|
|
||||||||
| ABSTRACT |
|
|
|---|
| INTRODUCTION |
|
|
|---|
The emergence of DHF has been linked to: 1) economic and ecologic changes that affected Asia during and after the World War II, and 2) the consequent dramatic expansion of the urbanized vector of dengue viruses, Ae. aegypti. In the second half of the 20th century, industrialization and rapid human population growth led to uncontrolled urbanization in Asia. In the absence of proper water supply, residents had to store water for domestic use, creating the ideal ecologic niche for Ae. aegypti. Except for some West African strains, this mosquito prefers to feed on humans and to lay its eggs in artificial breeding sites. Finally, the growth of air travel also contributed to the expansion of DHF, providing the means for viremic people to move very rapidly from one place to another.2 This was well documented on the island of Tahiti in French Polynesia, where the interepidemic periods shortened and DHF appeared after the international airport opened.4
Should Ae. aegypti be held responsible for the emergence of DHF? Endemic vectors were present in Asia and French Polynesia before the expansion of Ae. aegypti. Several studies have compared the oral receptivity for dengue viruses of Ae. aegypti with that of endemic vectors such as Ae. albopictus in Asia and Ae. polynesiensis in French Polynesia. Those studies led to the same conclusion: Ae. aegypti was not a good host for dengue viruses, i.e., it needed more virus to become infected than did Ae. albopictus and Ae. polynesiensis. It has been suggested that Ae. aegypti may have selected strains of virus producing a higher viremia in humans, assuming that a high viremia might be associated with severe clinical manifestations and hemorrhagic symptoms.5
In previous studies, we tested the oral receptivity of Ae. aegypti mosquitoes from different parts of the world where dengue is endemic (South Vietnam),6 or epidemic (French Polynesia, French Guiana).7,8 We found a very high oral receptivity with all samples. When we infected Ae. albopictus with the same virus strain, the oral receptivity was always lower (La Réunion, Albania).9,10 To study the relative importance of the two mosquito species in dengue transmission, we tested, whenever possible, Ae. aegypti and Ae. albopictus from the same geographical area and, perhaps more important, from the same breeding site. This was done for mosquitoes from both South and North Vietnam and from Thailand where the two species are sympatric. Moreover, to complete the study, we compared the oral receptivity of the two species derived from old laboratory strains (generations higher than 40) and from strains recently collected from the field (generations up to five) for different strains of dengue type 2 collected in Bangkok (Thailand, 1974), (Phnom Penh, Cambodia 2001), or the Seychelles archipelago (Indian Ocean, 1977). In all instances, we found a lower oral receptivity for Ae. albopictus recently collected from the field than for the strain reared in the laboratory for many years.
| MATERIALS AND METHODS |
|
|
|---|
|
|
Virus. The dengue type 2 Bangkok virus strain, provided by L. Rosen, was isolated in 1974 from human sera from Bangkok, Thailand. This virus had been passed only in different mosquito species by intrathoracic inoculation: two passages in Ae. albopictus, two passages in Toxorhynchites amboinonsis . We used the fourth mosquito passage to inoculate females of Ae. aegypti, Paea strain (pool D2BanP5-AA), and females of Ae. albopictus, Oahu strain (pool D2BanP5-AL).
The dengue type 2 Seychelles virus strain, also provided by L. Rosen, was isolated from human sera from the Seychelles archipelago in 1977, in an epidemic propagated by Ae. albopictus.11 This virus had been passed three times by intrathoracic inoculation in Tx. amboinensis. We used the third mosquito passage to inoculate Ae. aegypti, Paea strain, females (pool D2SeyP3-AA) and Ae. albopictus, Oahu strain, females (pool D2SeyP3-AL).
The Cambodia strain was isolated from human sera collected in 2001 in Phnom Penh, Cambodia. This serum was inoculated in Ae. aegypti, Paea strain, females (pool D2CamP1-AA) and in Ae. albopictus, Oahu strain, females (pool D2CamP1-AL).
The inoculated mosquitoes were incubated 14 days at 28°C, then triturated in heated (56°C for 30 min) fetal calf serum (FCS). The supernatant fluid recovered after low-speed centrifugation was used either as a source of virus in the mosquito blood-meals or, after filtration through 220-nm pores, as an inoculum for the production of virus in Ae. albopictus C6/36 clone cells.12 Viral stocks were produced by inoculating C6/36 cells with triturated infected Ae. aegypti, Paea strain. The mosquito cells were maintained at 28°C on RPMI-1640 medium supplemented by non-essential amino acids, penicillin, streptomycin, and 10% heated FCS. The percentage of infected cells was monitored during the incubation period by the indirect fluorescent antibody (IFA) assay13 When 100% of cells were infected, the supernatant fluid was collected and the pH adjusted to 7.5 with 10% sodium bicarbonate. The virus stock was divided into aliquots and stored at -80°C until used.
Viral stocks produced on C6/36 were D2Ban-C6/36 for the D2 Bangkok strain (inoculum = pool D2Ban P5 AA) and D2Sey-C6/36 for the D2 Seychelles strain (inoculum = D2Sey P3 AA).
Titrations of virus stocks produced in culture cells and ground up mosquito pools were carried out in Ae. aegypti, Paea strain, by inoculating serial dilutions of the supernatant intrathoracically.14 Mosquito infection was detected by IFA assay on head squashes. Titers were calculated by the 50% endpoint method and expressed as mosquito infectious doses (MID50) per mL. 15
Oral infection of mosquitoes.
The oral susceptibility of females was tested by a feeding protocol described elsewhere.7 Briefly, 5- to 7-day-old females were deprived of sucrose solution 24h before the infectious meal and then allowed to feed for 20 min through a chicken skin membrane covering an apparatus containing the feeding mixture maintained at 37°C. The infectious meal consisted of two-thirds washed rabbit erythrocytes, one-third virus suspension, and ATP (as a phagostimulant) at a final concentration of 5 x 10-3 M. Each meal yielded 108,2 MID50/mL for the samples in Tables 1
and 2
and 108,1 MID50/mL for the strains tested in Table 3
and Table 4
. Only fully engorged females were transferred to small cardboard containers and maintained at 28 ± 1°C for 14 days. Surviving females were killed and tested for the presence of dengue virus by an IFA assay on head squashes.13
|
|
| RESULTS |
|
|
|---|
Infection rates of Ae. albopictus and Ae. aegypti samples from allopatric breeding sites.
Infection rates of each sample and of the corresponding Paea control strain are shown in Table 2
. Rates were 97.3% and 98.7% for Ae. aegypti, and 24.6% and 28.8% for Ae. albopictus. Compared with the infection rate of the corresponding control using Fishers exact test, only one sample of Ae. aegypti exhibited a significant difference (AET2, P = 0.015) related to a lower rate of infection for the corresponding control, whereas the two Ae. albopictus samples exhibited a highly significant difference (P < 10-4). Differences in infection rates between Ae. aegypti and Ae. albopictus were highly significant (P < 10-4). Infection rates were similar when comparing the two Ae. aegypti samples (P = 0.614) and the two Ae. albopictus samples (P = 0.697).
Infection rates of Ae. albopictus and Ae. aegypti from old or recent laboratory strains for the Bangkok strain of dengue type 2 virus.
Infection rates of each strain of mosquitoes (results on the F4 generation for Ta Promh and Isohanoi strains) with three different pools of the D2Ban strain are shown in Table 3
. Females of the Ta Promh strain of Ae. albopictus exhibited lower infection rates (5.325.0%) for the three pools of virus compared with that of the old laboratory colony of Ae. albopictus, Oahu (80.5100%) or of the two Ae. aegypti strains (77.8.96.9% for the Isohanoi strain and 83.3100% for the Paea strain). Infection rates were similar for the two replicates when tested with the same virus pool (P > 0.05). Infection rates also were similar when comparing the three pools of the D2Ban strain for each species: P = 0.636 for the Ae. albopictus Ta Promh strain, P = 0.063 for the Ae. albopictus Oahu strain, P = 0.139 for the Ae. aegypti Isohanoi strain, and P = 0.122 for the Ae. aegypti Paea strain. Infection rates were not different when comparing the two Ae. aegypti strains, Isohanoi and Paea, for each virus pool (P = 0.169 for D2Ban-C6/36, P = 0.441 for D2BanP5-AA, and P = 0.928 for D2BanP5-AL) but were different between the two Ae. albopictus strains, Ta Promh and Oahu (P < 10-4). When considering infection rates of the two Ae. aegypti strains and of the Oahu strain of Ae. albopictus, homogeneity of infection rates was accepted for each virus pool: P = 0.087 for D2Ban-C6/36, P = 0.058 for D2BanP5-AA, P = 0.993 for D2Ban-AL. The homogeneity was rejected (P < 10-4) when data from the Ta Promh strain of Ae. albopictus were included.
Results obtained for F4 and F5 individuals infected with D2Cam and D2Sey are shown on Table 4
. As observed for the D2Ban virus strain, the old laboratory species (Ae. albopictus Oahu and Ae. aegypti Paea) and the recently colonized Ae. aegypti Isohanoi strain exhibited higher infection rates than the Ta Promh Ae. albopictus strain. However, when examining individuals from the Ta Promh strain of Ae. albopictus, an increase in infection rates was observed between the F4 and F5 generations. This increase was significant in all cases (P < 0.05) except for infections with D2SeyP3-AL (P = 0.414) where the rate of infection of F4 individuals from the Ta Promh strain of Ae. albopictus was the highest observed (60.7%). This tendency also was observed for the Isohanoi Ae. aegypti strain, for which the increase was significant when infected with D2CamP1-AL (P = 0.001). Infection rates of F4 individuals from the Ta Promh strain of Ae. albopictus were not different when considering the two mosquito pools, D2BanP5-AL and D2CamP1-AL (P = 0.939), but were different when results obtained with D2SeyP3-AL were added or when considering infection rates of D2SeyP3-AL with one pool or the other (P < 10-4).
| DISCUSSION |
|
|
|---|
These results contradict the generally accepted belief that Ae. aegypti is much less receptive to oral infection with dengue viruses of all four types than are most other Aedes capable of transmitting such viruses.5,18 It should be stressed, however, that this conclusion was based mainly on data obtained in one study, published in 1985, on females that originated from field collections made before 1971 for Ae. albopictus and before 1974 for Ae. aegypti.18 Therefore, the infection rates obtained with these individuals could not, in our opinion, allow conclusions to be drawn as to the real status of the competence of the two species. Furthermore, all the experiments were not performed with the same amount of virus, and a control strain was not included in each meal to allow a more accurate comparison of the data. We demonstrated in a previous study that with virus titers below 108 MID50/mL a small decrease in the titer of the meal led to a large decrease in the infection rate.7
One puzzling question is why our infection rates were consistently higher than those found in other studies.1820 We have shown previously that the quality of erythrocytes used in the meal is an important factor. When using erythrocytes collected and washed 72 h, instead of 24h, before the blood-meal, the rate of infected females for the control strain dropped dramatically from 9019.35%.7 It also should be noted that we used a feeding mixture with a much higher concentration of erythrocytes (two-thirds instead of one-third) than in the earlier studies. In one unpublished experiment, conducted with the D2Sey-C6/36 pool, the infection rate dropped from 97% (65/67) to 80% (20/25) for the Paea strain of Ae. aegypti fed on a mixture containing two-thirds or one-third proportions of erythrocytes, respectively. Otherwise, highly dengue-susceptible Ae. aegypti populations are commonly found in Southeast Asian, South American, and South Pacific regions, leading us to conclude that such characteristics are related to the domestic form of Ae. aegypti. Conversely, the native form of Ae. aegypti (i.e., the formosus form) shows lower infection rates.21
To determine if our results were due to a particularity of the Bangkok viral strain, e.g., adaptation or selection in the course of the different passages in the laboratory, or to the genotype of this viral strain,22 we tested two other dengue type 2 strains. We chose one strain from the Seychelles archipelago (isolated in 1977 during an epidemic transmitted by Ae. albopictus) and one collected in 2001 in Phnom Penh, where Ae. aegypti is responsible for dengue transmission. We used supernatant fluid from cell cultures and also triturated infected mosquitoes as a source of virus for the mosquito blood-meals, as was done in previous studies.1820 In all instances, we observed a significantly lower oral receptivity of the Ta Promh Ae. albopictus strain collected in 2001 compared with all other mosquito strains tested. We also observed a significant increase in the rate of infected females in the Ta Promh strain from one generation to the next. This demonstrates the importance of using individuals recently collected from the field to better estimate the "natural" oral receptivity to dengue viruses. We observed the same phenomenon in one population of Ae. aegypti formosus collected in Africa, which showed an increase in the infection rate from 52.0% at the F2 generation to 87.3% at the F20 generation (Vazeille, unpublished data).23
We have shown that the mode of production of the virus (cell culture, various number of passages on mosquitoes), and even the dengue strain itself (old or recently collected) did not alter the fact that the Ta Prohm strain of Ae. albopictus recently collected from the field was less receptive to oral infection than Ae. aegypti. However, the D2SeyP3-AL pool was able to infect the F4 females of this mosquito strain to a significantly higher degree than did the D2BanP5-AL or the D2CamP1-AL. It has been suggested that Southeast Asian genotypes of dengue 2 replicate more efficiently than American genotypes in field-derived mosquito strains.22 In our study, the Ae. albopictus strain seemed to be more receptive to infection with a virus collected during an epidemic transmitted by the same species. This result should, of course, be confirmed to better appreciate the impact of the genetic variability of virus strains, i.e., the variability in their ability to infect and replicate in a specific vector, in dengue epidemiology.
In view of these results, we conclude that Ae. albopictus populations now present in Southeast Asia are less susceptible to infection with dengue type 2 virus than are Ae. aegypti populations. This result is supported by another study showing that both infection and transmission rates for dengue type 1 were higher for Ae. aegypti than for Ae. albopictus in Taiwan.24 We did not evaluate transmission rates but only infection rates. However, prior studies have already demonstrated that if Ae. aegypti strains differ in the level of oral receptivity (or midgut receptivity) once the virus has disseminated from the midgut, the infected mosquito will always transmit the virus.22,25
We conclude that in Southeast Asia, Ae. aegypti is a very efficient vector for dengue viruses, highly receptive to oral infection, well adapted to an urban environment (e.g., laying eggs only in artificial containers), and feeding exclusively on humans. It is the perfect vector in endemic and epidemic situations. However, it does not transmit the virus to its progeny vertically very efficiently and may not thus contribute to the maintenance of the virus during interepidemic periods. On the other hand, Ae. albopictus, which is not highly orally receptive to dengue type 2 virus, is present mainly in rural areas and does not feed exclusively on humans. However, male Ae. albopictus can transmit dengue virus sexually in the course of mating, and females can transmit it vertically more efficiently than can Ae. aegypti females.26,27 These two mechanisms could explain the maintenance of the virus in nature between epidemics in non-endemic areas where susceptible human or primate populations are not always present. Ae. albopictus also could bridge a putative sylvatic and an urban cycle of dengue since it colonizes both rural and periurban breeding sites. Even when the only vector present is Ae. albopictus, it can be responsible for dengue epidemics, as shown for the outbreaks in the Seychelles,11 Japan,28 and the recent small outbreak in Hawaii (2001). It also is also important to stress that all Ae. aegypti in the study reported here were Ae. aegypti aegypti, the light domestic form. When we compared Ae. aegypti formosus (dark sylvatic form) from Madagascar with sympatric Ae. albopictus, we found opposite results, i.e., a much higher oral receptivity for Ae. albopictus for dengue type 2 virus.29 On this island, Ae. aegypti formosus is scarce and found only in rural breeding sites, whereas Ae. albopictus is the species encountered in cities.
The two dengue vectors in Southeast Asia, Ae. albopictus and Ae. aegypti, exhibit differences in their ecology and their oral receptivity. However, in a suburban environment, they can share a breeding site. Could this larval competition affect the oral receptivity of adult females? When pooling data from the four sympatric breeding sites, we observed a homogeneity of infection rates for Ae. aegypti samples whereas infection rates among Ae. albopictus were heterogeneous. This raises the question of the role of selection in the differentiation of Ae. albopictus populations in the drastic conditions encountered at sympatric breeding sites. Of course, further investigations are needed to better analyze the consequence of larval competition on the oral receptivity of Aedes mosquitoes for dengue viruses or arboviruses in general.
Received June 12, 2002. Accepted for publication October 17, 2002.
Acknowledgments: We thank Srisucha Chawprom and Somsak Prajakwong from Vector Borne Disease Control, Chiang Mai, Thailand, Luu Le Loan and Tran Huu Hoang from the Pasteur Institute of Ho Chi Minh City, and Vu Sinh Nam and Nguyen Thi Yen from the National Institute of Hygiene and Epidemiology of Hanoi for their help in collecting mosquitoes. We also thank Catherine Thouvenot for her assistance in rearing mosquitoes.
Financial support: This work was supported by the ACIP n° 41204 of the Pasteur Institute.
Authors addresses: Marie Vazeille, Leon Rosen, Laurence Mousson, and Anna-Bella Failloux, Unité décologie des systèmes vectoriels, Institut Pasteur, 25 rue du Docteur Roux, 75 724 Paris Cedex 15, France. Telephone: 33-1-40-61-36-17, Fax: 33-1-40-61-30-89.
Reprint requests: Anna-Bella Failloux, Unité décologie des systèmes vectoriels, Institut Pasteur, 25 rue du Docteur Roux, 75 724 Paris Cedex 15, France, Telephone: 33-1-40-61-36-17, Fax: 33-1-40-61-30-89.
| REFERENCES |
|
|
|---|
This article has been cited by other articles:
![]() |
B. W Alto, L. P. Lounibos, C. N Mores, and M. H Reiskind Larval competition alters susceptibility of adult Aedes mosquitoes to dengue infection Proc R Soc B, February 22, 2008; 275(1633): 463 - 471. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. H. B. VAN BENTHEM, S. O. VANWAMBEKE, N. KHANTIKUL, C. BURGHOORN-MAAS, K. PANART, L. OSKAM, E. F. LAMBIN, and P. SOMBOON SPATIAL PATTERNS OF AND RISK FACTORS FOR SEROPOSITIVITY FOR DENGUE INFECTION Am J Trop Med Hyg, February 1, 2005; 72(2): 201 - 208. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. L. DE OLIVEIRA, M. VAZEILLE, A. M. B. DE FILIPPIS, and A.-B. FAILLOUX LARGE GENETIC DIFFERENTIATION AND LOW VARIATION IN VECTOR COMPETENCE FOR DENGUE AND YELLOW FEVER VIRUSES OF AEDES ALBOPICTUS FROM BRAZIL, THE UNITED STATES, AND THE CAYMAN ISLANDS Am J Trop Med Hyg, July 1, 2003; 69(1): 105 - 114. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |