Cocirculation of two dengue virus serotypes in individual and pooled samples of Aedes aegypti and Aedes albopictus larvae Cocirculação de dois sorotipos do vírus dengue em amostras individuais e pools de larvas de Aedes aegypti e Aedes albopictus

Introduction: To detect dengue virus, eggs of Aedes sp were collected in the city of Belo Horizonte, Brazil, in 2007. Methods: Egg samples were subsequently hatched and the larvae were tested for the presence of dengue virus RNA by RT-PCR. Results: Among the Aedes aegypti larvae samples, 163 (37.4%) out of 435 were positive, including 32 (10.9%) of 293 individual larvae samples concomitantly positive for two serotypes. Conclusions: Virological surveillance detecting coinfected vectors in the field could represent an important strategy for understanding the numerous factors involved in the transmission and clinical presentation of dengue.

Dengue is an acute viral disease transmitted primarily by Aedes aegypti and the infectious agent is one of four serotypes of the dengue virus (DENV): DENV-1, DENV-2, DENV-3 or DENV-4.The incidence of dengue fever and dengue hemorrhagic fever (DHF) has increased in tropical regions of the Americas 1 .In Brazil, over the past 22 years, the geographic expansion of the vector has enhanced the cocirculation of serotypes DENV-1, DENV-2 and DENV-3 in 25 of the 27 Brazilian states 2 .
In the fall of 1996, DENV-1 was introduced to the City of Belo Horizonte, the capital of the State of Minas Gerais, resulting in a localized outbreak.In 1997, other outbreak occurred in the city during the summer and fall.
After a silent period during the winter and spring of 1997, a new and large epidemic cycle began in Belo Horizonte and continued throughout the summer and fall of 1998, during which the mean incidence rate was 381 per 10,000 inhabitants [3][4][5] .At that time, DENV-2 was isolated cocirculating with DENV-1.
A seroepidemiological survey was conducted by our group using a serum-neutralization assay.A seroprevalence of 23% was determined, with one in four of these cases demonstrating concurrent seropositivity for DENV-1 and DENV-2 5 .Other studies during the same period, using reverse transcription-polymerase chain reaction (RT-PCR) in pooled larvae collected in the nine administrative districts of the city, also detected samples positive for DENV-1 and DENV-2 6 .
In 2002, DENV-3 was isolated from human serum for the first time in the city.Since then, cocirculation of DENV-1, DENV-2 and DENV-3 has occurred in the State of Minas Gerais and this cocirculation has been documented molecularly and by isolation, demonstrating the genotypic variation of DENV in the state 7 .
During epidemiological weeks 13 and 14 of 2007 (March 25 th to April 7 th ), eggs were collected from positive ovitraps (of 486 placed) in three administrative districts of Belo Horizonte (Central-South, East, and Venda Nova).The larvae of A. aegypti and A. albopictus that hatched from these eggs were reared to 4 th instar, identified at a species level and examined for DENV.
In order to determine whether more than one serotype could be present in a single larva, the hatched larvae were distributed into individual and multiple larval pools.The A. aegypti pools had an average of eight larvae (range 2 to 10) and the two Aedes albopictus pools had 3 and 6 larvae each.The samples were then sent to the Molecular Virology and Bioproducts Laboratory of the Ezequiel Dias Foundation for RT-PCR analysis.The silica method was used to extract RNA from the larva samples 8 .Approximately 1µg to 5µg of extracted RNA was added to 50pmol of oligonucleotide antisense primers and 200U of reverse transcriptase (SuperScript II RT -Gibco/Life Technologies) to generate the cDNA.The oligonucleotide primers used for the detection of DENV have been previously described 9 .As a positive control for the assays, cDNA prepared from RNA extracted from the supernatant taken from the Pessanha JEM et al -Cocirculation of dengue virus serotypes in samples of Aedes sp cellular culture of C6/36 cells infected with standard specimens of DENV-1, DENV-2 and DENV-3 was used.For negative controls, sterile water was used.The amplification cycles were: one cycle at 95ºC (3min), 30 cycles at 95ºC (/15s each), one cycle at 55ºC (15s) and finally one cycle at 72ºC (30s).
A total of 13,436 Aedes sp eggs were detected in 247 (50.8%) of the 486 ovitraps, with an egg density Index (EDI) 10 of 54.4.Of the 3,604 (26.8%) eggs that hatched, 3,587 (99.5%) were A. aegypti and 17 (0.5%) were A. albopictus.Due to logistical constraints, a sample of 1,400 (39%) of the A. aegypti larvae were tested with RT-PCR.These larvae were selected using a stratified sampling process to assure proportional representation from all three administrative districts (Table 1).All 17 A. albopictus larvae were tested.
Infection rates (IR) for the various dengue viruses were determined using an Excel add-in 11 .The infection rates of the different dengue viruses, expressed as maximum likelihood estimates (MLE) are presented in Table 3.The overall IR for dengue virus in the : negative samples, DENV: dengue virus.
---293  A. aegypti larvae was 138.6/1,000 or 13.9% and DENV-2 was the predominant serotype with an overall IR of 11.1%.The IR of DENV-2 was 7.3% when was detected alone in pools, 2.1% when detected in combination with DENV-1 and 1.1% when detected in combination with DENV-3.The second most frequently detected was DENV-1 with an overall IR of 3.3%.The IR of DENV-1 was 1.2% when was detected alone in pools and 2.1% when detected in combination with DENV-2.The least frequently detected was DENV-3 with an overall IR of 1.8%.The IR of DENV-3 was 0.8% when detected alone in pools and 1% when detected in combination with DENV-2.No DENV1 and DENV-3 coinfections were detected.
The authors declare that there are no conflicts of interest.

REFERENCES FINANCIAL SUPPORT ACKNOWLEDGMENTS
In 1996 and 1997, only DENV-1 was detected in human sera submitted to the Belo Horizonte City Health Department Epidemiological Surveillance Service.Beginning in 1998, DENV-2 was also isolated 3 .Since 2002, DENV-3 has been predominantly isolated.It is interesting to note that in the present study, DENV-2 was the most frequently detected serotype.
Sabin demonstrated the occurrence of transovarial transmission of DENV-2 by A. aegypti in nature by the isolation of dengue virus from naturally infected mosquito larvae showing a minimum infection rate of 1:2,067 for DENV-2 12 .Natural transovarial transmission of DENV-4 by A. aegypti was demonstrated from adult mosquitoes reared in the laboratory from eggs collected in Trinidad, with a minimum infection rate of 1:1,855 13 .The results of laboratory experiments described above with A. aegypti showed a very inefficient minimum infection rate.Cecilio et al determined an infection rate for DENV-2 of 1:35.45,indicating that transovarial transmission of DENV by A. aegypti could provide a mechanism for the maintenance of DENV when continuous mosquito breeding is interrupted 6 .Transovarial transmission was confirmed by the detection of DENV-2 in male mosquitoes hatched from eggs which were collected in nature and reared in controlled conditions.
The detection of coinfection by two serotypes in a single larva of the vector means that the new generations of mosquitoes could be coinfected by transovarial transmission.To our knowledge, this is the first report in the Brazilian literature of double infection by different serotypes of the dengue virus detected in individual larvae.
Human cases of concurrent infections by two serotypes have been reported in Brazil: in the City of Cuiabá, State of Mato Grosso (DENV-1 and DENV-2) 14 and more recently in the City of Fortaleza, State of Ceará (DENV-2 and DENV-3) 15 .
The true epidemiological significance of the transmission of two or more serotypes by vectors that acquired the virus by transovarial transmission has yet to be established.The present results raise several questions related to the infectivity of these coinfected vectors, their significance in transmission during silent periods (interepidemic) and the relation between simultaneous infection by two serotypes and the clinical severity of cases.Virological and entomological surveillance detecting coinfected vectors in the field could represent an important strategy for increasing current understanding of the various factors involved in the transmission and clinical presentation of dengue and improving the monitoring of the dynamic of its occurrence.
The authors would like to thank the Zoonosis Control team of the City of Belo Horizonte Health Department.This study was partially funded by the Pan-American Health Organization and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq).The authors EGK, FAP and WTC are recipients of research fellowships from the CNPq.