Introduction

The Asian tiger mosquito, Aedes albopictus (Skuse), is native to Asia and the South Pacific and has been recently introduced into the continental United States and South America (Kambhampati and Rai, 1991; Kambhampati et al, 1991). Ae. albopictus is known to be an important vector of dengue in Southeast Asia (Gould et al, 1968; Chan et al, 1971) and has been implicated in a recent dengue outbreak in Mexico (Ibanez-Bernal et al, 1997). Maternally inherited Wolbachia infections were first found in the ovaries of this mosquito by Wright and Barr (1980). It is now known that nearly all populations of Ae. albopictus are superinfected with two different Wolbachia strains designated wAlbA (A group) and wAlbB (B group) (Sinkins et al, 1995b; Zhou et al, 1998; Kittayapong et al, 2002).

Cytoplasmic incompatibility (CI) is a common phenomenon in insects, caused by Wolbachia (Yen and Barr, 1971) and is known to occur in Ae. albopictus (Kambhampati et al, 1993). When a Wolbachia-infected male mates with an uninfected female, the eggs or embryos die. The net effect is a decrease in the fitness of uninfected females, which over time results in the spread of the infection through the host population (Turelli and Hoffmann, 1991). Individuals may be infected with more than one strain of Wolbachia. Among such superinfected individuals, CI occurs if the female is uninfected with respect to the strain that the male carries. The net effect is a decrease in the fitness of single-infected females, and thus the superinfection spreads (Sinkins et al, 1995b). Through the action of CI and related phenotypes, Wolbachia is estimated to have successfully invaded approximately 15–20% of all insect species (Werren et al, 2000).

The development of transgenic mosquitoes that lack the ability to transmit disease has been suggested as a possible future method for the control of insect-transmitted diseases. This requires a method to introduce and spread transmission-blocking genes into natural vector populations. The phenomenon of CI makes Wolbachia a prime candidate in this endeavor (Sinkins et al, 1997; Curtis and Sinkins, 1998). The success of this long-term goal for disease control is critically dependent on the efficiency of CI expression and maternal transmission of Wolbachia under field conditions (Turelli and Hoffmann, 1999). While these factors have been well studied in Drosophila, little attention has been paid to the dynamics of Wolbachia infections in insects of medical importance. While maternal transmission rates have been examined recently in field populations of Ae. albopictus and shown to be very high (Kittayapong et al, 2002), no studies have examined the expression of CI in field-caught Ae. albopictus. Moreover, in published Drosophila studies a clear difference has been reported between laboratory and field-reared individuals, the latter displaying reduced CI expression (Hoffmann et al, 1990, 1998; Turelli and Hoffmann, 1995). Therefore, the main objective of this study was to investigate the strength of CI expression in field populations of the mosquito vector Ae. albopictus. We have also determined whether the age of male hosts has any effect on the expression of CI in this species.

Materials and methods

Field collection of mosquitoes

Aedes albopictus were collected from their natural habitats in two different localities approximately 300 km apart between August 1999 and January 2000. The collection sites were located in Plaeng Yao District, Chachoengsao Province, eastern Thailand and Sai Yok District, Kanchanaburi Province, western Thailand. Two different methods were used to collect Ae. albopictus. First, eggs were collected from several ovitraps placed near bamboo stumps. An ovitrap consisted of a black plastic container lined with a paper towel and half-filled with water. The paper towels containing mosquito eggs were collected from both locations and were brought to the laboratory in Bangkok within a 1-week period. Second, swarming male mosquitoes were collected from locations adjacent to bamboo forests in Chachoengsao and Kanchanaburi Provinces using a simple mosquito net. Live males were put in screen-top containers and were fed with a 10% sugar solution before being transported to the laboratory in Bangkok on the same day.

Laboratory-reared mosquito strains

There were three colonies of Ae. albopictus mosquitoes used in the crossing experiments. The superinfected colony (KLPP) was generated in 1997 from crosses between males and females collected from Kanchanaburi Province, western Thailand and the Phi Phi Islands off the west coast of Thailand. The superinfected status of this colony for wAlbA and wAlbB Wolbachia was confirmed by both RFLP analysis of ftsZ PCR products and group-specific wsp primers as described in Kittayapong et al (2002). The single-infected wAlbA colony (KOH) was established before 1970 from mosquitoes collected at Samui Island (Koh Samui), off the east coast of Thailand and provided by Dr KS Rai, University of Notre Dame, USA. The uninfected colony (UJU) was artificially generated by tetracycline treatment and was provided by Dr Yasushi Otsuka, Oita Medical University, Japan. All these colonies were maintained in the insectary at 27 ± 2°C and with 70 ± 10% relative humidity.

Crossing experiments

Eggs collected from individual ovitraps were placed in separate rearing trays filled with 1 litre of distilled water and larvae were fed fish food. To avoid evaluating individuals from the same cohort, only one male per ovitrap was used in this study. Individual adult males were isolated and allowed to mate for 3 days in a confined cup with either three, 1 to 3-day-old, uninfected females from the UJU laboratory colony, or three, 1 to 3-day-old, wAlbA infected females from the KOH colony. After mating, each female was blood-fed with a hamster and allowed to oviposit F1 eggs. The same individual males were then mated again with three, 1 to 3-day-old, superinfected females from the laboratory reared KLPP colony to serve as a control cross. After 3 days, superinfected females were isolated for blood-feeding and egg-laying. Eggs laid by individual females were counted and then allowed to hatch for 24 h in deoxygenated water. CI expression of the superinfected phenotype was determined by the average number of eggs hatched from each crossing experiment. Single pair crosses between field-collected adult males and either UJU or KOH females were also done as described above.

Several crosses were done between laboratory-reared uninfected, wAlbA infected and wAlbA + wAlbB superinfected Ae. albopictus to investigate the effect of host age on CI expression, since it was not possible to determine the exact age of field-collected mosquitoes. In our experiments, the age of males was varied in three different incompatible crosses. These crosses were uninfected females × superinfected males, wAlbA infected females × superinfected males, and uninfected females × wAlbA infected males. Normal control crosses were performed with 2 to 5-day-old individuals. In these experiments, the age of males was varied from 2–5 days old to 10, 20, 30, 40, 50 and 60 days old. Individual males were first aged then mated to young individual females (1–3 days old) using an artificial force-mating technique to ensure that individual pairs were successfully mated. After mating, each female was allowed to oviposit in a confined cup. Eggs laid on each paper towel were left in the wet container to mature, then dried and hatched in distilled water. Spermatheca of individual females were dissected and the presence of sperm was identified by microscopy to confirm insemination. Eggs obtained from mating pairs with no sperm insemination were discarded. CI expression was determined as the average egg hatch rate of each cross. Statistical comparisons of egg hatch rate were performed between methods of collection and between locations using the Mann-Whitney U test from SPSS for windows version 7.5. P values at or below alpha 0.05 were considered significant.

PCR detection of Wolbachia

After crossing experiments, individual field-collected males were checked for their infection status of Wolbachia (Kittayapong et al, 2002). In this study, we used A and B group specific wsp primers, which were designed from the conserved regions of the outer surface protein gene, to determine the type of infection of individual field-collected mosquitoes (Braig et al, 1998; Zhou et al, 1998). The universal eukaryotic 12S rDNA primers (12 SAI and 12 SBI) were also used for checking the quality of DNA extraction (O'Neill et al, 1992). DNA extraction of individual gonad tissues from a colony of laboratory-reared superinfected Ae. albopictus was used as a positive control in the PCR reaction.

Reproductive tissues from individual mosquitoes were used for DNA extraction. The extraction procedure followed the crude boiling method of O'Neill et al (1992). Gonad tissues from individual mosquitoes were dissected under distilled water on a sterile microscope slide and were then homogenised in 100 μl of STE buffer. Homogenates were heated at 95°C for 10 min and were centrifuged at 14 000 rpm for 1 min. Supernatant of 1 μl from individual mosquitoes was used in each 20-μl reaction volume. Each PCR reaction mixture contained 2 μl 10× buffer (Promega), 2 μl 25 mM MgCl2, 0.5 μl dNTPs (10 mM each), 0.5 μl of each primer and 1 unit of Taq DNA polymerase (Promega). The temperature profile for PCR was as follows: 95°C for 3 min, followed by 30 cycles of 95°C for 1 min, 50°C for 1 min and 72°C for 1 min per cycle. Ten μl of each PCR product was electrophoresed on a 1% agarose gel and was visualised by ethidium bromide staining. The size of the PCR product was determined using a 1 Kb DNA ladder (Gibco). Expected sizes of the PCR products with A and B group-specific wsp primers were 397 bp and 501 bp respectively.

Results

Expression of CI in field-collected males

For each cross, a total of 35–40 field-collected adult males from each location were sequentially crossed with either 1 to 3-day-old, laboratory-reared uninfected (UJU) or wAlbA infected (KOH) females and then superinfected (KLPP) females. The cross to UJU females was able to measure the combined CI effect of the superinfection. The cross to KOH enabled CI due to the presence of the wAlbB infection in field males to be evaluated independently of the wAlbA infection. The final cross was a control to verify that the males were fertile, only data from fertile males was subsequently used. Table 1 shows the number of eggs laid by individual females and the percentage of egg hatch from all crosses. CI expression, as determined by egg hatch rate, was very strong (0.19–0.40% hatching) in all of the crosses between field- collected males and young UJU or KOH females.

Table 1 Results of crossing experiments between field-collected adult males and 1–3 days old, uninfected (UJU), single wAlbA infected (KOH) or superinfected (KLPP) laboratory-reared females
Full size table

In addition to field-caught adults, eggs were collected from field ovitraps and reared in the insectary. The resulting adult males were crossed as above to either UJU or KOH females, which in all cases resulted in low egg hatch rates (0.02–0.74% hatching) as shown in Table 2. Males derived from ovitraps or caught as adults from swarms show no significant differences in the rate of egg hatch when crossed with either UJU (P > 0.05) or KOH (P > 0.05) females. In addition, males collected from different locations, eastern or western Thailand, did not show significant differences in the strength of CI expression when crossed with either UJU (P > 0.05) or KOH (P > 0.05) females. In control crosses between field males and superinfected females, the mean percent egg hatch ranged between 87.91% to 91.01%.

Table 2 Results of crossing experiments between laboratory-raised adult males derived from field ovitrap eggs and 1–3 days old, uninfected (UJU), single wAlbA infected (KOH) or superinfected (KLPP) laboratory-reared females
Full size table

Host age effect on CI expression

Results of egg hatch rates from the crosses that varied male age are shown in Table 3. In these crosses, strong CI was expressed at all ages for superinfected KLPP males regardless of whether their young female partners were uninfected or wAlbA infected. There were no viable eggs produced at all male age classes. In the crosses between single infected KOH males and uninfected females, strong CI expression was observed until the males were 10 days old. The wAlbA infected males that were 20 to 60 days old produced some viable offspring ranging from 18.27% to 54.44% egg hatch.

Table 3 Results of egg hatch between incompatible crossing types that should show strong CI expression only if males were young. In this experiment, the age of individual males was varied from 2–5 days old to 10, 20, 30, 40, 50, and 60 days old
Full size table

Discussion

A number of studies have documented a reduction in the strength of Wolbachia-mediated CI expression when old males are used in crosses. This was first characterised for Culex (= fatigans) quinquefasciatus (Singh et al, 1976) and has since been reported for Drosophila simulans (Hoffmann et al, 1986) and Armigeres subalbatus (Jamnongluk et al, 1999). The underlying mechanistic hypothesis is that Wolbachia densities decrease with age of males (Binnington and Hoffmann, 1989; Bressac and Rousset, 1993). However, in the studies reported here, no evidence could be found for this effect in aged superinfected males. When KOH wAlbA males were aged and then mated, a pronounced weakening of CI was detected which mirrored studies in other species. The KOH strain has been shown to be infected at much lower densities than superinfected strains (Sinkins et al, 1995a) and as such might be more susceptible to density related aging effects. Considering that nearly all individuals sampled from field populations to date have been shown to be superinfected (Kittayapong et al, 2002), it would indicate that male age effects should not be influencing CI dynamics in the field.

Adult males caught from the field, transported back to the lab and mated to colony females all produced strong CI, indicating that field rearing does not appear to influence the strength of CI expression in this mosquito species. This is in contrast to studies with Drosophila that have indicated that CI expression can be much stronger in the laboratory than the field (Hoffmann et al, 1998). In our study, no difference was seen between laboratory or field-reared individuals. Turelli and Hoffmann (1999) have suggested that in order for Wolbachia to be used effectively in an applied setting to invade target insect populations, maternal transmission rates and CI expression should both be high. The results presented in this and previous studies (Kittayapong et al, 2002) indicate that Wolbachia infections of Ae. albopictus, a known dengue vector, have CI properties in the field that suggest Wolbachia may be a better candidate for applied strategies to modify insect vector competence than parameter estimates from Drosophila would initially indicate.