosen L. 1997. Mosquito vectors and dengue virusvector relationships. In: Gubler DJ, Kuno G, editors. Dengue and dengue hemorrhagic fever. New York: CAB International. p 45± 60.
528 Seah CL, Chow VT, Tan HC, Can YC. 1995. Rapid, single-step RTPCR typing of dengue viruses using ®ve NS3 gene primers. J Virol Methods 51:193±200. Sudiro TM, Ishiko H, Green S, Vaughn DW, Nisalak A, Kalayanarooj S, Rothman AL, Raengsakulrach B, Janus J, Kurane I, Ennis FA. 1997. Rapid diagnosis of dengue viremia by reverse transcriptasepolymerase chain reaction using 30 -noncoding region universal primers. Am J Trop Med Hyg 56:424±429.
Warrilow et al. Tanaka M. 1993. Rapid identi®cation of ¯avivirus using the polymerase chain reaction. J Virol Methods 41:311±322. Wang WK, Lee CN, Kao CL, Lin YL, King CC. 2000. Quantitative competitive reverse transcription-PCR for quanti®cation of dengue virus RNA. J Clin Microbiol 38:3306±3310. Westaway EG. 1968. Greater speci®city of the 19S and 7S antibodies on hemagglutination-inhibition tests with closely related group B arboviruses. Nature 219:78±79.
