Am. J. Trop. Med. Hyg., 67(6), 2002, pp.

656661 Copyright 2002 by The American Society of Tropical Medicine and Hygiene

DETECTION AND STABILITY OF JAPANESE ENCEPHALITIS VIRUS RNA AND VIRUS VIABILITY IN DEAD INFECTED MOSQUITOES UNDER DIFFERENT STORAGE CONDITIONS
CHERYL A. JOHANSEN, ROY A. HALL, ANDREW F. VAN DEN HURK, SCOTT A. RITCHIE, AND JOHN S. MACKENZIE Department of Microbiology and Parasitology, School of Molecular and Microbial Sciences, The University of Queensland, St. Lucia, Queensland, Australia; Tropical Public Health Unit, Queensland Health, Cairns, Queensland, Australia

Abstract. A semi-nested polymerase chain reaction (PCR) was evaluated for detection of Japanese encephalitis (JE) virus in infected mosquitoes stored under simulated northern Australian summer conditions. The effect of silica gel, thymol, and a combination of the two on RNA stability and virus viability in dead mosquitoes were also examined. While JE virus RNA was relatively stable in mosquitoes held for up to 14 days after death, viable virus was not detected after day 1. Thymol vapor inhibited fungal contamination. Detection of single mosquitoes infected with JE virus in large pools of mosquitoes was also investigated. Single laboratory-infected mosquitoes were detected in pools of 200 mosquitoes and in pools diluted to 0.2/100 and 0.1/100 mosquitoes, using the semi-nested PCR. However, the ability to detect live virus decreased as pool size increased. The semi-nested PCR proved more expensive than virus isolation for pools of 100 mosquitoes. However, the semi-nested PCR was faster and more economical using larger pools. Results indicate that surveillance of JE virus in mosquitoes using the semi-nested PCR is an alternative to monitoring seroconversions in sentinel pigs. INTRODUCTION Following the first appearance of Japanese encephalitis (JE) virus in the Torres Strait in 1995,1 control and prevention measures implemented have included surveillance of JE virus activity in the Torres Strait and northern Queensland, Australia by detection of seroconversions to JE virus in sentinel pigs.2 The sentinel system has since detected JE virus activity in the Torres Strait in all years except 1999, and detected activity in Cape York in 199824 (Lee JM, Smith GA, unpublished data). However, the use of sentinel pigs for surveillance of JE virus activity in Cape York and the Torres Strait is expensive, and since pigs are amplifying hosts of JE virus, they pose a potential risk to the community. Thus, other methods of surveillance are urgently required. Although virus detection in field-collected mosquitoes is a gold standard of surveillance, the routine collection of mosquitoes for virus isolation is not practical in this instance because the remoteness of Cape York and the Torres Strait prohibit daily servicing of standard light traps. A new mosquito trap developed by American Biophysics Pty., Ltd., (East Greenwich, RI) overcomes some of these difficulties by using propane gas to produce CO2, allowing it to function for up to three weeks without a propane refill.5 By using these traps together with alternative methods of JE virus detection, it may be possible to develop a safer and more effective surveillance system instead of using sentinel pigs. With respect to viral detection methods, both West Nile (WN)6 and St. Louis encephalitis (SLE)7 viruses have been detected with a reverse transcriptasepolymerase chain reaction (RT-PCR) in dead mosquitoes held at 20C and 27C for 14 days and 20 days, respectively. However, studies of the stability of JE virus and viral RNA in dead mosquitoes held in humid, hot conditions typical of the north Queensland summer have not been conducted. Furthermore, to rapidly process the large numbers of mosquitoes required for surveillance purposes, mosquitoes would need to be examined in large pool sizes (> 100), as was recently demonstrated for Rift Valley fever (RVF) virus.8 This paper describes investigations into the stability of JE virus RNA and virus viability in dead mosquitoes stored on desiccating and anti-fungal reagents, as well as detection of infectious JE virus and viral RNA in large pools of mosquitoes using standard virus isolation techniques9 and a seminested RT-PCR. MATERIALS AND METHODS Mosquito infections. Laboratory-reared Culex sitiens (colonized from Coomera Island, Queensland) were infected with JE virus for determination of RNA stability and virus viability in dead mosquitoes and pool size experiments. Two-to-three day-old mosquitoes were starved for 24 hours prior to infection. Mosquitoes were allowed to feed through a pig intestine membrane feeder containing approximately 3 mL of blood/ virus mixture comprising stock JE virus (TS3306) diluted in heparinized (25 units/mL) rabbit blood and 1% sucrose. The final titer of the blood/virus mixture was between 106.21 and 107.61 log10 tissue culture infective dose50 (TCID50)/mL. Mosquitoes were allowed to feed for 3 hours, after which nonengorged mosquitoes were discarded. Mosquitoes were held at 28C at a relative humidity of 7075% with a 12-hour light/day photoperiod with 45-minute crepuscular periods. Apple slices and 10% honey water were offered as a nutrient source. After a 12-day extrinsic incubation, mosquitoes were killed with CO2. The head and salivary glands from each mosquito were removed and assayed by immunofluorescence10 to confirm infection. Infected mosquitoes were stored directly at 70C for the pool size experiment or placed into an incubator for up to 14 days post-death to investigate RNA stability and virus viability in dead mosquitoes. Stability of RNA and virus viability in dead mosquitoes. To determine whether different storage substrates were beneficial or detrimental to virus viability or RNA stability, four substrates were tested: 1) 300 grams of eight-mesh silica gel desiccant (Ajax FineChem, Sydney, New South Wales, Australia); 2) cotton wicks saturated with the anti-fungal agent 5-methyl-2-isopropylphenol (thymol; Crown Scientific, Fortitude Valley, Queensland, Australia); 3) eight-mesh silica gel plus thymol-saturated wicks; and 4) no substrate. Thymol wicks were prepared by placing cotton wicks (Commonwealth Dental Supplies, Brisbane, Queensland, Australia) into a saturated solution of thymol crystals in acetone. The wicks

656

DETECTION OF JE VIRUS RNA IN DEAD MOSQUITOES

657

were then removed and allowed to dry in a fume cupboard before being stored in a sealed jar. Experiments using thymolsaturated wicks were performed separately due to the pervasive nature of the thymol vapor. Silica gel was replaced when the humidity inside the container reached 40%. Replicates of single infected mosquitoes were placed in petri dishes inside rectangular (17 11.5 5.5 cm) plastic Reko Food storage containers (Reko Pty., Ltd., Lisarow, New South Wales, Australia) containing one of the substrates. Holes (5-cm diameter) were cut into the lid of each container, which was placed directly below an operating encephalitis vector surveillance trap11 in a humidified environmental incubator that maintained storage conditions at a relative humidity of 7080% and a temperature between 28C (7:00 PM to 7:00 AM) and 32C (7:00 AM to 7:00 PM). These conditions were chosen to simulate environmental conditions that mosquitoes would be exposed to when collected at Cape York or the Torres Strait, based on weather records from the Australian Bureau of Meteorology. Infected mosquitoes were removed from each container at days 0, 1, 2, 4, 7, and 14, then stored at 70C until use. Individual infected mosquitoes and uninfected negative control mosquitoes were subsequently homogenized9 in 600 L of M199 cell culture medium (containing 2% fetal bovine serum). Homogenate was used directly for RNA extractions and clarified briefly by centrifugation at 2,000 rpm for two minutes for titration for infectious virus content. Negative control uninfected mosquitoes were homogenized and assayed concurrently with each group of samples processed. Virus assay. Fifty microliters of undiluted clarified homogenate was added to duplicate wells of a confluent 96-well plate of C6/36 cells to determine if viable virus and fungal contaminants were present. To determine the virus titer in each mosquito, a 100- L aliquot of homogenate was diluted 1/10 in M199 (containing 2% fetal bovine serum), filtered through a 0.2- m syringe filter to remove potential fungal contaminants, and 10-fold dilutions were inoculated onto the remaining wells of the 96-well plate seeded with C6/36 cells. The C6/36 cells were incubated at 28C for six days in a 5% CO2-enhanced environment before being fixed and assayed by tissue culture enzyme immunoassay (TC/EIA)12 using the JE virus-specific monoclonal antibody 995.13 The infectious titers of individual mosquitoes were determined using a formula for calculating 50% end points14 and expressed as the log10 TCID50/mosquito. Extraction of RNA. RNA was extracted from 100 L of unclarified homogenate immediately following homogenization. Extractions of RNA from mosquito homogenates were carried out using TRIZOL Reagent (Gibco-BRL Life Technologies, Gaithersburg, MD) according to the manufacturers instructions. The RNA pellet was air-dried prior to resuspension in 20 L of diethylpyrocarbonate (DEPC)treated double-distilled water and stored at 70C. Pellets that did not completely dissolve were heated for five minutes at 55C prior to storage at 70C. The RNA was assayed by RT-PCR and semi-nested PCR as described in this report. RT-PCR and semi-nested PCR. A single-tube RT-PCR was used to minimize the number of manipulations required for each sample, thereby helping to avoid sample contamination. An additional semi-nested PCR was used to enhance the sensitivity of the JE virus detection system. Positive and negative controls were included in each assay to confirm adequate

assay performance. The primers used in the RT-PCR were forward primer FV-128 (5 -CCGGGCTGTCAATATGCT3 , designed by J. Conlon; The University of Queensland, Brisbane, Queensland, Australia) and reverse primer prMR3 (5 -CATGAGGTATCGCGTGGC-3 ). In the semi-nested PCR, FV-128 was also used as the forward primer; however, FV-prM (5 -CACCAGCAGTCAATGTCTTC-3 ) and JE659 (5 -CACCAGCAATCCACGTCCTC-3 ) (also designed by J. Conlon) were used as the reverse primers. The flavivirus group-reactive primers FV-128, FV-prM, and JE659 were designed using nucleotide sequence alignments of JE virus strains FU, M15, and M40, Murray Valley encephalitis (MVE) virus strain MVE/1/51, Kunjin (KUN) virus strain MRM61C, WN virus strain Wengler, and SLE virus strain MSI.7, and were based in part on dengue consensus primers described elsewhere.15 Primer prMR3 was designed using nucleotide sequence alignments of isolates of JE virus strains FU, K94P05, Nakayama, and JKT6468, which represented four different genotypes,1618 MVE virus, and KUN virus. Primers annealed to regions encoding the capsid (C) to membrane (M) proteins in the RT-PCR and C to pre-membrane (prM) regions in the semi-nested PCR. This region was chosen because it has elements conserved in several mosquitoborne flaviviruses,15 and is commonly used for nucleotide sequence analysis of isolates of JE virus.9,1620 The two reverse primers were used in the semi-nested PCR to maximize annealing to the genome of different genotypes of JE virus and to other flaviviruses, thereby improving sensitivity. The primers described were capable of detecting representatives of the four genotypes of JE virus in addition to MVE, KUN, WN, Kokobera, and Stratford viruses. Working stocks of primers (Sigma-Aldrich Pty., Ltd., Castle Hill, New South Wales, Australia) (100 ng/ L) were made in DEPC-treated doubledistilled water and stored at 20C. In addition to noninfected individual mosquito or pools of mosquitoes, DEPCtreated water and JE virus stock (TS3306) were used as negative and positive controls, respectively. Reverse transcription and amplification was carried out using two microliters of denatured RNA in a 25- L reaction volume containing Red Hot buffer (Advanced Biotechnologies, Surrey, United Kingdom), 50 ng of each primer (FV128 and prMR3), one unit of avian myeloblastosis virus reverse transcriptase (Promega, Madison, WI), 2.5 units of Red Hot Taq polymerase (Advanced Biotechnologies), four units of RNasin (Promega), 1.5 mM MgCl2, 0.4 mM dNTPs (Biotech International, Ltd., Perth, Western Australia), and 0.4 mM dithiothreitol (Promega). Reverse transcription was carried out at 55C for one hour before beginning a touchdown procedure21 that included denaturation for one minute at 94C, followed by two cycles of denaturation at 94C for 30 seconds, annealing at 60C for 30 seconds, and elongation at 72C for one minute. The annealing temperature was then varied for two cycles of annealing temperatures at 58C, 56C, 54C, and 52C. Finally, 20 cycles of the PCR were carried out with a denaturing temperature of 94C for 30 seconds, annealing at 50C for 30 seconds, and elongation at 72C for one minute, followed by a final elongation at 72C for 10 minutes. The resulting PCR products were visualized after electrophoresis on a 1.5% agarose gel in Tris-acetate-EDTA (TAE) buffer. The expected DNA product had a size of 714 nucleotides based on the nucleotide sequence of JE virus strain FU.20 The semi-nested PCR was carried out in a 25- L volume of

658

JOHANSEN AND OTHERS

the manufacturers (Advanced Biotechnologies) buffer containing 5 L of primary RT-PCR product (after diluting 1:100 in DEPC-treated water), 50 ng of forward primer (FV-128), 50 ng of each reverse primer (JE659 and FV-prM), 0.4 mM dNTPs, 1.5 mM MgCl2, and 2.5 units of Red Hot Taq polymerase. The touchdown PCR included denaturation at 94C for one minute, followed by two cycles of denaturation at 94C for 30 seconds, annealing at 56C for 30 seconds, and elongation at 72C for one minute. The annealing temperature was then varied for two cycles of annealing temperatures at 54C, 52C, 50C, and 48C. Finally, 20 cycles of the PCR were carried out with a denaturing temperature of 94C for 30 seconds, annealing at 46C for 30 seconds, and elongation at 72C for one minute, followed by a final elongation at 72C for 10 minutes. The resulting PCR products were visualized after electrophoresis on a 1.5% agarose gel in TAE buffer. The expected DNA product had a size of 554 nucleotides based on the nucleotide sequence of JE virus strain FU.20 Detection of JE virus and RNA in large mosquito pools. Single, confirmed infected mosquitoes were added to different sized pools of colony-reared non-infected mosquitoes to determine if individual infected mosquitoes could be detected in pools of various sizes. Single JE virus-infected mosquitoes were added to pools of up to 199 non-infected laboratory-reared mosquitoes of mixed species. Five replicates and single negative control pools were examined per pool size. Pool sizes, volume of diluent, and homogenization times were as follows: one infected mosquito triturated in 600 L for 1.5 minutes, 25 mosquitoes triturated in 2.5 mL for 1.5 minutes, 50 mosquitoes triturated in 5.0 mL for 3.0 minutes, 100 mosquitoes triturated in 5.0 mL for 3.0 minutes, and 200 mosquitoes triturated in 5.0 mL for 5 minutes. Pool sizes > 200 mosquitoes were too large for the 5-mL tubes used to homogenize smaller mosquito pools. To simulate pools of 500 and 1,000 mosquitoes, 1:5 and 1:10 dilutions of pools of 100

mosquitoes containing single infected mosquitoes were made in homogenates from negative control pools of 100 mosquitoes following the procedures of Armstrong and others,22 such that the final concentrations were 0.2/100 and 0.1/100 mosquitoes. Samples were homogenized as previously described and RNA was extracted and tested by RT-PCR and semi-nested PCR. Furthermore, the infectious titer of each pool was determined by 10-fold dilutions using C6/36 cells and the TC/EIA as previously described, and expressed as the log10 TCID50/mL. In addition, 33 confirmed JE virus positive pools from Papua New Guinea9 and the Torres Strait4,18,19 (Johansen C, unpublished data) were assayed by semi-nested PCR to further validate the procedure. RESULTS The viability of JE virus decreased rapidly after the death of laboratory-infected mosquitoes stored in simulated climatic conditions of northern Queensland (Table 1). The mean SD infectious titer of mosquitoes (minus head and salivary glands) immediately after death and prior to storage was 3.39 0.88 log10 TCID50/mosquito (no substrate and silica gel alone) or 3.71 0.37 log10 TCID50/mosquito (thymol alone and thymol plus silica gel). However, only one mosquito (1.1%) had detectable virus in the TC/EIA 1 day after death. Fungal contamination was observed in some samples stored without substrate, or on silica gel alone; no fungal contamination was detected in mosquitoes held in the presence of thymol vapor or thymol vapor plus silica gel. The RNA appeared to remain relatively stable in mosquitoes stored for up to 14 days after death. All mosquitoes held in thymol alone or thymol plus silica gel had detectable viral RNA in the RT-PCR and semi-nested PCR, while viral RNA was detected in 85% and 90% of mosquitoes stored without a preserving substrate by RT-PCR and semi-nested PCR, re-

TABLE 1 Effect of different storage substances on infectivity, fungal contamination, and RNA stability in single mosquitoes infected with Japanese encephalitis virus, dead for up to 14 days
Substrate No. of days* since death TC/EIA Fungal contamination RT-PCR Semi-nested PCR

No substrate No substrate No substrate No substrate No substrate Silica gel alone Silica gel alone Silica gel alone Silica gel alone Silica gel alone Thymol alone Thymol alone Thymol alone Thymol alone Thymol alone Thymol and silica Thymol and silica Thymol and silica Thymol and silica Thymol and silica

gel gel gel gel gel

1 2 4 7 14 1 2 3 7 14 1 2 4 7 14 1 2 3 7 14

0 0 0 0 0 1/4 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1/4 0 0 0 0 1/4 1/4 0 0 1/4 0 0 0 0 0 0 0 0 0 0

2/4 4/4 4/4 4/4 3/4 3/4 4/4 4/4 3/4 4/4 5/5 5/5 5/5 5/5 5/5 5/5 5/5 5/5 5/5 5/5

3/4 4/4 4/4 4/4 3/4 4/4 4/4 4/4 4/4 4/4 5/5 5/5 5/5 5/5 5/5 5/5 5/5 5/5 5/5 5/5

* Infected mosquitoes were held on different substrates under the fan of an encephalitis vector surveillance trap in a humidified incubator at 2832C and a relative humidity of 7080%. TC/EIA tissue culture enzyme immunoassay. The mean SD log10 50% tissue culture infectious doses (TCID50) per infected mosquito were 3.39 0.88 (no substrate and silica gel alone) and 3.71 0.37 (thymol alone and thymol and silica gel). Numerator number of samples positive, denominator number of samples tested. Fungal contamination detected macroscopically and microscopically in undilated homogenate on C6/36 monolayers. RT-PCR reverse transcriptasepolymerase chain reaction.

DETECTION OF JE VIRUS RNA IN DEAD MOSQUITOES

659

spectively. Similarly, 90% of the mosquitoes stored in silica gel alone had detectable RNA in the RT-PCR, although virus RNA was detected in all such samples after the additional amplification cycles in the semi-nested PCR. No PCR products were detected in negative control samples. The infectious titer of mosquito pools containing single laboratory-infected mosquitoes (mean SD log10 TCID50/ infected mosquito 3.71 0.37) decreased sharply with increasing pool size (Table 2). Despite a two-fold dilution factor between pools of 25 and 50 mosquitoes, the infectious titer decreased by more than one log10 TCID50/mL. The infectious titer continued to decrease at a similar rate as pool size increased from 50 to 200 mosquitoes, which were all homogenized in a volume of 5 mL. No viable virus was detected in simulated pools of 500 and 1,000 mosquitoes. There was also a decrease in the ability to detect JE virus RNA by RT-PCR when pool size was increased. However, all samples (including 500 and 1,000 mosquitoes) were positive by the seminested PCR, with little difference in PCR product intensity. Furthermore, the semi-nested PCR product was of sufficient intensity to obtain nucleotide sequence information and perform phylogenetic analyses of positive pools of mosquitoes. No PCR products were detected in the negative controls. In addition, 100% of the previously confirmed JE virus infected pools of mosquitoes from Papua New Guinea and the Torres Strait were positive by the semi-nested PCR. DISCUSSION Irrespective of the preserving substrate used for storing mosquitoes (silica gel, thymol wicks, silica gel and thymol wicks, or no substrate), viral RNA from JE virus-infected mosquitoes was detected in all but two instances by the seminested PCR. The two negative mosquitoes were not stored on a preserving substrate. Furthermore, 100% of the mosquitoes stored on thymol wicks (either alone or with silica gel) were detected using the RT-PCR alone, compared with 85% and 90% of mosquitoes stored without a preserving substrate or on silica gel alone, respectively. Thymol wicks appeared to inhibit development and growth of mold in the humid conditions the mosquitoes were exposed to, and they also may also have improved the stability of RNA in dead mosquitoes. A preliminary study found that there was no significant different in the number of Culex mosquitoes (t 0.001, P > 0.05; n
TABLE 2 Effect of increasing pool size on infectious titer and detection of Japanese encephalitis virus RNA by RT-PCR and seminested PCR
Pool size* Log10 TCID50/mL RT-PCR Semi-nested PCR

1/25 1/50 1/100 1/200 0.2/100 0.1/100

4.04 1.46 2.98 1.98 1.96 1.67 Negative Negative Negative

5/5 5/5 4/5 4/5 2/5 1/5

5/5 5/5 5/5 5/5 5/5 5/5

* One infected mosquito in pools of uninfected mosquitoes; pools of 0.2/100 and 0.1/100 comprised 1:5 and 1:10 dilutions of each infected pool of 1/100 mosquitoes diluted in homogenate from pools of 100 non-infected mosquitoes and simulated pools of 500 and 1,000 mosquitoes, respectively. Values are the mean SD. The log10 50% tissue culture infectious dose (TCID50) per single infected mosquito was 3.71 0.37. RT-PCR reverse transcriptasepolymerase chain reaction. Numerator number of samples positive, denominator number of samples tested.

10) or total number of mosquitoes (t 1.327, P > 0.05; n 10) collected in Centers for Disease Control mosquito traps23 with or without thymol vapor. These results suggest that thymol wicks are suitable for preventing fungal contamination when storing mosquitoes in the field for long periods of time. The failure to isolate virus from all but one dead mosquito stored from day 1 post-death indicates that virus isolation is unsuitable for detection of JE virus in long-dead mosquitoes held in hot, humid conditions, and is consistent with findings for other flaviviruses.6,7,24 The semi-nested PCR was potentially capable of detecting one JE virus-infected mosquito in 1,000 non-infected mosquitoes. In contrast, virus infectivity decreased dramatically as pool size increased, until it was no longer detectable in pools of 200 mosquitoes. This indicates that the semi-nested PCR is superior to detection by virus isolation and TC/EIA when using large pools of mosquitoes. Numerous PCR assays have been previously described that detect arboviruses in mosquitoes in pool sizes between 50 and 100 mosquitoes.7,22,2530 However, few have looked at larger pool sizes. Armstrong and others22 were able to detect the equivalent of 0.1/100 mosquitoes infected with eastern equine encephalomyelitis (EEE) virus by pooling aliquots from homogenates of pools of 100 mosquitoes, and Ross River virus-infected mosquitoes were detected in pools of up to 500 mosquitoes (Sellner LN, unpublished data). More recently, Jupp and others8 found that single mosquitoes infected with RVF virus could be detected in pools of up to 600 non-infected mosquitoes, and the equivalent of 1/16,000 when pools of 1/100 mosquitoes were diluted in tissue culture medium. Our results confirm large pools of mosquitoes or diluted mosquito pools can be tested using the semi-nested PCR. Advantages of processing large pools of mosquitoes include reduced sample processing time and cost. However, processing of large pool sizes would result in a loss of valuable information about infection rates in mosquitoes during outbreaks of arboviral disease. Furthermore, processing of large pool sizes would increase the possibility that more than one flavivirus could be present, prohibiting confirmation of virus identification by nucleotide sequence analysis. To avoid this, mosquitoes could be processed in pools of 200, and 20- L aliquots from five pools of 200 mosquitoes could be pooled for RNA extraction and subsequent assay by the semi-nested PCR, enabling processing of up to 1,000 mosquitoes per sample for JE virus detection. Any samples deemed positive may then be re-tested individually to confirm infection, undertake nucleotide sequencing of the viral RNA and calculate minimum infection rates. A similar procedure has been previously described for EEE virus.22 Detection of JE virus RNA directly from mosquito pools enables phylogenetic analysis of strains of JE virus without possible nucleotide base changes following passage through cell culture.31,32 If mosquito pools are found to contain more than one flavivirus, virus-specific primers (Conlon J, unpublished data) can be used to identify the flaviviruses present. One of the major advantages of processing large pools of mosquitoes is the reduced sample processing time and cost. The cost of processing 100 pools of 100 mosquitoes for virus isolation and identification by TC/EIA is less expensive, at least in terms of consumables, reagents, and labor than RNA extractions and the semi-nested PCR (Table 3). However, the time required to extract RNA and test 100 pools of mosqui-

660

JOHANSEN AND OTHERS

TABLE 3 Comparison of cost ($US)* and time to process 100 pools of 100 mosquitoes for virus isolation and semi-nested PCR
Variable Virus isolation Semi-nested PCR

Centre, Nedlands 6009, Western Australia, Telephone: 61-8-93464656, Fax: 61-8-9346-2912, E-mail: cjohanse@cyllene.uwa.edu.au

REFERENCES
1. Hanna JN, Ritchie SA, Phillips DA, Shield J, Bailey MC, Mackenzie JS, Poidinger M, McCall BJ, Mills PJ, 1996. An outbreak of Japanese encephalitis in the Torres Strait, Australia, 1995. Med J Aust 165: 256260. 2. Shield J, Hanna J, Phillips D, 1996. Reappearance of the Japanese encephalitis virus in the Torres Strait, 1996. Commun Dis Intell 20: 191. 3. Hanna JN, Ritchie SA, Phillips DA, Lee JM, Hills SL, van den Hurk AF, Pyke AT, Johansen CA, Mackenzie JS, 1999. Japanese encephalitis in north Queensland, Australia, 1998. Med J Aust 170: 533536. 4. Pyke AT, Williams DT, Nisbet DJ, van den Hurk AF, Taylor CT, Johansen CA, Macdonald J, Hall RA, Simmons RJ, Mason RJV, Lee JM, Ritchie SA, Smith GA, Mackenzie JS, 2001. The appearance of a second genotype of Japanese encephalitis virus in the Australasian region. Am J Trop Med Hyg 65: 747 753. 5. Kline DL, 1999. Comparison of two American Biophysics mosquito traps: the professional and a new counterflow geometry trap. J Am Mosq Control Assoc 15: 276282. 6. Turell MJ, Spring AR, Miller MK, Cannon CE, 2002. Effect of holding conditions on the detection of West Nile viral RNA by reverse-transcriptase polymerase chain reaction from mosquito (Diptera: Culicidae) pools. J Med Entomol 39: 13. 7. Kramer LD, Chiles RE, Do TD, Fallah HM, 2001. Detection of St. Louis encephalitis and western equine encephalomyelitis RNA in mosquitoes tested without maintenance of a cold chain. J Am Mosq Control Assoc 17: 213215. 8. Jupp PG, Grobbelaar AA, Leman PA, Kemp A, Dunton RF, Burkot TR, Ksiazek TG, Swanepoel R, 2000. Experimental detection of Rift Valley fever virus by reverse transcriptionpolymerase chain reaction assay in large samples of mosquitoes. J Med Entomol 37: 467471. 9. Johansen CA, van den Hurk AF, Ritchie SA, Zborowski P, Nisbet DJ, Paru R, Bockarie MJ, Macdonald J, Drew AC, Khromykh TI, Mackenzie JS, 2000. Isolation of Japanese encephalitis virus from mosquitoes (Diptera: Culicidae) in the Western Province of Papua New Guinea, 19971998. Am J Trop Med Hyg 62: 631638. 10. Beaty BJ, Thompson WH, 1975. Emergence of La Crosse virus from endemic foci. Fluorescent antibody studies of overwintered Aedes triseriatus. Am J Trop Med Hyg 24: 685691. 11. Rohe D, Fall RP, 1979. A miniature battery powered CO2 baited light trap for mosquito borne encephalitis surveillance. Bull Soc Vector Ecol 4: 2427. 12. Broom AK, Hall RA, Johansen CA, Oliveira N, Howard MA, Lindsay MD, Kay BH, Mackenzie JS, 1998. Identification of Australian arboviruses in inoculated cell cultures using monoclonal antibodies in ELISA. Pathology 30: 286288. 13. Gould EA, 1991. Antigenicity of flaviviruses. Arch Virol (Suppl 1): 137152. 14. Reed LJ, Meunch H, 1938. A simple method for estimating fifty percent end points. Am J Hyg 27: 493497. 15. 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: 545551. 16. Chen WR, Tesh RB, Rico-Hesse R, 1990. Genetic variation of Japanese encephalitis virus in nature. J Gen Virol 71: 2915 2922. 17. Chen WR, Rico-Hesse R, Tesh RB, 1992. A new genotype of Japanese encephalitis virus from Indonesia. Am J Trop Med Hyg 47: 6169. 18. Ritchie SA, Phillips D, Broom A, Mackenzie J, Poidinger M, van den Hurk A, 1997. Isolation of Japanese encephalitis virus from Culex annulirostris in Australia. Am J Trop Med Hyg 56: 8084. 19. Johansen CA, van den Hurk AF, Zborowski P, Phillips DA, Pyke AT, Mackenzie JS, Ritchie SA, 2001. Entomological inves-

Cost (consumables/reagents) Labor costs (person hours) Time (preliminary identification) Additional time (confirmation)#

$159 $380 814 days 713 days

$269 $422 4 days 3 days

* Based on currency exchange rates of 0.5260 US$/unit on March 15, 2002. Includes sample homogenization, inoculation onto 96-well plates of C6/36 cells, and assay by tissue culture enzyme immunoassay (TC/EIA). Includes sample homogenization, extraction of RNA, reverse transcriptasepolymerase chain reaction (RT-PCR), and semi-nested PCR. Based on work time (excluding incubations) for a higher education worker (HEW) Level 601 (Research Assistant) at the University of Queensland as of July 1, 2001 (salary every two weeks $844.76). Time to tentatively identify Japanese encephalitis (JE) virus in 100 pools of 100 mosquitoes. # Time to confirm the presence of JE virus in 100 pools of 100 mosquitoes by repeating isolation or identification by TC/EIA or RT-PCR/semi-nested PCR/nucleotide sequencing, respectively.

toes is substantially less than for virus isolation and TC/EIA. Furthermore, given that JE virus RNA can potentially be detected in pools of up to 1,000 mosquitoes, detection by the semi-nested PCR would prove substantially more economical than virus isolation. Nawrocki and others27 reached a similar conclusion during a study investigating the detection of SLE virus in pools of mosquitoes. The primary objective of this study was to investigate the viability of a PCR-based system for detecting JE virus in infected mosquitoes as a component of a mosquito-based surveillance system. The method of mosquito collection and sample processing for surveillance of JE virus described in this paper and elsewhere (Johansen CA and others, unpublished data) is currently being field-tested using laboratoryinfected mosquitoes. Given that the semi-nested PCR described in this paper can detect numerous flaviviruses, including MVE and Kunjin viruses, the assay may also prove useful for surveillance of other flaviviruses enzootic in northern Australia.
Acknowledgments: We thank James Conlan and Dr. Helle Bielefeldt-Ohmann for providing primer sequences; Major Bob Cooper and Cassie Jensen (Army Malaria Institute, Brisbane, Queensland, Australia) for kindly supplying colony mosquitoes for laboratory infections and for assistance with salivary gland dissections; and Professor Brian Kay (Queensland Institute of Medical Research, Brisbane, Queensland, Australia) and Kay Marshall for providing mosquitoes for pool size construction. We also thank Dr. Greg Smith and Alyssa Pyke (Virology, Queensland Health Scientific Services, Brisbane, Queensland, Australia) for allowing us to use their laboratory for mosquito infections; and Kerry Huxham (Australian Quarantine Inspection Service, Mareeba, Queensland, Australia) for providing the method for preparing thymol wicks. Financial support: This work was funded by Queensland Health and the National Health and Medical Research Council of Australia. Authors addresses: Cheryl A. Johansen, Department of Microbiology, The University of Western Australia, Queen Elizabeth II Medical Centre, Nedlands 6009, Western Australia. Roy A. Hall and John S. Mackenzie, Department of Microbiology and Parasitology, School of Molecular and Microbial Sciences, The University of Queensland, St. Lucia, Queensland 4072, Australia. Andrew F. van den Hurk, Queensland Health Scientific Services, Queensland Health, Coopers Plains, Queensland 4108, Australia. Scott A. Ritchie, Tropical Public Health Unit, Queensland Health, P.O. Box 1103, Cairns, Queensland 4870, Australia. Reprint requests: Cheryl A. Johansen, Department of Microbiology, The University of Western Australia, Queen Elizabeth II Medical

DETECTION OF JE VIRUS RNA IN DEAD MOSQUITOES

661

20.

21. 22.

23. 24.

25. 26.

tigations of an outbreak of Japanese encephalitis outbreak in the Torres Strait, Australia, in 1998. J Med Entomol 38: 581 588. Williams DT, Wang LF, Daniels PW, Mackenzie JS, 2000. Molecular characterization of the first Australian isolate of Japanese encephalitis virus, the FU strain. J Gen Virol 81: 2471 2480. Don RH, Cox PT, Wainwright BJ, Baker K, Mattick JS, 1991. Touchdown PCR to circumvent spurious priming during gene amplification. Nucleic Acids Res 19: 4008. Armstrong P, Borovsky D, Shope RE, Morris CD, Mitchell CJ, Karabatsos N, Komar N, Spielman A, 1995. Sensitive and specific colorimetric dot assay to detect eastern equine encephalomyelitis viral RNA in mosquitoes (Diptera: Culicidae) after polymerase chain reaction amplification. J Med Entomol 32: 4252. Sudia WD, Chamberlain RW, 1962. Battery-operated light trap, an improved model. Mosq News 22: 126129. Tsai TF, Happ CM, Bolin RA, Montoya M, Campos E, Francy DB, Hawkes RA, Roehrig JT, 1988. Stability of St. Louis encephalitis viral antigen detected by enzyme immunoassay in infected mosquitoes. J Clin Microbiol 26: 26202625. Vodkin MH, Streit T, Mitchell CJ, McLaughlin GL, Novak RJ, 1994. PCR-based detection of arboviral RNA from mosquitoes homogenised in detergent. Biotechniques 17: 114116. Wasieloski LP, Rayms KA, Curtis LA, Blair CD, Beaty BJ, 1994.

27.

28.

29.

30. 31. 32.

Reverse transcription-PCR detection of La Crosse virus in mosquitoes and comparison with enzyme immunoassay and virus isolation. J Clin Microbiol 32: 20762080. Nawrocki SJ, Randle YH, Vodkin MH, Siegel JP, Novak RJ, 1996. Evaluation of a reverse transcriptase-polymerase chain reaction assay for detecting St. Louis encephalitis virus using field-collected mosquitoes (Diptera: Culicidae). J Med Entomol 33: 123127. Chow VTK, Chan YC, Yong R, Lee KM, Lim LK, Chung YK, Lam-Phua SG, Tan BT, 1998. Monitoring of dengue viruses in field-caught Aedes aegypti and Aedes albopictus mosquitoes by a type-specific polymerase chain reaction and cycle sequencing. Am J Trop Med Hyg 58: 578586. 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: 26342639. Paranjpe S, Banerjee K, 1998. Detection of Japanese encephalitis virus by reverse transcription/polymerase chain reaction. Acta Virol 42: 511. Borucki MK, Kempf BJ, Blair CD, Beaty BJ, 2001. The effect of mosquito passage on the La Crosse virus genotype. J Gen Virol 82: 29192926. Kerr PJ, Weir RC, Dalgarno L, 1993. Ross River virus variants selected during passage in chick embryo fibroblasts: serological, genetic and biological changes. Virology 193: 446449.