Journal of the American Mosquito Contol Association, 18(l):26-31,20O2Copyright @ 2OO2 by the American Mosquito Control Association, Inc.

SIMULTANEOUS DETECTION OF THREE MOSQUITO-BORNEENCEPHALITIS VIRUSES (EASTERN EQUINE, LA CROSSE, AND

ST, LOUIS) WITH A SINGLE-TUBE MULIIPLEX REVERSETRANSCRIPTASE POLYMERASE CHAIN REACTION ASSAY

JOON-HAK LEE,I KEN TENNESSEN., BRUCE G. LILLEYI ,cNo THOMAS R. UNNASCH'3

ABSTRACT, Three mosquito-bome human encephalitis viruses (eastem equine encephalitis virus [EEE], St. l-ouisencephalitis virus [SLE], and La Crosse encephalitis virus [LAC]) are sympatric in the southeastem United States.However, little is known conceming the temporal and spatial pattem of the distribution of these viruses in this area.As part of surveillance activities to detect the transmission of these 3 viruses in the Tennessee Valley area, wedeveloped a single-tube multiplex reverce transcriptase polymerase chain reaction (RT-PCR) assay capable of detectingthese 3 mosquito-bome viruses in a single reaction. Three viruses were differentiated by size of amplified products.Sensitivities of the multiplex RT-PCR assay for SLE, EEE, and LAC were 1-3 log median tissue culture infectivedoses per pool, roughly comparable to the reported sensitivity of PCR detection assays for the individual viruses, andI log more sensitive than antigen-capture assays for SLE and EEE. The sensitivity of the multiplex PCR was notchanged significantly when carried out in the presence of extracts prepared from 50 uninf'ected mosquitoes. The costof the assay is estimated at $2.98 per test, similar to the cost of other RT-PCR-based assays for viruses. However,adaptation of the RT-PCR to a multiplex format adds less than $0.01 to the per-unit cost of an RT-PCR assay targetinga single virus species. Analysis of these data suggests that the single-tube multiplex RT-PCR assay represents asensitive, specific, cost-effective, and rapid method for monitoring activities of the 3 endemic mosquito-bome humanencephalitis viruses in mosquito populations in the southeastem United States.

KEY WORDS Reverse transcriptase polymerase chain reaction, St. Louis encephalitis, eastern equine en-cephalitis, La Crosse encephalitis, arbovirus

INTRODUCTION

Arboviral encephalitis viruses represent an im-portant health threat to humans. In the eastern Unit-ed States, 3 major encephalitis viruses are knownto be transmitted. These include eastern equine en-cephalitis virus (EEE), St. Louis encephalitis virus(SLE), and La Crosse encephalitis virus (LAC). All3 of these viruses have been documented in thesoutheastern United States (Morris 1988, Tsai andMitchell 1989, Mancao et al. 1996). Although doc-umented cases of human infections with these vi-ruses are relatively rare, the severe morbidity andmortality associated with these viruses makes thempotentially serious health threats. As a result of thisthreat, several states carry out surveillance pro-grams whose goal is to provide early detection ofoutbreaks of these viruses.

One of the most efficient ways to detect trans-mission of the encephalitis viruses is to documentthe presence of the viruses in the vector mosquitopopulation. However, present methods for the de-tection of these viruses in mosquitoes are time-con-suming, which delays getting results to surveillancepersonnel and increases costs. Several differentmethods have been developed to detect the pres-ence of these viruses in vector mosquitoes. Theseinclude direct viral isolation by tissue culture (Nas-

I Division of Geographic Medicine, University of Ala-bama at Birmingham, 1530 3rd Avenue South, Birming-ham, AL 35294-2170.

2 Tennessee Valley Authority, Environmental ResearchCenter 2P, PO Box lol0, Muscle Shoals, AL 35662-1010.

3 To whom correspondence should be addressed.

ci and Mitchell 1996), antigen-capture assays withmonoclonal antibodies specific for each of the vi-ruses (Hildreth et al. 1984. Tsai et al. 1987. Olsonet al. 1991), and amplification of viral-specificDNA products from viral RNA by reverse tran-scriptase polymerase chain reaction (RT-PCR;Armstrong et al. 1995, Nawrocki et al. 1996). Eachof these methods has its advantages and disadvan-tages. The tissue culture method, although the mostsensitive method of viral identification, is costlyand requires a well-equipped laboratory with BSL3biohazard certification. The tissue culture methodrequires several days to obtain a result. The enzymeimmunoassay (EIA) and RT-PCR-based assays areless expensive, more rapid than tissue culture (Na-wrocki et al. 1996) and are highly specific for viralRNA. In some cases, particular mosquito specieshave been shown to play an important role in main-taining or transmitting more than 1 of the enceph-alomyelitis viruses (Morris 1988, Reisen and Mon-ath 1989, Tsai and Mitchell 1989). In the case ofthese species, it may be necessary to assay mos-quitoes for each of the viruses that may be trans-mitted. Thus, in areas such as the southeasternUnited States, where more than 1 virus is found, itcan be necessa.ry to run multiple tests on a singlesample to ensure the detection of all viruses.

As part of a surveillance effort to detect EEE,SLE, and LAC transmission in the Tennessee Val-ley, we have developed a multiplex RT-PCR ca-pable of detecting viral RNA from the 3 viruses ina single reaction. This method employs a single-tube protocol, minimizing the manipulations need-ed to carry out the RT-PCR assay. The adaptation

26

Mencs 2002 ENcEpHALrrrs Vrnus MurrplEx RT-PCR

Table 1. Oligonucleotide primer sequences and multiplex reverse transcriptase polymerase chain reaction products.

Primer sequences

Expected size ofamplicons

(base pairs)

2'7

Virusesr

EEE

SLE

LAC

5' TACCCTACACTTAACTACCCGC 3'5' TGTCGTTTGCCTGGTTTAGGT 3'5' CGATTGGATGGATGCTAGGTAG 3'5, ACTCGGTAGCCTCCATCTTC 3'5' TCAAGAGTGTGATGTCGGATTTGG 3'5' GGAAGCCTGATGCCAAATTTCTG 3'

298

240

7 1 5

IEEE, eastem equine encephalitis; SLE, St. Louis encephalitis; LAC, La Crosse encephalitis.

of RT-PCR assays for each of the viruses to a sin-gle-tube format also means that any of the 3 virusesendemic to the southeastern United States may bedetected in a single reaction, minimizing the num-ber of tests needed to fully characterize each sam-ple.

MATERIALS AND METHODS

Virus stocks: Eastern equine encephalomyelitisvirus (NJ/60 strain, passage 6), SLE (TBH-28, un-known passage number), and LAC (Prototype)were kindly provided by the Centers for DiseaseControl and Prevention (Fort Collins, CO). Theseviruses were amplified in Vero cell culture andstored at -80'C. Titers of amplified viruses weredetermined by microtitration assay (Hierholzer andKillington 1996) and calculated as previously de-scribed (Reed and Munch 1938). The titers of stockviruses were 7.37 (LAC), 8.73 (EEE), and 8.03(SLE) log median tissue culture infective doses(TCIDro)/ml.

Single-tube multiplex RT-PCR: The RNA ex-traction was performed with Trizol LS (Life Tech-nologies, Grand Island, NY) according to manufac-turer's instructions. Multiplex RT-PCR wasperformed with a single-tube RT-PCR kit (Pro-STAR HF Single-Tube RT-PCR System, Strata-gene, Lalolla, CA) according to the manufacturer'sinstructions. Reactions were carried out in a totalvolume of 40 pl, and contained 3 primer pairs (leach for EEE, SLE, and LAC; Table l). Oligonu-cleotide primers for SLE were selected from themembrane M protein and envelope protein genes.The EEE primer pairs were modified slightly frompreviously published primers for this virus (Arm-strong et al. 1995). Similarly, the LAC primerswere adapted from previously described sequences(Wasieloski et al. 1994). The primer pairs were de-signed so that the predicted melting points and op-timal annealing temperatures for PCR were within3.3'C of one anothe!: as determined by the algo-rithms contained in the Oligo@ program package(Rychlik 1992). Two hundred picomoles of eachprimer set were used in the reaction. The thermalprofile consisted of 42"C for 15 min, 95"C for 1min, and 40 cycles of 95'C for 30 sec, 58"C for 30sec, and 68'C for 2 min, followed by 68'C for lO

min. The reaction products were separated on a2.OEa agarose gel (SeaKem LE, FMC Bioproducts,Rockland, ME) and detected by staining in 2 p'g/liter ethidium bromide.

To test the speciflcity of the assay and its abilityto detect mixed infections, RNA was extractedfrom I ml of 6-6.5 log TCID.o/ml of each of the 3viral stocks. The purified RNA was resuspended inl0 pl of RNAse-free water. One microliter of thepurified RNA (corresponding to 5-5.5 TCID.')from each virus was combined in each of the 8possible combinations and used as a template in theRT-PCR.

To determine the sensitivity of the RT-PCR, se-rial dilutions of 3 stock viruses were prepared incold BA-1 medium (199 medium containing l%obovine serum albumin with antibiotics and adjustedto pH 7.6 with Tris and sodium bicarbonate; Tsaiet al. 1987), and RNA was extracted. To assess theeffect of mosquito extract on the sensitivity of theassay, the serial dilutions of the viral stocks weremixed with 5O Aedes aegypti (L.) in a final volumeof 1 ml of medium. Homogenates were preparedfrom the virus mosquito mixture by vortexing themixtures in the presence of 4 copper-plated steelbeads (Copperhead@ BBs, Crosman, East Bloom-field, NY). Homogenates were transferred to 1.5-ml centrifuge tubes and clarified by centrifugationat l3,O0O X g for 1 min. The RNA was extractedfrom the supernatants as described above, and usedas a template in the RT-PCR assay. After RT-PCR,the products were treated with 5 units of RNAse(United States Biochemical Corp., Cleveland, OH)for 30 min at 37"C to remove the mosquito RNApresent in the reaction and to assist in the visuali-zation of the PCR products.

Antigen-capture assctys: Antigen-capture assaysfor EEE and SLE were carried out following pre-viously published methods (Tsai et al. 1987, Olsonet al. l99l). Antibodies to EEE and SLE and pos-itive control antigens for both viruses were kindlyprovided by Roger Nasci of the Centers for DiseaseControl (Fort Collins, CO). Serial dilutions of theviral stocks were prepared as described above, and100 u,l of each dilution of each virus was tested ineach well by EIA. Assays were calried out in qua-druplicate. The average optical density (OD) value

Me.ncn 2002 EttcppunLtns Vtnus Mr;lnplpx RT-PCR 3 I

where little information on the viruses is availableand where multiple arboviruses are circulating inhosts and mosquito populations, such as in thesoutheastern United States.

ACKNOWLEDGMENTS

The research received financial support from theTennessee Valley Authorit;u (contract 98RE4-233022) and the National Institutes of Health (pro-ject 1 ROl 4149724). We are grateful to Nick Kar-abatos and Roger Nasci (Centers for DiseaseControl, Fort Collins, CO), Scott Weaver (Univer-sity of Texas Medical Branch [UTMB]-Galveston,Galveston, TX), Laura Chandler (UTMB-Galves-ton), and Robert Tesh (UTMB-Galveston) for pro-viding materials used in this study. We are also indebt to Jun Isoe of the University of Arizona for agift of uninfected Ae. aegypti mosquitoes.

REFER.ENCES CITED

Armstrong P, Borovsky D, Shope RE, Morris CD, Mitch-ell CJ, Karabatsos N, Komar N, Spielman A. 1995. Sen-sitive and specific colorimetric dot assay to detect east-ern equine encephalomyelitis viral RNA in mosquitoes(Diptera: Culicidae) after polymerase chain reactionamplification. J Med Entomol 32:42-52.

Hierholzer JC, Killington RA. 1996. Virus isolation andquantification. In: Mahy BWJ, Kangro HO, eds. Virol-ogy method manual New York: Academic Press. p 25-36.

Hildreth SW Beaty BJ, Mayfield HK, Gilfillan RE Ro-senau BJ. 1984. Detection of eastern equine encepha-lomyelitis virus and Highlands J virus antigens withinmosquito pools by enzyme immunoassay (EIA). II. Ret-rospective field test of the EIA. Am J Trop Med Hyg33:973-98O.

Howe DK, Vodkin MH, Novak RI, Mitchell CJ, Mclaugh-lin GE. 1992. Detection of St. Louis encephalitis in

mosquitoes by use of the polymerase chain reaction. "IAm Mosq Control Assoc 8:333-335.

Janousek TE, Kramer WL. 1998. Surveillance for arthro-pod-borne viral activity in Nebraska, 1994-1995. J MedEntomol 35:758-762.

Mancao MY. Law IM. Roberson-Trammell K. 1996. Cal-ifornia encephalitis in Alabama. South Med J 89:992-993.

Morris CD. 1988. Eastern equine encephalomyelitis. In:Monath TB ed. The arboviruses: epidemiology andecology Volume 3. Boca Raton, FL: CRC Press. p l-20.

Nasci RS, Mitchell CI. 1996. Arbovirus titer variation infield-collected mosquitoes. J Am Mosq Control Assoc12:167-17 l.

Nawrocki SJ, Randle YH, Vodkin MH, Siegel Jf; NovakRJ. 1996. Evaluation of a reverse transcriptase-poly-merase chain reaction assay for detecting St. Louis en-cephalitis virus using field-collected mosquitoes (Dip-tera: Culicidae). I Med Entomol 33:123-127.

Olson JG, Scott TW' Lorenz LH, Hubbard JL. 1991. En-zyme immunoassay for detection of antibodies againsteastern equine encephalomyelitis virus in sentinelchickens. J Clin Microbiol 29:1457-1461.

Reed LJ, Munch H. 1938. A simple method of estimatingfifty percent endpoints. Am J Hyg 27:493-497.

Reisen WK, Monath TP 1989. Western equine encepha-lomyelitis. In: Monath T\ ed. The arboviruses: epide-miology and ecology Volume 5. Boca Raton, FL: CRCPress. p 89-137.

Rychlik W. 1992. Oligo, 4.O5 Plymouth, MN: NationalBiosciences.

Tsai TE, Bolin RA, Montoya M, Francy DB, Roehrig JT1987. Detection of St. Louis encephalitis virus antigenin mosquitoes by capture enzyme immunoassay. "/ ClinM icrobio I 25 :3'l O-37 6.

Tsai TE Mitchell CJ. 1989. St. Louis encephalitis. In:Monath TP, ed. The arboviruses: epidemiology andecology Volume 4. Boca Raton, FL: CRC Press. p I 13-143.

Wasieloski LP Jr, Rayms-Keller A, Curtis LA, Blair CD,Beaty BJ. 1994. Reverse transcription-PCR detectionof LaCrosse virus in mosquitoes and comparison withenzyme immunoassay and virus isolation. J Clin Mi'crobiol 32:2O76-2080.