Skip to main content
  • Loading metrics

The effect of repeat feeding on dengue virus transmission potential in Wolbachia-infected Aedes aegypti following extended egg quiescence

  • Meng-Jia Lau,

    Roles Conceptualization, Investigation, Writing – original draft, Writing – review & editing

    Affiliations Department of Biology, The Pennsylvania State University, University Park, Pennsylvania, United States of America, The Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, Pennsylvania, United States of America

  • Andrés R. Valdez,

    Roles Formal analysis, Methodology, Writing – original draft, Writing – review & editing

    Affiliations The Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, Pennsylvania, United States of America, Biomedical Engineering, The Pennsylvania State University, University Park, Pennsylvania, United States of America

  • Matthew J. Jones,

    Roles Investigation, Writing – review & editing

    Affiliations Department of Biology, The Pennsylvania State University, University Park, Pennsylvania, United States of America, The Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, Pennsylvania, United States of America

  • Igor Aranson,

    Roles Supervision, Writing – review & editing

    Affiliations The Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, Pennsylvania, United States of America, Biomedical Engineering, The Pennsylvania State University, University Park, Pennsylvania, United States of America

  • Ary A. Hoffmann,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliation Pest and Environmental Adaptation Research Group, Bio21 Institute and The School of Biosciences, University of Melbourne, Parkville, Victoria, Australia

  • Elizabeth A. McGraw

    Roles Conceptualization, Funding acquisition, Project administration, Writing – original draft, Writing – review & editing

    Affiliations Department of Biology, The Pennsylvania State University, University Park, Pennsylvania, United States of America, The Huck Institutes of the Life Sciences, The Pennsylvania State University, University Park, Pennsylvania, United States of America


As Wolbachia pipientis is more widely being released into field populations of Aedes aegypti for disease control, the ability to select the appropriate strain for differing environments is increasingly important. A previous study revealed that longer-term quiescence in the egg phase reduced the fertility of mosquitoes, especially those harboring the wAlbB Wolbachia strain. This infertility was also associated with a greater biting rate. Here, we attempt to quantify the effect of this heightened biting behavior on the transmission potential of the dengue virus using a combination of assays for fitness, probing behavior, and vector competence, allowing repeat feeding, and incorporate these effects in a model of R0. We show that Wolbachia-infected infertile mosquitoes are more interested in feeding almost immediately after an initial blood meal relative to wild type and Wolbachia-infected fertile mosquitoes and that these differences continue for up to 8 days over the period we measured. As a result, the infertile Wolbachia mosquitoes have higher virus prevalence and loads than Wolbachia-fertile mosquitoes. We saw limited evidence of Wolbachia-mediated blocking in the disseminated tissue (legs) in terms of prevalence but did see reduced viral loads. Using a previously published estimate of the extrinsic incubation period, we demonstrate that the effect of repeat feeding/infertility is insufficient to overcome the effects of Wolbachia-mediated blocking on R0. These estimates are very conservative, however, and we posit that future studies should empirically measure EIP under a repeat feeding model. Our findings echo previous work where periods of extensive egg quiescence affected the reproductive success of Wolbachia-infected Ae. aegypti. Additionally, we show that increased biting behavior in association with this infertility in Wolbachia-infected mosquitoes may drive greater vector competence. These relationships require further exploration, given their ability to affect the success of field releases of Wolbachia for human disease reduction in drier climates where longer egg quiescence periods are expected.

Author summary

Wolbachia pipientis is a naturally occurring, maternally inherited insect endosymbiont that was artificially introduced into the mosquito, Aedes aegypti, the dominant vector of dengue, Zika, chikungunya, and Yellow Fever viruses globally. This bacterium is being released into mosquito populations in the field for disease control because Wolbachia reduces the ability of these co-infecting viruses to replicate as well as spread easily through populations due to a form of reproductive manipulation. Previous research has shown that mosquito eggs experiencing long periods of storage make Wolbachia-infected adults more likely to be infertile and exhibit increased biting frequency. Using a combination of behavioral, vector competence, and fitness assays as well as modeling, we show that the increase in biting behavior leads to higher virus prevalence and load in the mosquito. Together, the decreased fertility and the increased virus transmission potential may decrease the efficacy of Wolbachia, particularly in dry climates where mosquitoes likely spend longer in the egg stage before hatching. We highlight a range of future experiments needed to fully ascertain the effect of egg quiescence on virus transmission.


Wolbachia, a maternally inherited endosymbiotic bacterium widespread in arthropods, reduces the replication of co-infecting viruses in its insect hosts [1,2]. Wolbachia is not naturally found in the global mosquito vector, Aedes aegypti, but several strains have been transinfected into the mosquito from multiple donor insect species [1,3], forming stable infections. Wolbachia limits the replication of the arboviruses dengue, chikungunya [2], Zika [4], and Yellow Fever [5] inside the vector, reducing the potential for transmission. It also limits the reproductive success of Wolbachia-free females in populations via the action of cytoplasmic incompatibility (CI), leaving Wolbachia-infected females to populate a greater proportion of the next generation, assisting with symbiont spread [6,7]. Wolbachia-mediated ‘pathogen blocking,’ in combination with its self-spreading abilities, form the basis of a global strategy to use the symbiont to limit the incidence of human arboviral diseases. The release of Wolbachia into native Ae. aegypti populations have led to its spread and the replacement of local mosquito populations in numerous field trials [811]. High rates of Wolbachia infection in mosquitoes post-release have then been associated with substantial reductions in dengue fever incidence in humans [10,12].

The long-term success of these Wolbachia-based ‘population replacement’ strategies depends on at least four factors: the successful vertical transmission of Wolbachia, the strong expression of CI, limited fitness costs associated with Wolbachia infection, and the ongoing induction of Wolbachia-mediated pathogen blocking [13]. The strength of CI expression is known to be correlated with symbiont density (reviewed in [14]). High ambient temperatures can reduce Wolbachia densities, causing weakening of CI [15] and maternal transmission failure [16]. Fitness costs to hosts infected with Wolbachia, while often mild, can be greater in the presence of high temperatures [17]. High-temperature conditions have been proposed to explain the poor spread and maintenance of Wolbachia in at least one release site [18]. The factors that will affect expression levels of pathogen blocking are less clear. Mosquito genetic background may matter [19], as well as Wolbachia strain x mosquito interactions [20]. In the releases in Brazil, for example, the efficacy of Wolbachia varied across release zones and did not entirely correlate with the frequency of Wolbachia in the population, suggesting an interaction with local environmental or genetic factors [11,21]. Thus, the success of the ‘population replacement’ strategies seems to depend on the local context.

Even though ~ten different Wolbachia strains have been established in Ae. aegypti from a range of donor insect species [22], only variants of two strains have been successfully released and established in the field in replacement strategies: wMel [10,21] and wAlbB [8]. The wAlbB strain, originally from Aedes albopictus [3], shows better resistance to both higher and lower temperatures [23,24] compared to wMel, originally from Drosophila melanogaster [1]. Subsequent laboratory investigation found, however, that females infected with wAlbB can be infertile as adults and lack matured ovaries if they spend an extended period in the egg stage [25]. Infertility rates can reach 75% in females hatched from eggs stored in a humidity-controlled environment for 11 weeks. This reduction suggests that field releases of this strain may encounter difficulty in regions with distinct dry seasons preventing continuous hatching of eggs. Interestingly, the resulting infertile females also demonstrated an increased feeding frequency, with the potential to drive greater transmission of viruses when the rainy season does arrive [26].

The degree of virus transmission is highly dependent upon mosquito population size, survival, feeding frequency, infection rate, and the extrinsic incubation period (EIP) [27]. Wolbachia-mediated pathogen blocking results in a reduction in the infection rate and the viral load [2] and lengthens the EIP [28]. Increased feeding due to egg storage may act counter to pathogen blocking. Studies of Wolbachia’s effect on virus transmissibility in the laboratory are frequently carried out on mosquito lines with little appreciable egg storage time and usually after a single blood feed, where the mosquitoes feed to repletion [1,2]. In the field, egg quiescence may be common in areas with dry seasons, and successive feeding within a single gonotrophic cycle is common, likely due to the consumption of smaller blood meals on live hosts [29,30]. In this study, we examine how Wolbachia-associated infertility and increased blood-feeding post-egg quiescence affect DENV transmissibility. We also examine key life history traits, including longevity and probing behavior. With vector competence and fitness measures, we are then able to model the effect of infertility and increased feeding rate on transmissibility.

Materials and methods

Mosquito rearing

Two mosquito lines, a Wolbachia wAlbB-infected and wildtype (Wolbachia-free), were used in this experiment. The wAlbB Wolbachia strain originated from Aedes albopictus and was previously microinjected into Ae. aegypti [3]. Just prior to these experiments, wAlbB-infected Ae. aegypti females were backcrossed for 3 generations to wild-type males collected from Mérida, Mexico, in 2018. This Mérida line also served as the wildtype, Wolbachia-free control. All mosquitoes were maintained in the laboratory under 26 ± 1°C and 68 ± 5% relative humidity, with a 12:12 hour light: dark cycle. A previous study compared the effect of egg storage (quiescence) for 3, 6, and 11 weeks and showed maximal reductions in fertility and fecundity of Wolbachia-infected Ae. aegypti [25] at 11 weeks. We, therefore, stored all mosquito eggs (+/- Wolbachia) for 11 weeks prior to hatching under the environmental conditions defined above.

Mosquito larvae were reared at a density of 100 larvae/L of water and fed Tropical Fish Food (Tetramin: Tetra Werke, W. Germany). Adults were housed in 18 × 18” square breeding cages (BioQuip) in populations of 300 individuals and fed with 10% sucrose solution ad libitum. Fresh human blood from anonymous donors (BioIVT: Westbury, NY) was used for all blood-feeding using a Hemotek membrane feeder (Hemotek Ltd., UK). Post-blood feeding, mosquitoes were immobilized by chilling to sort them into smaller experimental containers.

Life history traits

We measured the longevity and the probing frequency of Wolbachia-infected or uninfected female mosquitoes hatched from quiescent eggs as they are key determinants of mosquito virus transmission at the population level. At 5 ± 1 days post-emergence, mosquitoes were blood-fed (no virus), and then engorged females were relocated into 32 oz plastic containers with mesh lids for experimental groups and replicates. As per a previous report, adult females infected with Wolbachia, having experienced extended egg quiescence, exhibited high rates of infertility as caused by undeveloped ovaries (see images within [26]). To confirm that our Wolbachia-infected lines (all stored as eggs) also demonstrated high rates of infertility compared to Wolbachia-uninfected controls, we dissected 100 Wolbachia wAlbB-infected and 20 Wolbachia uninfected female mosquitoes and calculated the rates of undeveloped ovaries for each. For longevity, each line was represented by five replicate containers, each containing 30–40 individuals. Mosquito survival was checked every two days from 0 to 62 days post-feeding. During this period, mosquitoes were provided with 10% sucrose and 50% larval rearing water using 2 cotton balls, respectively each changed every 2 days. Once per week, mosquitoes were offered a blood meal for 15 minutes using the Hemotek through mesh lids.

For the probing frequency experiment, each line was represented by 7 replicate containers, each containing 30–40 individuals. Over seven days post the first blood meal, warmed blood was provided once per day (24 ± 1 hours). A probing event was described as follows: an individual landed on the feeder and showed probing or probing attempts for at least 3 seconds. Multiple separate feeding events could be scored for the same mosquito if it left the feeder and then returned for probing during the ten-minute assay window. All mosquitoes/treatments were fed within 2 hours, and the order of feeding each day was randomized. A subset of containers was observed and scored in real time by the experimenter. Given the scale of the design, the remainder were scored after watching videos. Mosquitoes were deprived of both sucrose and water one hour before the start of feeding until the end of the trial.

Dengue virus infection and quantification

We used the DENV-2 (ET-300; GenBank: EF440433.1) strain, cultured in C6/36 Ae. albopictus cells to test the transmissibility of mosquitoes after hatching from quiescent eggs. Cells were grown at 25°C in RPMI 1640 medium (Life Technologies, Carlsbad, CA, USA) with 10% sterile fetal bovine serum (FBS, Life Technologies) and 20 mM HEPES (Sigma-Aldrich, St. Louis, MO, USA). For infection, cells were allowed to grow to 80% confluency, then the medium was replaced with fresh RPMI with 2% FBS, and the virus was inoculated at an MOI of 0.1. Viral concentration of the supernatant was quantified 7 days post-infection via squash buffer extraction (10 mM Tris (pH 8.2), 1 mM EDTA, and 50 mM NaCl) and DENV-G qPCR [31]. The virus was diluted in RPMI medium to a concentration of 1 x 107 copies/ml and then mixed 1:1 with blood in a flask for mosquito feeding. This concentration was chosen according to a previous study, where ~half of the mosquitoes were expected to be infected 12 days post-infection [32]. The concentration of virus in blood was also quantified, 30 μL blood was pipetted into 270 μL TRI Reagent (Sigma-Aldrich, Cat. no. T9424) and extracted using a Direct-zol RNA 96 Magbead Zymo kit (Zymo Research) on a MagMAX Express 96 system (Applied bio- systems), then we qPCR to quantify the copy of dengue virus using a LightCycler 480 instrument (Roche) following the methods described previously [31]. The viral load in blood for feeding each day was determined by averaging the results from two technical replicates from the same flask.

Wolbachia infected or wildtype lines hatched from eggs that had been stored for 11 weeks were separated into two groups at 5 ± 1 day post-emergence: one was fed with dengue infectious blood once; the other was fed and then provided with an infectious blood meal once every subsequent day for the following 7 days. After the initial infectious blood meal, mosquitoes were immobilized on ice, and engorged individuals were sorted into 16 oz containers for further collection. Each container held ~ 30 mosquitoes. Each infectious feed lasted for 30–40 minutes, and cotton balls soaked with sucrose and water were provided during the feeding. Mosquito leg samples were collected at 8, 11, and 14 days post the initial infectious (dpi) feeding for virus quantification. We chose to study legs for safety and ease in a large-scale design (dissection post-freezing) and because a recent study demonstrated the high efficacy of leg tissue loads in predicting transmission potential with an animal model [33]. A biological replicate was repeated with mosquitoes from different containers under the same experimental conditions. We tested the prevalence and viral load in legs to indicate virus dissemination [33], using 24 individuals per line. At the point of dissection, legs were stored in 300 μL of TRI Reagent and stored at -80°C for virus quantification via high throughput RNA extraction methods. Bodies from the Wolbachia infected line were also dissected for Wolbachia load determination via DNA extraction. Before legs collections, female fertility was first determined under a microscope (Carl Zeiss Microscopy GmbH, Göttingen, Germany) by assessing the existence of ovaries [26]. This allowed us to score whether the individual leg sample was from a fertile or infertile female. All viral loads were then assessed across 6 treatment groups. All Wolbachia uninfected mosquitoes were fertile, and so there were only 2 treatment classes- ‘fed once’ or ‘fed multiple times’. Wolbachia infected lines, in contrast, exhibited mixed infertility, and so in addition to ‘fed once’ and ‘fed multiple times,’ there was also an additional classification of ‘fertile’ or ‘infertile’ (4 total classes). Samples were homogenized using a Bead Ruptor Elite, and then RNA was extracted using a Direct-zol RNA 96 Magbead Zymo kit (Zymo Research, Irvine, CA) on a MagMAX Express 96 system (Applied bio- systems) according to the manufacturer’s instructions. Dengue virus copy number was then determined in a LightCycler 480 instrument following methods described previously [31]. Samples with a Ct value > 32 were considered as negative because at late Ct the standard curve no longer exhibited a linear relationship. After qPCR, DENV infection prevalence in legs was analyzed using a generalized linear model following a Wald test in R studio software. A log10 transformation was applied to viral load data prior to ANOVA.

Wolbachia quantification

We quantified Wolbachia densities in bodies (minus legs) after the single feeding. Bodies were chosen to include the numerous tissues [34] where blocking is likely to occur (midgut, fat body, etc.). In brief, after dissection, bodies were stored in 300 μL of TRI Reagent, and a Direct-zol DNA/RNA Miniprep kit (Zymo Research, Cat No. R2080) was used to extract DNA, followed by a qPCR method to test the relative densities of Wolbachia using primers for a Wolbachia gene and a host control gene as described previously [1,2]. Wolbachia densities were also log-transformed and compared using a Student’s t-test.

Statistical analysis

Data analysis and visualization were performed in R studio software (2022.07.1). We used a log-rank survival test for longevity data through package ‘survival’ [35] and ‘survminer’ [36]. For the probing behavior data, viral loads, and prevalence, a generalized linear model for Gaussian distribution was used to compare the effect of Wolbachia and the day of feeding, with “emmeans” package used to perform post hoc analysis. The Mann-Whitney U test was used for pair-wise comparisons between Wolbachia-infected and wild-type mosquitoes on each day.


Following Smith et al. [27] we evaluated a Ross and Macdonald index, where R0 estimates the number of new dengue-infected hosts linked to a single mosquito. To avoid human and external factor dependencies, we re-write the R0 as: (1)

The parameters are as follows; g is the time-dependent death rate, obtained from the survival rate (Fig 1), a is the biting rate (Fig 2), c represents the viral infection rate or prevalence (Fig 3), and v represents the Extrinsic Incubation Period (EIP). Given our multiple feeding design, we could not directly estimate viral prevalence at early time points; we, therefore, selected the earliest days of DENV arrival in the saliva from a previously published model for ‘Wolbachia uninfected’ as 5 days and 7 days for both classes of Wolbachia-infected mosquitoes based on wMel and DENV serotype 2 data [37]. Holding EIP constant across infertile and fertile groups, and for both the single and repeat-feeding groups was a very conservative choice. We did attempt to fit a sigmoidal curve to our prevalence data to estimate EIP but found this unreliable without time points before 8 days. To introduce variation in all measures, we perturbed values by 25%. We used the ChaosPy library [38] to perform this task, 10000 samples were generated, retrieving 10000 model evaluations.

Fig 1. Longevity of Wolbachia-infected and uninfected Ae. aegypti females hatched from eggs that were stored for 11 weeks.

Five replicates, each with 30–40 mosquitoes, were measured for each group. Data are for females that became fully engorged after feeding 5 ± 1 days post-emergence. Females were followed for up to 66 days post-feeding. The shaded areas represent the 95% confidence intervals.

Fig 2. Probing frequency of Wolbachia-infected and uninfected Ae. aegypti females hatched from eggs that were stored for 11 weeks.

Seven replicates, each with 30–40 mosquitoes, were measured for each group. Data are for females that became fully engorged after feeding 5 ± 1 days post-emergence. The curves are fitted using a LOESS method and formula = ’y ~ x’. The shaded areas represent the 95% confidence intervals.

Fig 3. DENV-2 viral infection following repeat feeding after 11 weeks of egg storage for Replicate 1.

(A) infection prevalence and (B) load in adult female Aedes aegypti legs at 8, 11, and 14 days post-infection (dpi). Females were provided with infectious blood every day from 1 to 7 dpi (post the initial infectious blood meal at 0 dpi). Each collection point x treatment is represented by 24 individuals. The fertility status of Wolbachia-infected females was identified through ovarian dissection.


Life history traits

Life history traits can affect mosquito vectorial capacity. Here we tested the longevity and probing frequency between the Wolbachia wAlbB-infected line and the wildtype line after their eggs were quiescent for 11 weeks. Infertility was the dominant feature of the Wolbachia infected line post-egg storage, with 81% of females (n = 100) scoring as infertile based on dissection. In contrast, 0% of wild type females (n = 100) were infertile post egg storage. Wolbachia-infected mosquitoes also died faster than wildtype (Fig 1. χ2 = 12.8, df = 1, p < 0.001). The average survival for Wolbachia-infected females was 40 days versus 46 days for wild type. We then examined probing frequencies for fully engorged females after their first blood meal over a 7-day time course (Fig 2). The feeding frequency was significantly higher for the Wolbachia-infected line compared to the wildtype (GLM: family = Gaussian, t(108) = 2.07, p = 0.04). There was also a significant effect on the day of feeding (GLM: family = Gaussian, t (108) = 5.08, p < 0.001). The Wolbachia-infected line showed a significantly higher rate of probing on days 2, 3, 4, and 5 (Mann-Whitney U tests: day 1: Z = -1.00, p = 0.32; day 2: Z = -3.40, p < 0.001; day 3: Z = - 3.38, p < 0.001; day 4: Z = -2.26, p = 0.02; day 5: Z = -2.21, p = 0.03; day 6: Z = -1.63, p = 0.10; day 7: Z = -1.10, p = 0.27). In these first few days, the wild type line shows little interest in probing whereas the rates by the Wolbachia-infected line are ~10-fold greater. In later dpi, as probing interest rises in the wildtype line, the Wolbachia-infected still feed at roughly twice the rate.

Dengue virus prevalence and load

All mosquitoes were fed with blood mixed with DENV-2 virus at 5 ± 1 days post-emergence, and after this initial infectious blood meal, they were divided into two groups: ‘blood-fed once’ and ‘blood-fed repeatedly.’ Females from the latter were provided with an infectious blood meal every day from 1 to 7 dpi. Viral prevalence and load were quantified in leg tissues at 3 time points post-feeding. Legs were chosen as they have recently been described as a better proxy for transmission in wild type mosquitoes [33]. Before RNA extraction, the bodies of Wolbachia-infected females were dissected under a microscope and used to score individuals as either fertile or infertile based on ovary appearance. For uninfected mosquitoes, there are 2 classes (fed once or multiply). For Wolbachia-infected, there are four (fed once or multiply by x fertile vs. infertile). Because ‘experimental replicate’ was significant in the model for both infection prevalence (χ2 = 25.04, df = 1, p < 0.001), and viral load (F1,375 = 29.71, p < 0.001), we analyzed the two replicate experiments separately.

Not surprisingly, infection prevalence (Figs 3 and S1A) rose with time in both replicate experiments (replicate 1: χ2 = 16.97, df = 2, p < 0.001; replicate 2: χ2 = 38.01, df = 2, p < 0.001). Similarly, the multiple feeding events also increased prevalence in both replicates (replicate 1: χ2 = 4.65, df = 1, p = 0.03; replicate 2: χ2 = 16.20, df = 1, p < 0.001). Replicate 1 prevalence was much higher than replicate 2, with several reaching 100% (Figs 3 and S1A). Our data were surprising in that the wild type line either showed lower prevalence in the leg tissue (replicate 1: χ2 = 10.43, df = 2, p = 0.005) or there was no difference compared to Wolbachia-infected (replicate 2: χ2 = 1.63, df = 2, p = 0.44), suggesting that Wolbachia did not provide clear or consistent blocking at the level of dissemination to the legs. When we provided pairwise comparisons for Wolbachia-infected fertile and infertile females, infertile females showed higher infection prevalence than fertile females upon multiple feeding at 8 dpi (post hoc upon GLM: replicate 1: z = 2.197, p = 0.028), but not for 11 or 14 dpi or any of the time points in replicate 2 (all p > 0.05).

The viral load data, in contrast, did reveal strong evidence of blocking in the Wolbachia-infected across all dpi. In replicate 1 (Fig 3), ‘dpi’ (F2,215 = 16.21, p < 0.001), feeding frequency (F1,215 = 24.88, p < 0.001), and ‘mosquito line’ (F2,215 = 8.08, p < 0.001) were significant. Whereas in replicate 2 (S1B Fig), only the ‘mosquito line’ (F2,160 = 8.97, p < 0.001) exhibited a strong effect. The interactions between ‘dpi’ and the ‘feeding frequency’ (replicate 1: F2,215 = 3.54, p = 0.031; replicate 2: F2,160 = 7.96, p < 0.001) or ‘mosquito line’ (replicate 1: F4,160 = 3.02, p = 0.018; replicate 2: F4.160 = 2.49, p = 0.045) were also significant. On average, across treatments, DENV loads in Wolbachia-infected lines were 4.44 × 104 (infertile) and 4.84 × 104 (fertile), compared to uninfected mosquitoes where the copy number was 1.45 × 105. When the Wolbachia-infected infertile females were blood-fed repeatedly, the viral load increased ~3 fold to 1.33 × 105. There was only a two-fold increase for the other two lines between single and repeat feeding to 1.06 × 105 (Wolbachia-infected fertile) and 2.14 × 105 (wildtype).

Last, we tested for Wolbachia density after a single DENV-2 feed (S2 Fig), and all mosquitoes from the Wolbachia-infected line were infected. We saw no large difference in Wolbachia loads between infertile and fertile females, although there was a trend with the former being higher (F1,68 = 3.76, p = 0.057).


As described above, we used published estimates of EIP (5 days for wild type, 7 days for Wolbachia infected). These are conservative estimates, as EIP may be shorter with repeated feeding, especially for the Wolbachia-infertile mosquitoes given their probing behavior in the first few days (Fig 2). Our estimates of survival also contained only two classes (wild type, Wolbachia). Given the predominance of Wolbachia-infected mosquitoes were infertile (~80%), and that previous work has shown that Wolbachia-fertile mosquitoes do not feed more frequently than wild type [26], we used the latter estimate for the Wolbachia-fertile biting rate. Because the distributions for R0 are non-symmetrical (Fig 4), metrics like the mode and the 90% confidence interval region, rather than the mean, are more useful in comparing across treatments (Table 1). After a single feeding, as expected the Wolbachia-free line exhibited the largest R0 (Fig 4A and Table 1) compared to the Wolbachia-infected lines. Only for the Wolbachia-infected lines does variation in R0, increase with repeat feeding (Fig 4 and Table 1). The opposite is true for the wild type line. The effect of repeat feeding is not large enough to overcome the effects of blocking as expressed as differences in EIP, ie, the wild type line still exhibits the greater R0 although there is a greater overlap of the three curves (less difference). While biting rate, death rate, and EIP were significant determinants of R0 for wild type (Fig 5) and Wolbachia-fertile mosquitoes (Fig 6), only the death rate and EIP mattered for Wolbachia-infertile mosquitoes (Fig 7). We expected greater differences to emerge between the Wolbachia fertile and infertile mosquitoes with multiple feeds (Fig 4), but this is likely due to the use of the same estimates for EIP and death rate for the two Wolbachia classes.

Fig 4.

R0 density distribution for single (A) and repeatedly blood-fed (B) experiments.

Fig 5. R0 values vs. experimental parameters for different blood-feeding experiments considering Wolbachia uninfected mosquitoes.

Green circles indicate the parameters contributing significantly.

Fig 6. R0 values vs. control parameters for different blood-feeding experiments considering Wolbachia-infected fertile mosquitoes.

Green circles indicate the parameters contributing significantly.

Fig 7. R0 values vs. control parameters for different blood-feeding experiments considering Wolbachia-infected infertile mosquitoes.

Green circles indicate the parameters contributing significantly.

Table 1. Statistical parameters for the R0 probability density distributions depicted in Fig 4 with EIP estimates of 5 and 7 days, respectively for ‘Wolbachia’ uninfected and ‘Wolbachia’-infected.

Highlighted in green is the population that transmits most effectively.


In the last decade, Wolbachia was released into Ae. aegypti field populations in several countries around the world to inhibit the transmission of arboviral diseases [1,2]. It was recently discovered Ae. aegypti females infected with the Wolbachia wAlbB strain can lack matured ovaries, become infertile, and show higher feeding rates than fertile Wolbachia-infected or wildtype uninfected females after experiencing an extended period of egg quiescence [25,26]. The increased feeding rate poses a potential challenge to the efficacy of Wolbachia as a biocontrol agent for controlling arboviral diseases. Here we conducted measurements of longevity, probing frequency, and quantified dengue virus in mosquito legs following a single or multiple infectious blood meal with DENV. Additionally, we modeled the effect of these measured parameters on virus transmission.

Our data show that adults infected with Wolbachia exhibit a slight reduction in survival compared to wild type after storage. We know, however, from previously published works in Ae. aegypti [3,4] that Wolbachia infection, even without egg storage, can cause similar scale reductions in lifespan to what was measured here. The reduction in survival is nearly linear, however, which is unusual and warrants future comparisons with Wolbachia-infected mosquitoes without a history of egg storage. Our detailed time course of feeding behavior in this study demonstrates how the probing behavior of Wolbachia-infected mosquitoes (dominated by the infertile phenotype) is increased. Importantly, the differences relative to the wild type are disproportionately due to behavior during the early dpi. After a single feed, wild type mosquitoes are not interested in probing again until at least day 4. The Wolbachia-infected individuals begin feeding almost immediately, at a rate roughly 10-fold higher. Even after dpi 4, the Wolbachia-infected population continued to probe at roughly twice the rate. These differences in behavior equate to both an increase in viral loads and viral prevalence in Wolbachia-infected infertile over fertile mosquitoes through increased blood meal consumption.

Our DENV prevalence data were not as expected because they lacked evidence of Wolbachia-mediated virus blocking. In one replicate, the wild type line exhibited the lowest rates of infection and in another replicate, it did not differ from either of the Wolbachia-infected lines. The viral load data, in contrast, exhibited the predicted pattern of blocking, with higher loads in the wild type line compared to the Wolbachia-infected line. Previous studies on pathogen blocking in disseminated proxy tissues (heads, wings, legs) have shown reduced prevalence of both DENV [3] and ZIKV [39] well beyond 8 dpi in the presence of wAlbB in contrast to our data. The lack of blocking could result from incomplete homogenization of the Wolbachia uninfected and infected lines, although at three rounds of backcrossing they should be >80% similar. Alternatively, these data suggest that long-term egg storage may affect Wolbachia’s blocking ability with respect to prevalence. Our viral load data, in contrast, did show strong evidence of blocking. We should note that the disconnect we see between prevalence and load viral load, while unusual [2], is not without precedent. A recent study from our group for Jamestown Canyon Virus indicated that these two traits can indeed be independent of one another [40]. In this latter case, Wolbachia was successful at preventing infection in individuals/tissues but was not effective at controlling viral load once infection was initiated. Here, we see the opposite, that infection of the tissues was very successful, but there was then subsequent moderation of viral loads.

Our modeling attempted to capture the additive effects of the increased biting rate due to Wolbachia-induced egg storage effects on transmission potential. We could not empirically measure the EIP in the critical early time points because we could not begin sampling until 8 dpi to allow for our repeated early feedings. We therefore used conservative estimates of EIP from the literature, with Wolbachia-infected lines exhibiting a later EIP. Repeat feeding, however, as seen in infertile females from stored eggs, might reduce the differential between wildtype and Wolbachia-infected fertile mosquitoes. As EIP is the dominant factor in transmission, given its power term in the Ross-MacDonald equation, any such reductions in EIP could substantially increase transmission. Additionally, other studies have shown that as viral loads rise, Wolbachia-mediated blocking is less effective [37]. Repeated feeding may reduce the efficacy of Wolbachia with greater exposure to virus. But while our modeling indicates Wolbachia-mediated blocking is more powerful than the infertility-induced repeat biting effects on load and prevalence, we do not think they have been fully captured in our design without a true empirical measure of EIP.


This study has raised several questions that should be explored in future empirical work. First, an expanded set of tissues, including salivary glands, should be assessed for viral prevalence and load in wAlbB-infected mosquitoes from stored eggs versus unstored to determine if blocking strength is indeed altered. Similarly, adult survival of stored and unstored lines should be compared. Second, we suggest that EIP for DENV be measured in the context of egg storage and Wolbachia wAlbB infection to get a better estimate of the relative importance of infertility on transmission. Third, previous work has suggested that infertility in response to egg storage is greater for the wAlbB than the wMel strain [25]. The effect of egg storage on virus transmissibility for wMel-infected mosquitoes should be measured and compared to our data. In terms of broader impact, our study also raises issues that need to be considered in field releases where there are long dry seasons that may force egg quiescence. The releases being developed in Saudi Arabia may provide an interesting test of the importance of these effects in the field [41]. If that quiescence is long enough and is experienced by a high proportion of individuals in the population, Wolbachia-infected individuals may be at a disadvantage reproductively, threatening spread. We found ~80% infertility after 11 weeks, but the previous study also found 25% at only 6 weeks [25]. Second, such storage may lead to greater biting in the resulting infertile females, which could increase transmission in the short term and reduce the efficacy of Wolbachia-mediated blocking. Last, any increased biting rates associated with Wolbachia infection may compromise community acceptance of its deployment. As the ideal scenario for successful Wolbachia replacement releases in the field includes both high female reproductive success and maximal reductions in virus transmission, further studies are needed to fully examine the relationships between storage length, infertility, and transmission and how they may affect the efficacy of Wolbachia over dryer landscapes.

Supporting information

S1 Fig. Viral infection following repeat feeding following 11 weeks of egg storage for Replicate 2.

(A) DENV-2 infection prevalence and (B) DENV-2 load in adult female Aedes aegypti legs at 8, 11, and 14 days post-infection (dpi). Females were provided with infectious blood every day from 1 to 7 dpi (post the initial infectious blood meal at 0 dpi). Each collection point x treatment is represented by 24 individuals. The fertility status of Wolbachia-infected females was identified through ovarian dissection.


S2 Fig. Wolbachia loads in mosquito bodies post-storage and a single feeding event did not differ (p = 0.057) between fertile and infertile individuals.



In memory of Professor Howie Weiss (Biology Department, The Pennsylvania State University) for his support of ARV and his enthusiasm for modeling host:parasite interactions.


  1. 1. Walker T, Johnson PH, Moreira LA, Iturbe-Ormaetxe I, Frentiu FD, McMeniman CJ, et al. The wMel Wolbachia strain blocks dengue and invades caged Aedes aegypti populations. Nature. 2011;476: 450–453.
  2. 2. Moreira LA, Iturbe-Ormaetxe I, Jeffery JA, Lu G, Pyke AT, Hedges LM, et al. A Wolbachia symbiont in Aedes aegypti limits infection with dengue, chikungunya, and Plasmodium. Cell. 2009;139: 1268–1278.
  3. 3. Xi Z, Khoo CC, Dobson SL. Wolbachia establishment and invasion in an Aedes aegypti laboratory population. Science (1979). 2005;310: 326–328.
  4. 4. Dutra HLC, Rocha MN, Dias FBS, Mansur SB, Caragata EP, Moreira LA. Wolbachia blocks currently circulating Zika virus isolates in Brazilian Aedes aegypti mosquitoes. Cell Host Microbe. 2016;19: 771–774.
  5. 5. Hurk AF, Hall-Mendelin S, Pyke AT, Frentiu FD, Mcelroy K. Impact of Wolbachia on infection with chikungunya and Yellow Fever viruses in the mosquito vector Aedes aegypti. PLOS Negl Trop Dis. 2012;6: 1892.
  6. 6. Sicard M, Bonneau M, Weill M. Wolbachia prevalence, diversity, and ability to induce cytoplasmic incompatibility in mosquitoes. Curr Opin Insect Sci. 2019;34: 12–20.
  7. 7. Dobson SL, Fox CW, Jiggins FM. The effect of Wolbachia-induced cytoplasmic incompatibility on host population size in natural and manipulated systems. Proc Biol Sci. 2002;269: 437–45.
  8. 8. Nazni WA, Hoffmann AA, NoorAfizah A, Cheong YL, Mancini M V, Golding N, et al. Establishment of Wolbachia strain wAlbB in Malaysian populations of Aedes aegypti for dengue control. Curr Biol. 2019;29: 4241–4248 e5.
  9. 9. Ryan PA, Turley AP, Wilson G, Hurst TP, Retzki K, Brown-Kenyon J, et al. Establishment of wMel Wolbachia in Aedes aegypti mosquitoes and reduction of local dengue transmission in Cairns and surrounding locations in northern Queensland, Australia. Gates Open Res. 2019;3: 1547.
  10. 10. Utarini A, Indriani C, Ahmad RA, Tantowijoyo W, Arguni E, Ansari MR, et al. Efficacy of Wolbachia-infected mosquito deployments for the control of dengue. N Engl J Med. 2021;384: 2177–2186.
  11. 11. Pinto SB, Riback TIS, Sylvestre G, Costa G, Peixoto J, Dias FBS, et al. Effectiveness of Wolbachia-infected mosquito deployments in reducing the incidence of dengue and other Aedes-borne diseases in Niteroi, Brazil: A quasi-experimental study. PLOS Negl Trop Dis. 2021;15: e0009556.
  12. 12. Hoffmann AA, Ahmad NW, Keong WM, Ling CY, Ahmad NA, Golding N, et al. Introduction of Aedes aegypti mosquitoes carrying wAlbB Wolbachia sharply decreases dengue incidence in disease hotspots. iScience. 2024;27: 108942.
  13. 13. Hoffmann AA, Ross PA, Rasic G. Wolbachia strains for disease control: ecological and evolutionary considerations. Evol Appl. 2015;8: 751–768.
  14. 14. López-Madrigal S, Duarte EH. Titer regulation in arthropod-Wolbachia symbioses. FEMS Microbiol Lett. 2019;366: fnz232.
  15. 15. Ross PA, Ritchie SA, Axford JK, Hoffmann AA. Loss of cytoplasmic incompatibility in Wolbachia-infected Aedes aegypti under field conditions. PLOS Negl Trop Dis. 20190419th ed. 2019;13: e0007357.
  16. 16. Zheng B, Guo W, Hu L, Huang M, Yu J. Complex Wolbachia infection dynamics in mosquitoes with imperfect maternal transmission. Math Biosci Eng. 2018;15: 523–541.
  17. 17. Foo IJ, Hoffmann AA, Ross PA. Cross-generational effects of heat stress on fitness and Wolbachia density in Aedes aegypti mosquitoes. Trop Med Infect Dis. 2019;4: 13.
  18. 18. Hien NT, Anh DD, Le NH, Yen NT, Phong T V, Nam VS, et al. Environmental factors influence the local establishment of Wolbachia in Aedes aegypti mosquitoes in two small communities in central Vietnam. Gates Open Res. 2021;5: 147.
  19. 19. Liang X, Tan CH, Sun Q, Zhang M, Wong PSJ, Li MI, et al. Wolbachia wAlbB remains stable in Aedes aegypti over 15 years but exhibits genetic background-dependent variation in virus blocking. PNAS nexus. 2022;1: pgac203.
  20. 20. Flores HA, Taneja de Bruyne J, O’Donnell TB, Tuyet Nhu V, Thi Giang N, Thi Xuan Trang H, et al. Multiple Wolbachia strains provide comparative levels of protection against dengue virus infection in Aedes aegypti. PLOS Pathog. 2020;16: e1008433.
  21. 21. Gesto JSM, Pinto SB, Dias FBS, Peixoto J, Costa G, Kutcher S, et al. Large-scale deployment and establishment of Wolbachia into the Aedes aegypti population in Rio de Janeiro, Brazil. Front Microbiol. 2021;12: 711107.
  22. 22. Ogunlade ST, Meehan MT, Adekunle AI, Rojas DP, Adegboye OA, McBryde ES. A Review: Aedes-borne arboviral infections, controls and Wolbachia-based strategies. Vaccines (Basel). 2021;9: 32.
  23. 23. Lau M, Ross PA, Endersby-Harshman NM, Hoffmann AA. Impacts of low temperatures on Wolbachia (Rickettsiales: Rickettsiaceae)-infected Aedes aegypti (Diptera: Culicidae). J Med Entomol. 2020;57: 1567–1574.
  24. 24. Ross PA, Wiwatanaratanabutr I, Axford JK, White VL, Endersby-Harshman NM, Hoffmann AA. Wolbachia infections in Aedes aegypti differ markedly in their response to cyclical heat stress. PLOS Pathog. 2017;13: e1006006.
  25. 25. Lau M-J, Ross PA, Hoffmann AA. Infertility and fecundity loss of Wolbachia-infected Aedes aegypti hatched from quiescent eggs is expected to alter invasion dynamics. PLOS Negl Trop Dis. 2021;15: e0009179.
  26. 26. Lau M-J, Ross PA, Endersby-Harshman NM, Yang Q, Hoffmann AA. Wolbachia inhibits ovarian formation and increases blood feeding rate in female Aedes aegypti. PLOS Negl Trop Dis. 2022;16: e0010913.
  27. 27. Smith DL, Battle KE, Hay SI, Barker CM, Scott TW, McKenzie FE. Ross, Macdonald, and a theory for the dynamics and control of mosquito-transmitted pathogens. PLOS Pathog. 2012;8: e1002588.
  28. 28. Ye YH, Carrasco AM, Frentiu FD, Chenoweth SF, Beebe NW, van den Hurk AF, et al. Wolbachia reduces the transmission potential of dengue-infected Aedes aegypti. PLOS Negl Trop Dis. 2015;9: e0003894.
  29. 29. Scott TW, Amerasinghe PH, Morrison AC, Lorenz LH, Clark GG, Strickman D, et al. Longitudinal studies of Aedes aegypti (Diptera: Culicidae) in Thailand and Puerto Rico: blood feeding frequency. J Med Entomol. 2000;37: 89–101.
  30. 30. Scott TW, Clark GG, Lorenz LH, Amerasinghe PH, Reiter P, Edman JD. Detection of multiple blood feeding in Aedes aegypti (Diptera: Culicidae) during a single gonotrophic cycle using a histologic technique. J Med Entomol. 1993;30: 94–9.
  31. 31. Ford SA, Allen SL, Ohm JR, Sigle LT, Sebastian A, Albert I, et al. Selection on Aedes aegypti alters Wolbachia-mediated dengue virus blocking and fitness. Nat Microbiol. 2019/08/28. 2019;4: 1832–1839.
  32. 32. Carrington LB, Simmons CP. Human to mosquito transmission of dengue viruses. Front Immunol. 2014;5: 290.
  33. 33. Gloria-Soria A, Brackney DE, Armstrong PM. Saliva collection via capillary method may underestimate arboviral transmission by mosquitoes. Parasit Vectors. 2022;15: 103.
  34. 34. Amuzu HE, McGraw EA. Wolbachia-based dengue virus inhibition is not tissue-specific in Aedes aegypti. PLOS Negl Trop Dis. 2016;10: e0005145.
  35. 35. Therneau TM, Lumley T, Atkinson E, Crowson C. Survival Analysis. 2023.
  36. 36. Kassmbara A, Kosinski M, Biecek P, Fabian S. Survminer. 2017.
  37. 37. Ferguson NM, Kien DT, Clapham H, Aguas R, Trung VT, Chau TN, et al. Modeling the impact on virus transmission of Wolbachia-mediated blocking of dengue virus infection of Aedes aegypti. Sci Transl Med. 2015;7: 279ra37.
  38. 38. Feinberg J, Langtangen HP. Chaospy: An open source tool for designing methods of uncertainty quantification. J Comput Sci. 2015;11: 46–57.
  39. 39. Chouin-Carneiro T, Ant TH, Herd C, Louis F, Failloux AB, Sinkins SP. Wolbachia strain wAlbA blocks Zika virus transmission in Aedes aegypti. Med Vet Entomol. 2020;34: 116–119.
  40. 40. Lau M-J, Dutra HLC, Jones MJ, McNulty BP, Diaz AM, Ware-Gilmore F, et al. Jamestown Canyon virus is transmissible by Aedes aegypti and is only moderately blocked by Wolbachia co-infection. PLOS Negl Trop Dis. 2023;17: e0011616.
  41. 41. Ross PA, Elfekih S, Collier S, Klein MJ, Lee SS, Dunn M, et al. Developing Wolbachia-based disease interventions for an extreme environment. PLOS Pathog. 2023;19: e1011117.