Cross-Generational Effects of Heat Stress on Fitness and Wolbachia Density in Aedes aegypti Mosquitoes

Abstract

Aedes aegypti mosquitoes infected with Wolbachia symbionts are now being released into the field to control the spread of pathogenic human arboviruses. Wolbachia can spread throughout vector populations by inducing cytoplasmic incompatibility and can reduce disease transmission by interfering with virus replication. The success of this strategy depends on the effects of Wolbachia on mosquito fitness and the stability of Wolbachia infections across generations. Wolbachia infections are vulnerable to heat stress, and sustained periods of hot weather in the field may influence their utility as disease control agents, particularly if temperature effects persist across generations. To investigate the cross-generational effects of heat stress on Wolbachia density and mosquito fitness, we subjected Ae. aegypti with two different Wolbachia infection types (wMel, wAlbB) and uninfected controls to cyclical heat stress during larval development over two generations. We then tested adult starvation tolerance and wing length as measures of fitness and measured the density of wMel in adults. Both heat stress and Wolbachia infection reduced adult starvation tolerance. wMel Wolbachia density in female offspring was lower when mothers experienced heat stress, but male Wolbachia density did not depend on the rearing temperature of the previous generation. We also found cross-generational effects of heat stress on female starvation tolerance, but there was no cross-generational effect on wing length. Fitness costs of Wolbachia infections and cross-generational effects of heat stress on Wolbachia density may reduce the ability of Wolbachia to invade populations and control arbovirus transmission under specific environmental conditions.

1. Introduction

Viral diseases such as dengue are on the rise. This is due to a suite of factors including shifting geographical distributions of vectors and human mobility around the world, exposing populations to new environmental sources of infectious agents [1]. Dengue is an increasing global threat and is one of the most significant arboviruses. With an increasing geographic range of dengue transmission, it is estimated that about 4 billion people are at risk of dengue infection in 128 countries [2].

Belonging to the Flaviviridae family, dengue virus consists of four different serotypes, namely DENV 1–4 [3]. Primary infection with dengue is usually self-limiting and confers life-long immunity against that serotype. However, a second infection with a different serotype from the first infection increases the risk of severe disease such as dengue shock syndrome and dengue haemorrhagic fever. This is possibly caused by antibody-dependent enhancement of virus infection, whereby cross-reactive antibodies will bind to but not neutralize the virus, worsening the condition of the infection [4]. Today, there are no available anti-viral or effective vaccines to target this disease.

The primary vector of dengue is Aedes aegypti, which is a mosquito that lives in tropical and sub-tropical regions of the world [5]. Aedes aegypti has evolved to live in urban environments in close proximity to humans. Female Ae. aegypti has a strong preference for feeding on humans, mainly during the daytime, and can enter houses to feed. Once infected with the virus, the mosquito is a carrier for life [6]. Ae. aegypti has also evolved to feed on blood from more than one person during a feeding period [7]. These behaviors of Ae. aegypti allow efficient transmission of dengue and other pathogenic arboviruses. Hence, reducing or modifying the population of Ae. aegypti in nature is essential to reducing the spread of arboviruses such as dengue.

Conventional approaches to limiting dengue transmission rely on controlling Ae. aegypti populations through trapping, removing stagnant water around homes, using mosquito nets and applying insect repellent. Authorities have also implemented routine fogging with adulticides during dengue outbreaks [8]. However, chemical vector control is becoming increasingly ineffective as Ae. aegypti has now developed resistance to multiple insecticides in many parts of the world [9].

Harnessing Wolbachia–mosquito symbiosis is an alternative approach to disease control that could potentially reduce the global burden of dengue and other mosquito-borne diseases [10]. Wolbachia are gram-negative bacteria found in many insect species, including mosquitoes, and can interfere with RNA virus replication by competing for cellular components needed for replication, particularly lipids [11]. When transferred from other insect species to Ae. aegypti through embryonic microinjection, some Wolbachia strains including wMel [12], which is found naturally in Drosophila melanogaster, and wAlbB [13], which occurs naturally in Aedes albopictus, limit the capacity for the mosquitoes to transmit viruses. In Ae. aegypti, defective cholesterol and cellular trafficking in Wolbachia-infected cells limits viral replication, while restoring cholesterol homeostasis recovers dengue replication in mosquito cells [14].

Wolbachia are transmitted maternally and often modify insect reproduction to enhance the production of infected female hosts, facilitating its spread throughout natural populations [10]. Several Wolbachia strains that have been transferred to Ae. aegypti cause cytoplasmic incompatibility, which results in sperm and eggs being unable to form viable offspring when an uninfected female mates with an infected male [12,15,16]. However, Wolbachia-infected females can successfully produce viable and Wolbachia-infected offspring with both infected and uninfected males. Cytoplasmic incompatibility can be exploited to introduce Wolbachia infections into mosquito populations, transforming natural populations with mosquitoes that are less capable of transmitting arboviruses. Large-scale releases of Wolbachia-infected mosquitoes have been carried out in Australia, Malaysia, Vietnam, Brazil and other countries in the tropics with this approach, in an attempt to reduce the burden of dengue [17].

In the field, Ae. aegypti larvae experience extreme diurnal fluctuating temperatures, especially in small containers of water that are exposed to direct sunlight or are made of good heat conductors, such as metal [18]. Ae. aegypti infected with the wMel strain have greatly reduced Wolbachia density when reared at maximum daily temperatures of 37 °C [19,20], similar to temperatures experienced in Cairns, Australia, during the wet season in some breeding sites [18]. Since Wolbachia density tends to be positively associated with virus blockage [21,22], heat stress could limit the ability of Wolbachia-infected mosquitoes to reduce virus transmission. Heat stress can also reduce the intensity of cytoplasmic incompatibility and fidelity of maternal transmission, potentially impairing the ability of wMel to invade natural populations [20]. However, it is not clear if these effects are transient or if the effects of heat stress persist across generations.

In this study, we looked at the cross-generational effects of heat stress on Ae. aegypti fitness and Wolbachia density over two generations. The effects on fitness were examined in terms of wing length, which is an indicator of fecundity, and adult starvation tolerance, which is an indicator of nutritional reserves and a trait which has not previously been evaluated in Wolbachia-infected mosquitoes. By understanding the cross-generational effects of heat stress, we aim to produce knowledge that can aid the use of Wolbachia as an agent for arbovirus control.

2. Materials and Methods

2.1. Ethics Statement

Blood feeding of female mosquitoes on human volunteers for this research was approved by the University of Melbourne Human Ethics Committee (approval 0723847). All adult subjects provided informed written consent (no children were involved).

2.2. Mosquito Strains and Colony Maintenance

Ae. aegypti mosquitoes infected with wMel were collected from Cairns, Queensland, Australia in 2013, in areas where the wMel infection had established [17], and uninfected mosquitoes were collected in 2016 from outside the release area. Ae. aegypti infected with wAlbB were generated previously by transferring Wolbachia from Ae. albopictus [16], followed by introgression into an Australian background [23]. Wolbachia-infected females were crossed to uninfected males regularly to maintain all colonies on a similar genetic background. All populations were maintained in the laboratory as described by Ross et al. [24].

2.3. Mosquito Rearing at Constant and Cyclical Temperatures

We investigated the effect of heat stress during larval development on mosquito fitness and Wolbachia density in Ae. aegypti across two generations. In the first generation, eggs of uninfected, wMel-infected and wAlbB-infected mosquitoes were hatched concurrently in 3 L trays of reverse osmosis (RO) water at 26 °C and provided with one 300 mg tablet of tropical fish food (TetraMin, Tetra, Melle, Germany). Hatching trays were kept in a controlled-temperature room at a constant temperature of 26 °C and a 12/12 h light/dark cycle.

Twenty-four hours after hatching, 200 larvae were added to plastic containers filled with 500 mL of RO water to control the larval density. Larvae were either reared at a constant temperature of 26 °C or subjected to a diurnal cyclical temperature of 26–37 °C (

Table 1. Experimental design. Aedes aegypti mosquito larvae for all three infection types (uninfected, wMel, and wAlbB) were reared at either 26 °C or 26–37 °C in the first generation. Offspring of parents from each treatment were then reared at either 26 °C or 26–37 °C.

Infection Type	Generation
	1	2
Uninfected	26 °C	26 °C
	26 °C	26–37 °C
	26–37 °C	26 °C
	26–37 °C	26–37 °C
wMel	26 °C	26 °C
	26 °C	26–37 °C
	26–37 °C	26 °C
	26–37 °C	26–37 °C
wAlbB	26 °C	26 °C
	26 °C	26–37 °C
	26–37 °C	26 °C
	26–37 °C	26–37 °C

) in incubators (PG50 Plant Growth Chambers, Labec Laboratory Equipment, Marrickville, NSW, Australia). The temperature cycle of 26–37 °C was consistent with previous studies [19,20] and was chosen to reflect temperatures experienced in breeding sites in Cairns during the wet season [18]. Four replicate trays of larvae were reared for each Wolbachia infection type and temperature. Temperature loggers (Thermochron; 1-Wire, iButton.com, Dallas Semiconductors, Sunnyvale, CA, USA) in zip-lock bags were placed in six containers at random to monitor temperature in the incubator (Figure A1). The position of the containers was randomly shuffled every day to account for location-dependent temperature effects.

The larvae were reared to the pupal stage by providing fish food tablets ad libitum and then returned to a constant temperature of 26 °C for the adult and egg stages. Emerging adults were collected at random and tested for their starvation tolerance (see below), while the remainder were maintained for experiments on the second generation. One week after adults emerged, the females were blood-fed on a single human volunteer, their eggs were collected and conditioned, then hatched four days after collection. Larvae from each infection type and rearing temperature in the first generation were reared at either 26 °C or 26–37 °C in the second generation, for a total of four temperature treatments for each infection type (Table 1). When adults from the second generation emerged, 20 males and 20 females from each treatment were selected at random and stored in absolute ethanol for wing length measurements and Wolbachia density quantification. The remaining adults were used in the starvation tolerance experiment.

2.4. Adult Starvation Tolerance

We investigated the effect of heat stress during larval development on adult starvation tolerance in the first and second generation. Starvation resistance has been extensively studied in other Diptera including Drosophila where this trait interacts with other forms of stress resistance and can be affected by carryover effects [25]. Adult mosquitoes were subjected to starvation conditions whereby 25 males and 25 females from each Wolbachia infection type and rearing temperature were transferred to 3 L cages and provided with water only. Each treatment was replicated four times. Adult starvation tolerance was determined by counting and removing dead mosquitoes from the cages each day until the last adult died.

For the experiment on the first generation, adult mosquitoes were provided with 10% sucrose for 4 days before being transferred to experimental cages. In experiments on the second generation, adult mosquitoes were not provided with sugar before the experiment and instead were transferred to 3 L cages within one day of emerging.

2.5. Wing Length

To determine the effect of heat stress on body size in the second generation, 15 wings from each sex and treatment were dissected and mounted on glass slides with Hoyer’s solution (dH₂O/gum arabic/chloral hydrate/glycerin in the ratio 5:3:20:2). Wing length was measured as the distance from the axial notch to the wing tip [26] using NIS-Elements BR (Nikon Instruments, Japan). Each measurement was repeated, and lengths were averaged to produce a final measurement.

2.6. Wolbachia Density

Wolbachia density for wMel-infected mosquitoes was quantified using a quantitative real-time polymerase chain reaction (RT-qPCR) assay. Genomic DNA was extracted from 12 adult mosquitoes from each treatment, using 150 μL of 5% Chelex 100 resin (Bio-Rad Laboratories, Hercules, CA, USA). Ae. aegypti-specific primers involving a region in the RpS6 gene and wMel-specific primers involving the w1 marker were then used to quantify Wolbachia density relative to the density of mosquito DNA with an RT-qPCR assay outlined by Lee and others [27]. PCR was carried out using a Roche LightCycler 480 system (Roche Applied Science, Indianapolis, IN) to obtain crossing point (Cp) values for these markers for each mosquito. Differences between the Cp of the Wolbachia and Ae. aegypti markers were transformed by 2ⁿ to obtain approximate estimates of Wolbachia density.

2.7. Statistical Analysis

All data collected were analysed using GraphPad Prism v.7. Log-rank Mantel Cox tests were performed to compare differences in adult starvation tolerance between groups. We used general linear models to investigate the effects of sex, Wolbachia infection type and temperature regime on wing length and to compare Wolbachia density between temperature regimes.

3. Results

3.1. Adult Starvation Survival

3.1.1. Generation 1

We tested the tolerance of uninfected, wMel-infected and wAlbB-infected adults to starvation when larvae were reared at 26 °C or 26–37 °C (Figure 1). Adults that were reared at 26 °C had a higher starvation tolerance than those reared at 26–37 °C, with differences being highly significant for both females (Log-rank test: χ² = 9.493, df = 1, p = 0.002) and males (χ² = 18.243, df = 1, p < 0.001). Differences between males and females were substantial (χ² = 537.901, df = 1, p < 0.001), with females outliving males under starvation conditions. wMel-infected (females: χ² = 65.814, df = 1, p < 0.001, males: χ² = 8.414, df = 1, p = 0.004) and wAlbB-infected (females: χ² = 94.552, df = 1, p < 0.001, males: χ² = 37.541, df = 1, p < 0.001) adults had reduced starvation tolerance relative to uninfected mosquitoes. wAlbB-infected adults had lower starvation tolerance than wMel-infected adults for both females (χ² = 5.546, df = 1, p = 0.019) and males (χ² = 19.960, df = 1, p < 0.001).

3.1.2. Generation 2

We evaluated the cross-generational effects of heat stress on adult starvation tolerance. In contrast to the experiments on the first generation, males had a higher tolerance to starvation than females (Log-rank test: χ² = 154.502, df = 1, p < 0.001). This likely reflects differences in methodology; adults were starved immediately after emergence in the second generation, while in the first generation they were fed sucrose first.

Adults reared at 26 °C in the second generation had higher starvation tolerance than adults reared at 26–37 °C (females: χ² = 94.661, df = 1, p < 0.001, males: χ² = 56.243, df = 1, p < 0.001), consistent with the first experiment. We found cross-generational effects of heat stress on starvation tolerance in females but not in males (Figure 2). Females reared at 26°C in the second generation had higher starvation tolerance when their parents had experienced heat stress (χ² = 49.504, df = 1, p < 0.001), but there was no cross-generational effect when second-generation females were reared at 26–37°C (χ² = 3.635, df = 1, p = 0.057). There was no cross-generational effect of heat stress in males; starvation tolerance when the second generation was reared at 26 °C (χ² = 1.386, df = 1, p = 0.239) or 26–37°C (χ² = 3.373, df = 1, p = 0.066) was unaffected by the rearing temperature of the first generation.

When considered across all temperatures, uninfected adults had a higher starvation tolerance than Wolbachia-infected adults for both males (wMel: χ² = 40.884, df = 1, p < 0.001; wAlbB: χ² = 16.743, df = 1, p < 0.001) and females (wMel: χ² = 37.821, df = 1, p < 0.001; wAlbB: χ² = 46.963, df = 1, p < 0.001, Figure 3). wAlbB-infected males had higher starvation tolerance than wMel-infected males (χ² = 7.227, df = 1, p = 0.007), but tolerance did not differ between the two strains for females (χ² = 0.059, df = 1, p = 0.809).

3.2. Wing Length

We measured wing length from a sample of adults from the second generation as an estimate of body size. Females (mean ± SE = 2.835 ± 0.0155 mm, n = 160) were much larger than males (mean ± SE = 2.167 ± 0.0113 mm, n = 147, general linear model: F_1,305 = 1182.338, p < 0.001, Figure 4). There was a significant effect of the Wolbachia infection type on wing length for males (F_2,135 = 13.982, p < 0.001), with wMel-infected males being smaller than the other two infection types, but there was no effect in females (F_2,148 = 1.553, p = 0.215). We found no cross-generational effects for this trait; heat stress during the first generation had no bearing on wing length in the second generation for both males (F_1,135 = 0.101, p = 0.751) and females (F_1,148 = 2.090, p = 0.150). However, differences related to the rearing temperatures during the second generation were substantial for both males (F_1,135 = 522.311, p < 0.001) and females (F_1,148 = 416.768, p < 0.001), with mosquitoes reared at 26 °C being much larger than mosquitoes reared at 26–37°C (Figure 4).

3.3. Wolbachia Density

We measured Wolbachia density in wMel-infected adults to see if the rearing temperature in the first generation affected density in the second generation (Figure 5). Wolbachia density was higher in adults reared at 26 °C during larval development than in adults reared at 26–37 °C (females: general linear model: F_1,22 = 176.844, p < 0.001, males: F_1,22 = 100.051, p < 0.001). We found cross-generational effects of heat stress on Wolbachia density in females but not in males. Wolbachia density was lower in females reared at 26°C in the second generation when the first generation was reared under heat stress (F_1,22 = 44.391, p < 0.001). Wolbachia density was also lower in females reared at 26–37 °C in the second generation when the first generation was reared under heat stress, but this difference was marginally non-significant (F_1,22 = 4.140, p = 0.054). In contrast to females, there was no cross-generational effect of heat stress on Wolbachia density in males for either rearing temperature in the second generation (26°C: F_1,22 = 0.466, p = 0.502, 26–37 °C: F_1,20 = 1.688, p = 0.211). These results indicate that the effects of heat stress on female Wolbachia density can accumulate across generations, but Wolbachia density can also partially recover in the next generation in the absence of heat stress.

4. Discussion

In this study we evaluated the cross-generational effects of heat stress on Wolbachia density and fitness in Wolbachia-infected and uninfected Ae. aegypti. We found that diurnal fluctuating temperatures of 26–37 °C during larval development reduced adult starvation tolerance, wing length and Wolbachia density. Heat stress during the first generation influenced Wolbachia density and adult starvation tolerance in the following generation, but only in females. We also found that Wolbachia infections reduced adult starvation tolerance, adding to the growing list of phenotypic effects influenced by Wolbachia infections in Ae. aegypti [28].

Heat stress during development reduces Wolbachia density and cytoplasmic incompatibility intensity in wMel-infected Ae. aegypti [20,29,30], though Wolbachia density seems to partially recover when individuals are returned to cooler temperatures [19]. Few studies have measured the effects of temperature on Wolbachia infections across generations, except in cases where high temperatures were being used to deliberately cure Wolbachia infections [31], though these tended to measure infection frequencies rather than density [32,33]. In the parasitoid wasp Leptopilina heterotoma, Wolbachia density in daughters was not influenced by the rearing temperature of the mother (20 or 26 °C) [34], but the effects of temperature on Wolbachia density across generations have not been tested previously in Ae. aegypti.

In this study, we found that Wolbachia density partially recovered in female offspring of parents that experienced heat stress. In contrast, there was no effect of heat stress during the first generation on Wolbachia density in the second generation for males. This result was unexpected given that Wolbachia density responses to high temperatures are usually similar between males and females in Ae. aegypti [20,29]. Our results suggest that the effects of heat stress on Wolbachia density can accumulate across generations, at least in females. Long periods of hot weather (across multiple mosquito generations) could lead to a continuous decline in Wolbachia density, while short periods within a single generation may affect density in the following generation. Given that Wolbachia density is positively associated with virus blockage [21,22,35,36] and cytoplasmic incompatibility [37,38] in other systems, there may also be cross-generational effects of heat stress on these phenotypes in Ae. aegypti. However, since Wolbachia density partially recovered in the next generation in the absence of heat stress, it is unlikely that these effects would persist for more than a couple of generations under cooler conditions.

A large range of fitness effects have now been identified for Wolbachia infections in Ae. aegypti, with most research focusing on the wMelPop and wMel strains which have both been deployed into the field in disease control programs [17,39,40,41]. Wolbachia infections in Ae. aegypti tend to reduce fitness, particularly in terms of fertility [42], adult lifespan [15,23,43] and quiescent egg viability [44,45]. Despite much research on fitness effects of Wolbachia infections, there is relatively little research on the effects of Wolbachia infections on starvation tolerance. Previous studies in adult Drosophila found no effect of Wolbachia on this trait [46,47], but the wMel, wAlbB and wMelPop infections in Ae. aegypti reduce the survival of larvae without food, with the severity of the effect depending on the strain [48]. Johnson et al. [49] evaluated the survival of wild-type and wMel-infected male Ae. aegypti adults after the removal of honey solution and found that wild-type males survived slightly longer, though the two infection types were not compared statistically.

Here, we show that the Wolbachia infections wMel and wAlbB reduced adult starvation tolerance in Ae. aegypti, with no consistent differences between the two strains when considering both experiments. Wolbachia infections in Ae. aegypti can increase adult metabolic rate [50] and compete with their host for resources [51], which may explain the costs to starvation tolerance seen here. While the effects of Wolbachia on adult starvation tolerance are relatively subtle, fitness may be reduced under certain conditions in the field where nutrition is not readily available. Since female Ae. aegypti rarely feed on sugar [52], Wolbachia infections may shorten lifespan if females are not able to feed on blood regularly. Wolbachia may also reduce the opportunity for males to reproduce in the absence of sugar, which is their only source of nutrition.

5. Conclusions

In summary, we demonstrated that diurnal fluctuating temperatures of 26–37 °C during Ae. aegypti larval development reduced wMel Wolbachia density, with effects persisting into the next generation for females but not males. These findings have implications for the stability of Wolbachia infections across generations in tropical environments. The costs of Wolbachia infections to adult starvation tolerance observed here add to an increasing array of traits that are adversely affected by Wolbachia, which together will reduce the potential for Wolbachia to invade natural populations of Ae. aegypti.