Combining the Sterile Insect Technique with the Incompatible Insect Technique: I-Impact of Wolbachia Infection on the Fitness of Triple- and Double-Infected Strains of Aedes albopictus

The mosquito species Aedes albopictus is a major vector of the human diseases dengue and chikungunya. Due to the lack of efficient and sustainable methods to control this mosquito species, there is an increasing interest in developing and applying the sterile insect technique (SIT) and the incompatible insect technique (IIT), separately or in combination, as population suppression approaches. Ae. albopictus is naturally double-infected with two Wolbachia strains, wAlbA and wAlbB. A new triple Wolbachia-infected strain (i.e., a strain infected with wAlbA, wAlbB, and wPip), known as HC and expressing strong cytoplasmic incompatibility (CI) in appropriate matings, was recently developed. In the present study, we compared several fitness traits of three Ae. albopictus strains (triple-infected, double-infected and uninfected), all of which were of the same genetic background (“Guangzhou City, China”) and were reared under the same conditions. Investigation of egg-hatching rate, survival of pupae and adults, sex ratio, duration of larval stages (development time from L1 to pupation), time to emergence (development time from L1 to adult emergence), wing length, female fecundity and adult longevity indicated that the presence of Wolbachia had only a minimal effect on host fitness. Based on this evidence, the HC strain is currently under consideration for mass rearing and application in a combined SIT-IIT strategy to control natural populations of Ae. albopictus in mainland China.


Introduction
Aedes albopictus, one of the most invasive mosquito species [1], is the primary vector of the chikungunya virus and secondary vector of dengue viruses. However, in some areas, like mainland China, this mosquito plays a primary role in the transmission of dengue [2], with more than 43,000 cases reported in Guangdong in 2014 (WHO Dengue situation update 453). With no vaccine available, controlling the vector population is the only way of preventing or limiting the transmission of the chikungunya and dengue viruses. For several decades, vector control methods have mainly relied on the extensive use of insecticides, and this has resulted in an increased appearance of insecticide resistance in Ae. albopictus populations [3]. Furthermore, insecticides have a negative impact on non-target insect populations and can cause major toxicological effects on human health and the environment [4][5][6]. In addition to insecticide use, source reduction (destruction of mosquito oviposition sites) is another common approach to control, but is not sustainable, especially on a large scale [7]. Consequently, additional control methods are needed for the control of Ae. albopictus.
There has recently been an increased interest towards the development and application of genetic control methods such as the sterile insect technique (SIT) and the incompatible insect technique (IIT) to control mosquitoes including Ae. albopictus. The main difference between the SIT and the IIT lies in the sterilization means: gamma or x-ray irradiation is used in the former instance, an infection with Wolbachia in the latter. The sterile insect technique (SIT) is a species-specific and environmentally-friendly method which includes the mass-rearing of the target species, sterilization and inundative releases of the male insects into the target population. The released sterile males mate with and inseminate wild females, and through sequential releases the target population is suppressed [7]. Several successful SIT programs are in progress or have been concluded, including the eradication programs against the New World screwworm, Cochliomyia hominivorax Coquel in North America [8] and the tsetse fly, Glossinaausteni [9] on the island of Unguja.
Wolbachia is a maternally-transmitted obligate intracellular Alphaproteobacterium which is widespread among insect species [10,11]. Although mutualistic nutrition as well as other beneficial associations have been reported [12][13][14][15], the presence of Wolbachia in insect hosts has been commonly associated with reproductive phenotypes such as male killing, feminization, parthenogenesis and cytoplasmic incompatibility (CI) [10,11]. In diplo-diploid species, CI usually manifests as embryonic mortality resulting from mating between a Wolbachia-infected male and a female which is either uninfected or infected with a different Wolbachia strain [16,17]. The IIT is based on the mechanism of Wolbachia-induced CI and relies on the repeated releases of incompatible males which will mate with wild type females leading to the gradual suppression of the target population [18][19][20]. The first successful application of IIT was achieved in Burma where the target population of the filariasis vector Culex pipiens was almost eliminated [21]. Feasibility studies for the use of IIT, with or without irradiation, to control populations of the mosquito species Ae. albopictus [22], Ae. polynesiensis [23], Cx. pipienspallens [24] and An. stephensi [25] have provided encouraging results both in the laboratory and in the field.
Ae. albopictus is naturally single (wAlbA or wAlbB) or double-infected (wAlbA and wAlbB) with Wolbachia [26][27][28][29]. Double-infected Ae. albopictus males have been shown to express strong CI in crosses with either uninfected or single-infected females [28]. Several artificiallyinfected lines have also been established via embryonic microinjections mainly with the aim of using them for the control of natural populations of Ae. albopictus. Four different Wolbachia strains (wMel, wMelPop, wRi and wPip) have been transferred to Ae. albopictus, establishing single-or triple-infected lineswith diverse CI patterns (see Table 1 in Bourtzis et al. 2014) [7,22,[30][31][32][33][34]. The effect of Wolbachia infection on the fitness of the naturally double-infected as well as the transinfected lines has been investigated. However, in most cases, this analysis has not been done in a thorough and systematic manner, failing to take into consideration the traits and factors important for mass rearing and large scale applications including the genetic background of the target natural population and its fitness in the field, especially mating competitiveness, flight ability and longevity [28,31,33,[35][36][37].
Recently, a triple-infected (wAlbA, wAlbB and wPip) Ae.albopictus HC strain has been established which expresses strong CI against double-infected or uninfected strains with the same genetic background (Xi et al., unpublished data). This line could be used for the control of natural populations of Ae. albopictus by IIT, SIT or using a combination of irradiation-and Wolbachia-induced sterility which, in the absence of a perfect sexing system, would allow male-only releases [7]. However, the mass rearing and quality control analysis of the strain which will be used to introduce sterility in nature are prerequisites prior to large scale implementation of any of these approaches [7]. In the present study, we performed a comparative analysis of the impact of Wolbachia infection on several fitness traits of three Ae. albopictus strains with a high similarity of genetic background but with different Wolbachia infection status (triple-infected, double-infected and uninfected). Our results are discussed in view of using the triple-infectedstrain as a tool to suppress Ae. albopictus populations in pilot sites in mainland China.

Ethics Statement
The Ae. albopictus larvae were collected by the Center of Disease Control and Prevention of Guangzhou in strict accordance with the principles of the institutional and national Committees of Animal Use and Protection. No specific permissions were required for Ae. albopictus studies in Guangzhou. The 20 locations, where the larvae were collected, were not private or protected areas and did not involve endangered or protected species (Table 1). The blood used for routine blood-feeding was collected in Vienna, Austria during routine slaughtering of pigs or cowsin a nationally authorized abattoir, conducted at the highest possible standards strictly following EU laws and regulations.

Mosquito strains and rearing conditions
Three Wolbachia-infected strains of Ae. albopictus were used in the present study: the doubleinfected strain of GUA from Guangzhou City, China (wAlbA and wAlbB), the HC strain (triple-infected with wAlbA, wAlbB and wPip) and the Wolbachia-free aposymbiotic line (GT) strain. The GUA strain was established from 2,000 mosquito larvae collected from 20 different districts of the Guangzhou metropolitan area in March 2013 (100 larvae collected per district). The strain was maintained at a high population size (several thousand mosquitoes) in the laboratory for ten generations [eight generations in Guangzhou, China and two generations in the Insect Pest Control Laboratory (IPCL), FAO/IAEA Joint Division, Seibersdorf, Austria] before being used for any experimental work. The triple Wolbachia-infected HC strain was generated by the transfer of the wPip strain (originally from Culex pipiens molestus), through embryonic cytoplasmic injections, into the Ae. albopictus Houston strain. A single triple-infected female (wAlbA, wAlbB and wPip) was used for the establishment of the HC line. Until submission of thismanuscript, the infection has been stably maintained for at least three years. The transinfected line induces almost complete unidirectional CI toward the wild type Houston strain (Xi, unpublished data). HC females were outcrossed with GUA males for 6 consecutive generations. The outcrossing involved 100 virgin triple-infected females with 100 virgin GUA males in each generation in order to establish the HC line, with similar genetic background as the GUA strain. The strain was maintained at high population size (several thousand mosquitoes at each generation) in the laboratory for an additional two generations before being used for experimental work. The GT line was developed in Guangzhou by rearing several thousand adult mosquitoes with 10% sugar solution containing tetracycline (1.0 mg/ml) for five consecutive generations. The strain was maintained at high population size (several thousand mosquitoes) in the laboratory for two generations in 10% sugar solution without tetracycline before being used for any experimental work. The successful removal of Wolbachia and the maintenance of the symbiont-free status of the line was monitored by a standard wsp gene-based PCR assay using the primers 81F and 691R [26]. Since their establishment in the IPCL (September 2013), the GUA, HC and GT strains have been maintained in a climate-controlled room at 27 ± 1°C, 80 ± 10% RH, and a photoperiod of 12:12 (L:D) h. Taken together, the impact of Wolbachia infection type on the fitness of the three Ae. albopictus strains was assessed in generation G10 for the GUA strain, G8 for the HC strain and G7 for the GT strain.
Larvae were reared in plastic trays (40×29×8 cm) at a density of 3,000 first-instar larvae (L 1 ) per tray that contained 1 liter of deionized water and were fed on IAEA2 diet regime [38] with the following minor modifications. The diet (7.5%) used for our experiments was made of 26.26g (35%) Bovine Liver Powder (MP Biomedicals, Santa Ana, CA), 37.5g (50%) Tuna Meal (T.C. Union Agrotech, Thailand)and 11.24g (15%) Brewer Yeast(Sigma Aldrich Inc., St. Louis, MO) mixed in 1 liter of deionized water. Pupae were collected and placed in small plastic cups inside a clean adult cage for emergence. Adults were kept in standard 30×30×30 cm plastic cages (BugDorm 1; MegaView, Taichung, Taiwan) and continuously supplied with a 10% sugar solution. Bovine or porcine blood was provided to female mosquitoes (4 to 5 days old) twice a week and moist oviposition papers (white crêpe paper IF C140, Industrial FiltroS.r.l., ColognoMonzese, Italy) were put into cage 48 hours after blood-feeding for collection of eggs. Forty-eight hours later, the filter paper was removed from the cage and allowed to dry for 24 hours, then the egg paper was placed in a plastic bag and stored in a sealed box in a climatecontrolled room (27 ± 1°C, 80 ± 10% RH).
Immature development of the Ae. albopictus HC, GUA and GT strains For each strain, 3 replicates of 200-300 7 day-old eggs were transferred into a hatching solution made of 700 ml deionized water containing 0.25 g Nutrient Broth and 0.05 g Brewer Yeast at 27 ± 1°C. The parental fertility was measured by counting the number of eggs hatched from the total egg number laid per female, observed under the stereomicroscope.
Time to pupation and time to emergence were both recorded as the time required for the development of L 1 to pupa and L 1 to adult stages, respectively. Survival to pupation and survival to adult emergence were calculated as the proportion of larvae that survived from L 1 to pupal stage and from L 1 to adult stage, respectively.

Wing length
To assess adult size, the left wing of 20 males and 20 females from each strain was collected for viewing under a stereomicroscope. The wing length was determined by measuring the distance from the distal edge of the alula to the end of the radius vein (excluding fringe scales). A digital image of the wing was made using a CC-12 camera mounted on a stereomicroscope and the actual measurement was performed using the analysis B software (Olympus Soft Imaging Solutions GmbH, Munster, Germany). Three replicates of 20 males and 20 females each were conducted per strain.

Adult longevity and female fecundity
To assess adult longevity, 50 newly emerged males (maintained in the absence of females) and 30 newly emerged females (maintained together with 30 males) were placed in 30×30×30 cm plastic cages with constant access to 10% sugar solution. Female longevity was determinedunder two feeding regimes: (a) 10% sugar solution and (b) 10% sugar solution plus a blood meal on the 5 th and 25 th day. Oviposition cups were provided to the gravid females 48 hours after blood-feeding. Dead adults were removed and recorded daily until all females died. Three replicates were performed per strain.
To assess female fecundity, 25 newly emerged males and 25 newly emerged females (ratio 1:1) were placed in insect cages as previously described. Ten day old females were blood-fed and transferred into individual plastic tubes with moist filter paper. Fecundity was determined by counting the total number of eggs laid per female. Three replicates were conducted for each strain.

Statistical analysis
The statistical analysis was conducted using the SPSS 13.0 software and GraphPad Prism 6.0 software. Egg hatch rate, sex ratio (expressed as proportion of females per total number of adults), survival to pupal stage (from L 1 to pupa) and survival to adult stage (from L 1 to adult) data were arcsine transformed. Normalcy of the data was assessed by the D'Agostino-Pearson omnibus normality test of the GraphPad Prism 6.0 software. Comparisons between strains were performed by one-way analysis of variance (ANOVA) and Tukey's post hoc tests (P<0.05). One-way ANOVA and Tukey's post hoc tests were also used to compare the mean time topupation and the mean time to emergence, wing length and fecundity. Independent t-test was used to assess potential differences according to sex observed on the mean time to pupation, the mean time to adult emergence and wing length within the same strain. Kaplan-Meier survival analyses (P<0.05) were conducted to determine differences in adult longevity between strains.

Results
The artificially triple-infected (wAlbA, wAlbB, wPip) HC strain produced by backcrossing and the Wolbachia-free GT strain produced by tetracycline treatment were considered to have a highly similar genetic background to the naturally double-infected (wAlbA, wAlbB) Ae. albopictus GUA strain originating from the Guangzhou metropolitan area in China.

Immature development of the Ae. albopictus HC, GUA and GT strains
Under the same rearing conditions, there was no difference in the mean hatch rate of the eggs of the Ae. albopictus HC, GUA and GT strains (F = 0.28, df = 2, P>0.05) which ranged from 82% to 85% (Table 2). Survival both to pupation and to adult emergence did not differ significantly among the HC, GUA and GT strains (F = 3.82, df = 2, P>0.05 and F = 3.82, df = 2, P>0.05, respectively) ( Table 2). In addition, no significant difference in adult sex ratio was observed between the three Ae. albopictus strains (F = 0.55, df = 2, P>0.05) ( Table 2).
As shown in Table 3, the triple-infected HC strain exhibited faster development from first instar larvae to pupation and from first instar larvae to adult emergence compared to the double-infected GUA and the uninfected GT strains for both males (F = 21.86, df = 2, P<0.05; F = 22.65, df = 2, P<0.05, respectively) and females (F = 5.66, df = 2, P<0.05; F = 7.57, df = 2, P<0.05, respectively). For all the strains, the time both to pupation and to adult emergencewas always about one day shorter for males than females (HC, t = 12.83, df = 533. 33

Female fecundity and adult longevity
No significant difference was observed in the mean number of eggs laid per female between the three strains (F = 0.39, df = 2, P>0.05) ( Table 4). With regards to male longevity, no significant difference was observed between the HC, GUA, and GT strains (χ 2 = 0.22, df = 2, P>0.05; Fig  2A). No difference was observed either in female longevity between these three strains in either feeding treatments (Fed on sugar only: χ 2 = 5.23, df = 2, P>0.05, Fig 2B; Fed on sugar and blood: χ 2 = 3.49, df = 2, P>0.05, Fig 2C). The mean male longevity was 51.0 ± 3.1 d, 50.9 ± 3.1 d and 54.0 ± 3.1 d for the HC, GUA and GT strains, respectively. The mean longevity of females fed on sugar only was 47.85 ± 3.77 d, 44.57 ± 3.54 d and 38.57 ± 3.34 d for the HC, GUA and GT strains, respectively. The mean longevity of females fed on sugar and also offered blood meals was 47.51 ± 3.45 d, 43.14 ± 3.22 d and 40.3 ± 3.06 d for the HC, GUA and GT strains, respectively.

Discussion
The present study presents a comparative analysis of the impact of Wolbachia infections on the fecundity, fertility, developmental time, sex ratio, wing length and longevity of the three Ae. albopictus lines: the transinfected line HC (wAlbA, wAlbB and wPip), the naturally double-infected line GUA (wAlbB and wAlbB) and the aposymbiotic line GT. The comparative analysis was performed after all the lines were backcrossed to have a highly similar genetic background (>98%) as the naturally infected GUA line which originated from the Guangzhou metropolitan area, China. Our study showed that, except for a minor impact on the wing length and immature development time, Wolbachia infections (double or triple) had no effect on the fitness of the Ae. albopictus lines studied. Previous studies have reported that Wolbachia infections could have either a positive or negative impacton the fecundity of both fruit flies and mosquitoes. In the early 1990's it was reported that Wolbachia had a negative impact on the fecundity of D. simulans [39]; however, within less than three decades, the presence of Wolbachia become beneficial by enhancing the host fecundity [40]. In Ae. albopictus, different studies have produced different results. For example, Islam and colleagues reported that naturally double-infected (wAlbA and wAlbB) Ae. albopictus females could produce more offspring than uninfected females [41,42]. In contrast, the transinfected ARwP females of the single wPip-infected Ae.albopictus strain were shown to produce fewer eggs than either naturally double-infected (wAlbA and wAlbB) or aposymbiotic females during their lifetime [22]; however, the fecundity of ARwP females recovered in later generations [22]. The present study found that there was no difference in the fecundity of the three Ae. albopictus lines (HC, GUA and GT).This is consistent with the other studies on the triple Wolbachia-infected Ae. albopictus strains HouR (wAlbA, wAlbB and wRi) and two single wMel-infected Ae. albopictus strains Uju.wMel and HTM; however, wMelPop was shown to have major effects on the fitness of Ae. albopictus [31,32,34].  There have been reports showing that transinfected lines of insect species, for example of the Mediterranean fruit fly Ceratitiscapitata, exhibited a significant decrease in egg hatch rates [43,44]. In most cases, these phenomena are usually due to incomplete rescue or the occasional failure of some eggs to maternally inherit Wolbachia [42,43]. In some cases, fertility can be recovered through selection [45]. In Ae. albopictus, the triple-infected (wAlbA, wAlbB and wRi) HouR line exhibited a similar egg hatch rate to the naturally double-infected (wAlbA andwAlbB) line while a reduction in egg hatch was observed in the single-infected (wRi) HTR line [30,31]. Similar patterns were unraveled with the wPip-transinfected lines: no significant effect was observed in the triple Wolbachia-infected (wAlbA, wAlbB and wPip) HC strain while the single wPip-infected APwP strain showed a decreased egg hatch which was recovered in later generations [22,33]. However, the present study did not reveal any effect of the Wolbachia infections on the egg hatch rates either in the transinfected (wAlbA, wAlbB and wPip) or in the naturally double-infected (wAlbA and wAlbB) line. All three Ae. albopictus lines (HC, GUA and GT) exhibited very high survival rates (>88%) to adulthood which is an important advantage for mass rearing and large scale applications. Previous studies showed that Wolbachia negatively affects survival in the immature stages of Ae. albopictus, but not the adult sex ratio [42].Our study did not reveal differences either in either trait between HC, GUA and GT strains.
The potential effect of Wolbachia infections on the immature development of Ae. albopictus lines has been studied previously, however the observed differences could not be clearly attributed to the symbiont [42]. Similarly, no differences were revealed in our study using the naturally double-infected GUA strain and aposymbiotic GT strain. Interestingly, however, the triple-infected HC strain developed much faster (from L 1 to pupationand from L 1 to adult emergence) in both females and males compared to the GUA and GT strains. In contrast, the single Wolbachia (wPip)-infected Ae. albopictus strain ARwP showed no difference in development rate compared to the wild type double-infected (wAlbA and wAlbB) or the aposymbiotic strain [22]. This difference could be due to the different genetic background of the mosquito lines, the infection status or relative levels, the rearing conditions, the co-evolutionary history of these lines, or a combination of these factors. Interestingly, the shorter developmental time of the triple-infected HC strain might have a significantly positive impact on the mass rearing process by decreasing the rearing duration (by approximately half a day) and thus reducing rearing costs. In addition, this line has a more pronounced difference in the time to pupation on males than females. This might be of paramount importance for SIT and IIT applications, both of which depend on the availability of a perfect sexing system because the accidental release of females might increase pathogen transmission and certainly reduces the efficacy of the population suppression program [7,46]. For the IIT strategy, the accidental release of transinfected females may result in population replacement, rather than population suppression [7]. However, a highly efficient sex separation method is not available for Ae. albopictus [47]. So, the additional difference observed in the development of the triple-infected HC line between two sexes could be further exploited and might provide an opportunity to solve the current lack of an efficient means of sex separation [38,47]. An alternative approach would be to combine irradiation with IIT [7,23,48]. The application of low dose irradiation would sterilize the females while the released males would be fully sterile due to effects of both Wolbachia and irradiation [7,33]. This combined strategy is currently being considered for the population control of Ae. albopictus, a primary dengue vector in mainland China.
Wing length is commonly used as a proxy to determine the size of adult mosquitoes. It has been shown that rearing larvae at low densities results in larger mosquitoes. Short developmental time has also been associated with large adult mosquitoes [38,49,50]. In contrast, our study shows that the male developmental time of the HC strain is shorter than that observed for GUA and GT strains ( Table 2) and does not result in larger adult male mosquitoes (Fig 1). On the other hand, significant differences were detected in the size of adult females between the GUA and GT lines (Fig 1) despite the fact that their developmental time was similar ( Table 2). Previous studies could not reveal an effect of Wolbachia infection on the adult size of Ae. albopictus mosquitoes [42]. Even though the present study shows that differences exist between HC and GT males (2.01 ± 0.09 mm vs 2.07 ± 0.11 mm), and GUA and GT females (2.60 ± 0.09 mm vs 2.55 ± 0.10 mm), these are essentially very minor differences and likely to be of no major importance. Although single wMel-infected males were found to live longer than the wild type double-infected (wAlbA and wAlbB) or aposymbiotic males [32], most studies have not revealed any effect of Wolbachia infections (single, double or triple) on the longevity of Ae. albopictus males [22,36,41]. In corroboration, our study could not detect any effect of Wolbachia infection (either double or triple) on male or female longevity, irrespective of the feeding regime.
The development and application of SIT and IIT (or their combined use) for the control of Ae. albopictus depends on the mass rearing and production of high quality sterile male insects which compete with wild type males to mate with wild females in order to introduce sterility into the natural population [46,51]. Taken all together, the presence of Wolbachia wPip appears to have a minimal effect on the fitness of the artificially triple-infected (wAlbA, wAlbB and wPip) HC strains, with fecundity and fertility, developmental time and longevity being comparable to the wild type Ae. albopictus. These results support the feasibility of applying mass rearing and integrated SIT/IIT to control Ae. albopictus.