Skip to main content
  • Loading metrics

Impact of Wolbachia on Infection with Chikungunya and Yellow Fever Viruses in the Mosquito Vector Aedes aegypti

  • Andrew F. van den Hurk ,

    Affiliation Public Health Virology, Communicable Diseases Unit, Queensland Health Forensic and Scientific Services, Coopers Plains, Australia

  • Sonja Hall-Mendelin,

    Affiliation Public Health Virology, Communicable Diseases Unit, Queensland Health Forensic and Scientific Services, Coopers Plains, Australia

  • Alyssa T. Pyke,

    Affiliation Public Health Virology, Communicable Diseases Unit, Queensland Health Forensic and Scientific Services, Coopers Plains, Australia

  • Francesca D. Frentiu,

    Affiliations School of Biological Sciences, Monash University, Clayton, Australia, Institute for Health and Biomedical Innovation, Queensland University of Technology, Kelvin Grove, Australia

  • Kate McElroy,

    Affiliation Department of Pathology, University of Texas Medical Branch, Galveston, Texas, United States of America

  • Andrew Day,

    Affiliation Public Health Virology, Communicable Diseases Unit, Queensland Health Forensic and Scientific Services, Coopers Plains, Australia

  • Stephen Higgs,

    Affiliation Department of Pathology, University of Texas Medical Branch, Galveston, Texas, United States of America

  • Scott L. O'Neill

    Affiliation School of Biological Sciences, Monash University, Clayton, Australia


Incidence of disease due to dengue (DENV), chikungunya (CHIKV) and yellow fever (YFV) viruses is increasing in many parts of the world. The viruses are primarily transmitted by Aedes aegypti, a highly domesticated mosquito species that is notoriously difficult to control. When transinfected into Ae. aegypti, the intracellular bacterium Wolbachia has recently been shown to inhibit replication of DENVs, CHIKV, malaria parasites and filarial nematodes, providing a potentially powerful biocontrol strategy for human pathogens. Because the extent of pathogen reduction can be influenced by the strain of bacterium, we examined whether the wMel strain of Wolbachia influenced CHIKV and YFV infection in Ae. aegypti. Following exposure to viremic blood meals, CHIKV infection and dissemination rates were significantly reduced in mosquitoes with the wMel strain of Wolbachia compared to Wolbachia-uninfected controls. However, similar rates of infection and dissemination were observed in wMel infected and non-infected Ae. aegypti when intrathoracic inoculation was used to deliver virus. YFV infection, dissemination and replication were similar in wMel-infected and control mosquitoes following intrathoracic inoculations. In contrast, mosquitoes with the wMelPop strain of Wolbachia showed at least a 104 times reduction in YFV RNA copies compared to controls. The extent of reduction in virus infection depended on Wolbachia strain, titer and strain of the virus, and mode of exposure. Although originally proposed for dengue biocontrol, our results indicate a Wolbachia-based strategy also holds considerable promise for YFV and CHIKV suppression.

Author Summary

Mosquito-transmitted viruses such as dengue, yellow fever and chikungunya, are responsible for significant morbidity and mortality throughout tropical and sub-tropical regions of the world. These viruses are primarily transmitted by Aedes aegypti, a mosquito that due to its close association with humans has historically been difficult to control. An innovative control strategy involving the release of mosquitoes infected with the intracellular bacterium Wolbachia is currently being developed. This approach is based on the recent discovery that Wolbachia reduces infection of mosquitoes with dengue virus, malaria parasites and filarial nematodes. In the current study, we demonstrated that Wolbachia also blocks infection of chikungunya and yellow fever viruses in Ae. aegypti. The degree of virus inhibition depended on the strain of Wolbachia, the route of virus exposure, the virus strain and the titer of virus that the mosquitoes were exposed to. The implementation of Wolbachia-based control strategies has the capacity to transform the way that mosquitotransmitted diseases are controlled in the future.


Mosquito-transmitted viruses cause significant human morbidity and mortality throughout the world and impose particularly heavy health and economic burdens on developing countries. Dengue, caused by infection with any of the four dengue virus (DENV) serotypes, is currently the leading arboviral disease, with millions of cases of classic dengue fever and tens of thousands of deaths annually due to hemorrhagic disease [1]. Yellow fever virus (YFV) has been implicated in an estimated 200,000 clinical cases and 30,000 human deaths annually in the equatorial regions of Africa and South America [2], [3]. Recently, chikungunya virus (CHIKV) emerged as a major threat, with unprecedented outbreaks on islands in the western Indian Ocean, as well as in India, Thailand, Malaysia and Italy [4]. Effective vaccines against all four DENV serotypes and CHIKV are still at various stages of development and clinical trial [5], [6]. Although a highly effective vaccine against YFV has been administered for over 50 years, rapid vaccination of susceptible populations either prior to or during an epidemic is financially and logistically challenging, particularly in developing countries [2], [3].

Current disease control measures focus on the suppression of mosquito vector populations to reduce virus transmission. The primary vector of DENVs, CHIKV and YFV is the mosquito Aedes aegypti, a highly domesticated species that feeds almost exclusively on humans. Its geographic range has expanded with increased urbanization, resulting in increased arbovirus transmission [7]. The primary mosquito control activities are source reduction to eliminate larval habitats, application of larvicides (such as Temephos or s-methoprene), or adulticiding with indoor residual spraying or ultra-low-volume application. While these approaches can be successful, they are often labor-intensive, can be prohibitively expensive to implement and require a sustained commitment from all levels of government [8]. Furthermore, increasing insecticide resistance and concerns with non-target effects on the environment have necessitated the development of alternative approaches to arbovirus control.

An emerging biocontrol approach to reducing transmission of arboviruses is provided by the transinfection of Ae. aegypti mosquitoes with Wolbachia pipientis from other insect hosts [9], [10]. Wolbachia is a maternally transmitted, endosymbiotic bacterium that manipulates host reproduction to enhance its own transmission [11]. Estimated to infect over 60% of insect species [12], Wolbachia can provide its host with nutritional benefits [13] and enhanced resistance to pathogens [14], [15]. Ae. aegypti transinfected with the wMelPop-CLA strain of Wolbachia from Drosophila melanogaster [16] displayed a shortened life span [16], a reduction in blood feeding success [17], [18] and dramatically lowered DENV serotype 2 (DENV-2) infection levels compared to Wolbachia-free control mosquitoes [19]. Although these phenotypes are likely to reduce virus transmission in the field, wMelPop-CLA may also impose fitness costs on Ae. aegypti, such as reduced fecundity due to poor blood feeding success [17], [18] and decreased embryonic viability [20].

A second strain of Wolbachia, wMel that is closely related to wMelPop-CLA and also occurs naturally in D. melanogaster was recently introduced into Ae. aegypti as an additional strain for the biocontrol of dengue [21]. Both strains induce cytoplasmic incompatibility and have high rates of maternal transmission [16], [21], phenotypes necessary for the invasion of Wolbachia in natural populations of mosquitoes. Unlike wMelPop-CLA infected mosquitoes, Ae. aegypti infected with wMel did not suffer any significant deleterious fitness costs when compared to uninfected controls [21]. Similar to wMelPop-infected Ae. aegypti, wMel-infected mosquitoes displayed dramatically reduced replication of DENV-2 [21]. Ae. aegypti infected with wMel were deployed in the field in north Queensland, Australia over the 2010–2011 summer [22]. Wolbachia was able to reach almost 100% fixation in wild mosquito populations only a few months following release, indicating that Wolbachia-infected mosquitoes present a very promising strategy for the control of dengue that is cost-effective and poses minimal environmental and social harm [23], [24].

Although originally developed as a biocontrol tool for DENVs, Ae. aegypti infected with wMelPop-CLA showed reduced infection with CHIKV [19], filarial nematodes [25] and avian malaria [19]. Therefore, wMelPop-CLA infected mosquitoes could potentially be used for biocontrol in areas where human pathogens other than DENVs occur. However, not all strains of Wolbachia protect equally well, with those phylogenetically most closely related to wMel and wMelPop conferring the greatest degree of protection [26]. In Drosophila, wMel confers protection against a range of viruses, for example Drosophila C virus, Flock House virus and Cricket paralysis virus [14], [15]. However, it remains unclear whether wMel infection is able to limit the replication of human pathogens other than DENVs in mosquitoes, information that is critical to evaluating different Wolbachia strains for biocontrol.

Here, we tested the ability of the wMel strain of Wolbachia to limit infection in Ae. aegypti with CHIKV and YFV. We also tested the ability of YFV to replicate in mosquitoes infected with wMelPop-CLA, in order to compare this strain to wMel. We found reduced levels of CHIKV but not YFV infection in wMel-infected Ae. aegypti, although the degree of virus inhibition depended on the mode of infection. By contrast, mosquitoes harboring the wMelPop-CLA strain of Wolbachia showed reduced infection rates with YFV with the extent of reduction virus strain and titer dependent.

Materials and Methods


Six different lines of Ae. aegypti were used in the experiments. The transinfection of Ae. aegypti with the wMelPop-CLA and wMel strains of Wolbachia and maintenance of infected lines has been previously described 16,21. The wMel-infected line, MGYP2, and its tetracycline-treated counterpart, MGYP2.tet, were exposed to both YFV and CHIKV. Both lines were in the F16 generation for the YFV experiments and the F24 generation for the CHIKV experiments. The original wMelPop-CLA infected line, PGYP1 [16], and the corresponding tetracycline-treated line, PGYP1.tet, were assessed in the YFV experiments only. The PGYP1 line was used in the combined F52 and F53 generations, while the PGYP1.tet was used in the combined F50 and F52 generations. A Wolbachia-uninfected wild-type line of Ae. aegypti, designated Cairns3, which originated from Cairns, Australia, was used as a positive control for all experiments. The YFV-susceptible Rex-D white-eye Higgs strain of Ae. aegypti that originated from Rexville, Puerto Rico, was used as an additional positive control for the YFV experiments [27], [28]. Both the Cairns3 and Rex-D lines had been in colony for >40 generations.

Virus strains

The CHIKV strain was isolated from a patient visiting Melbourne, Australia in 2006 and contained the alanine to valine mutation in the membrane fusion glycoprotein E1 gene (E1-A226V) that has been linked to increased infectivity in mosquitoes, especially Ae. albopictus [29], [30]. The CHIKV stock had previously been passaged three times in African green monkey kidney (Vero) cells prior to use in this study. Two YFV strains that have been characterized in mosquitoes were used for the experiments: BA-55 which was isolated from a fatal yellow fever case in Nigeria in 1987 [31] and Cinetrop 28 (OBS 7549), which was isolated from a yellow fever patient in Bolivia in 1999 (R.B. Tesh, University of Texas Medical Branch, personal communication). The BA-55 strain had been passaged three times in suckling mouse brains, whilst the Cinetrop 28 strain had been passaged twice in C6/36 (Ae. albopictus) cells.

Exposure of mosquitoes to YFV and CHIKV

The CHIKV experiments were undertaken in the Biosafety Level 3 (BSL-3) insectary at Queensland Health Forensic and Scientific Services, Brisbane, Australia, and the experiments with YFV were undertaken in the BSL-3 insectary located at the University of Texas Medical Branch, Galveston, Texas, USA. The MGYP2, MGYP2.tet and Cairns3 lines were exposed to CHIKV using both intrathoracic inoculation and oral feeding. Intrathoracic inoculation was employed because it circumvents the midgut infection and escape barriers, and allows a standard amount of virus to be delivered. The six Ae. aegypti lines were exposed to YFV via intrathoracic inoculation and in an infectious blood meal. However, low feeding rates in all lines coupled with unexpectedly low infection rates in control lines that did actually feed, compromised our ability to draw any meaningful conclusions from the YFV oral feeding data. Thus it was excluded from the current paper. For intrathoracic inoculation, immobilized mosquitoes were inoculated with 0.5 µl and 0.22 µl of YFV or CHIKV suspension, respectively. The suspension consisted of stock virus diluted in growth media (GM; Gibco-BRL, Gaithersburg, MD) which contained antibiotics and antimycotics and was supplemented with 10% fetal bovine serum (FBS) for YFV or 3% FBS for CHIKV. For oral feeding, mosquitoes were allowed to feed on a blood meal consisting of CHIKV diluted in commercially available defibrinated sheep blood (Institute of Medical and Veterinary Science, Adelaide, Australia) and 1% sucrose. The blood meal was housed within a membrane feeding apparatus that was fitted with pig intestine as the membrane [32]. Feeding was undertaken at 23°C and a 2 h period was used to ensure that a sufficient number of mosquitoes had imbibed a blood meal. Immediately following virus exposure, mosquitoes were anaesthetized with CO2 and engorged mosquitoes placed in 900 ml gauze-covered containers. All mosquitoes were maintained on 10% sucrose at 28°C, high relative humidity and 12L∶12D light cycle within an environmental growth cabinet.

Infection and dissemination

For the YFV experiments, mosquitoes were processed at either day 10 or 14 post-exposure. Mosquitoes were chilled and the heads removed from the bodies and placed separately in 0.5 ml of GM supplemented with 10% FBS, and antibiotics and antimycotics. Recovery of virus from the head demonstrated that the virus had infected head tissue and was potentially able to be transmitted.

For the CHIKV experiments, infection, dissemination and transmission were assessed using a modified in vitro capillary tube system [33]. Briefly, mosquitoes were anaesthetized with CO2 and the legs and wings removed. The saliva was collected by inserting the proboscis of the mosquito into a capillary tube containing GM with 20% FBS. After 30 min, the contents of the capillary tube were expelled into 600 µl of GM with 3% FBS. The salivary glands were then dissected from the body in a drop of GM with 3% FBS. The dissected salivary glands were washed in a drop of 1% bleach [34], before being rinsed in 2 drops of GM with 3% FBS. The body remnants and salivary glands were placed in separate 2 ml tubes containing 1 ml of GM with 3% FBS and 3 sterile glass beads.

Salivary glands were also excised for visualization of CHIKV infection via an immunofluorescence assay (IFA). The glands were dissected in a drop of PBS and washed in another two drops of PBS, before each gland was placed in separate wells of an 18 well microscope slide. Slides were air-dried, fixed in cold acetone for 30 min, before being stored at −80°C.

Virus assays

Titration of initial blood/virus mixtures.

Blood/virus mixtures were titrated as serial ten-fold dilutions in 96 well microtiter plates seeded with confluent monolayers of C6/36 cells. Plates were incubated at 28°C for 5 days or 7 days for YFV- or CHIKV- infected blood/virus mixtures, respectively, before being fixed in PBS/acetone and stored at −20°C. Infection with YFV was detected using an indirect immunofluorescence assay with the monoclonal antibody (Mab) 863 [35][37], whilst CHIKV infection was detected using a cell culture enzyme immunoassay (CC-EIA; [38]) and the alphavirus-reactive Mab, B10 (provided by Roy A. Hall, University of Queensland, Brisbane, Australia).

Detection of YFV and CHIKV in mosquito components and saliva expectorates.

Heads, salivary glands, bodies and body remnants were homogenized and clarified by centrifugation. The CC-EIA was used to detect CHIKV in all mosquito tissues and saliva expectorates. To further reduce bacterial and fungal contamination, the mosquito tissues and saliva expectorates were filtered through a 0.2 µm filter prior to being inoculated onto the C6/36 cell monolayers.

Salivary glands and saliva expectorates were also analyzed for CHIKV RNA using real time TaqMan RT-PCR. YFV viral RNA was detected in the heads and bodies using a YFV-specific TaqMan RT-PCR. Viral RNA was extracted from 140 µl of each of the mosquito components and the saliva expectorates using the QIAmp Viral RNA kit following the manufacturer's protocol (Qiagen, Clifton Hill, Australia), with modifications described in Moreira and others used for the YFV extractions [19]. The primer and probe sequences, preparation of synthetic controls, reaction mix components, and cycling conditions are the same as described previously [39], [40]. The CHIKV TaqMan RT-PCR was used to qualitatively detect viral RNA in the salivary glands and saliva samples, with a sample considered to be positive if the cycle threshold (Ct) value was <40. The number of YFV RNA copies in heads and bodies was quantified using the TaqMan RT-PCR, with RNA standards produced for the relative quantification of YFV RNA copy numbers normalized to RNA levels of the Ae. aegypti house-keeping gene RpS17 [19].

IFA analysis of CHIKV in Ae. aegypti salivary glands.

Fifteen microliters of B10 Mab was added to each well. After incubation at 37°C for 30 min in a humidified container, slides were rinsed in PBS pH 7.4 before being washed for 10 min in PBS pH 7.4. Slides were air-dried before the addition to each well of 15 µl Dako Anti-Mouse FITC Conjugate (Dako, Glostrup, Denmark) in 0.04% Evan's Blue in PBS pH 7.4. Slides were incubated and washed as described above, followed by application of two drops of Dako fluroescent mounting medium (containing NaN3; Dako, Glostrup, Denmark). A slide containing wells inoculated with Ross River virus, Barmah Forest virus, Sindbis virus and CHIKV infected cells was included as a positive control. Slides were examined under an Olympus CX41 fluorescent microscope (Olympus, Tokyo, Japan).

Statistical analysis

Infection, dissemination and transmission rates were analyzed using Fisher's Exact Tests with two-tailed P-values. YFV copy numbers were tested for differences among mosquito lines using Wilcoxon rank sum tests for non-parametric data and P-values <0.05 were considered statistically significant. P-values were adjusted for multiple tests using a Bonferroni correction. All analyses were performed in R [41].


wMel inhibits infection and dissemination of CHIKV in Ae. aegypti

We challenged wMel-infected (MGYP2 line) and uninfected mosquitoes (the tetracycline-treated MGYP2.tet and wild-type Cairns3 lines) with CHIKV in two replicate oral feeding experiments. In both experiments, fewer MGYP2 mosquitoes displayed CHIKV infection compared to MGYP2.tet and Cairns3 lines (Table 1). In experiment 1, virus was detected in significantly fewer MGYP2 (23%) mosquito bodies compared to MGYP2.tet (83%) and Cairns3 (58%) lines (P<0.01). Similarly, in experiment 2, CHIKV was detected in significantly fewer MGYP2 (30%) bodies compared to MGYP2.tet (70%) and Cairns3 (67%) lines (P<0.01).

Table 1. Effects of the wMel strain of Wolbachia (MGYP2 line) on CHIKV infection and dissemination in Ae. aegypti 12 days after virus exposure.

Next, we used both CC-EIA and a CHIKV-specific TaqMan RT-PCR assay to track dissemination of the virus to the salivary glands and saliva following oral feeding experiments. CHIKV was detected in significantly fewer MGYP2 (20%) salivary glands compared to those from MGYP2.tet (80%) and Cairns3 (66%) lines (P<0.01) in experiment 1, but only when using the TaqMan RT-PCR assay (Table 1, in bold). The TaqMan RT-PCR assay detected more CHIKV-positive saliva expectorates than the CC-EIA (Table 1), suggesting that, if infectious particles were present in the saliva, they were below the threshold of CC-EIA detection. Nonetheless, in both oral feeding experiments CHIKV was detected in significantly fewer saliva expectorates in MGYP2 (10% and 6%) than MGYP2.tet (58% and 47%) (in both experiments P<0.01), but not Cairns3 mosquitoes.

We tested whether wMel infection reduced CHIKV infection rates in Ae. aegypti using intrathoracic inoculations as the mode of virus delivery, in two separate experiments. Inoculations with CHIKV at titers of 104.5 and 106.3 TCID50/ml resulted in similar rates of virus infection among the three mosquito lines in bodies, salivary glands and saliva expectorates, using both CC-EIA and TaqMan RT-PCR for detection (Table 1, experiments 3 and 4). Rates of infection in bodies and salivary glands were consistently high across all three lines. A smaller percentage of saliva expectorates from MGYP2 mosquitoes were virus infected compared to MGYP2.tet and Cairns3 mosquitoes, but the difference was not statistically significant.

Finally, we visualized CHIKV infection in the salivary glands of MGYP2 and Cairns3 Ae. aegypti lines using immunofluorescence assays (IFA). Twelve days following oral exposure, 27% (3/11) and 67% (6/9) of the MGYP2 and Cairns3 lines, respectively, were positive in the IFA. There did not appear to be any consistency in the localization of virus within infected salivary glands from either Ae. aegypti line (Figure 1).

Figure 1. Chikungunya virus in the salivary glands of Ae. aegypti.

CHIKV antigen (green fluorescence) detected in the salivary gland from a wMel-infected mosquito (MGYP2) at 100× (A) and 400× (B) magnification, and from a Wolbachia-uninfected mosquito (Cairns3) at 100× (C) and 400× (D) magnification. Salivary glands negative for CHIKV antigen from MGYP2 (E) and Cairns3 (F) mosquitoes (Both glands are shown at 100× magnification).

Reduced YFV infection rates in wMelPop-CLA but not wMel-infected Ae. aegypti

We tested the effect of wMel and wMelPop-CLA (PGYP1 mosquito line) Wolbachia strains on body infection rates, using intrathoracic inoculation with two strains of YFV isolated from Nigeria (BA-55) and Bolivia (Cinetrop 28). Four Wolbachia-uninfected mosquito lines were used as controls: Cairns3, Rex-D, MGYP2.tet and PGYP1.tet. Due to high mortality rates in some lines, mosquitoes exposed to 103.5 TCID50/ml of YFV BA-55 were tested at 10 days post virus exposure. Mosquitoes exposed to104.5 TCID50/ml of YFV BA-55 and 104.0 TCID50/ml of YFV Cinetrop 28 were tested at 14 days post exposure.

At both low (103.5 TCID50/ml) and high (104.5 TCID50/ml) titers of YFV strain BA-55, MGYP2 mosquitoes had the same high (100%) rates of body infection as MGYP2.tet, Cairns3 and Rex-D mosquitoes (Table 2), detected using a YFV-specific Taqman assay. By contrast, at low titer, significantly fewer PGYP1 (12%) mosquitoes displayed body infections compared to the PGYP1.tet (100%), Cairns3 (100%) and Rex-D (100%) lines (P<0.001). PGYP1 mosquitoes also displayed significantly lower rates of body infection compared to MGYP2 mosquitoes (P<0.001). At the higher virus titer a similar number (100%) of PGYP1 mosquitoes was infected with YFV as for the PGYP1.tet, Cairns3 and Rex-D lines. A similar pattern was observed for the Cinetrop 28 strain of YFV, with mosquito body infection rates ranging from 96–100% and no significant differences among the mosquito lines, although this strain was tested at only one titer (104.0 TCID50/ml).

Table 2. Effects of the wMel and wMelPop-CLA (MGYP2 and PGYP1 lines, respectively) strains of Wolbachia on YFV infection and dissemination in Ae. aegypti.

Next, we assessed the impact of the two strains of Wolbachia on YFV dissemination to the head in Ae. aegypti. Using the YFV strain BA-55, MGYP2 mosquitoes displayed similar high rates of dissemination to the head as the MGYP2.tet, Cairns3 and Rex-D mosquitoes (92–100%), at both low and high virus titers (Table 2). At low virus titer, however, no dissemination to the head was observed in PGYP1 mosquitoes compared to control lines PGYP1.tet (100%), Cairns3 (100%), Rex-D (100%) and the MGYP2 (100%) (P<0.001). At high virus titer, a higher (78%) number of PGYP1 mosquitoes had virus in the heads and no significant differences in infection rates were found across the mosquito lines. High head dissemination rates (84–100%) were found across all mosquito lines challenged with the Cinetrop 28 strain of YFV, with no significant differences among them.

Reduced YFV replication in Wolbachia-infected Ae. aegypti is virus titer and strain dependent

Next, we explored the effect of Wolbachia on the number of YFV RNA copies in Ae. aegypti bodies following intrathoracic inoculation with the strains BA-55 and Cinetrop 28, at three different titers. Overall, infection with both Wolbachia strains resulted in a lower median number of virus copies per body in Ae. aegypti compared to control lines. However, PGYP1 mosquito bodies consistently displayed several log fewer copies than MGYP2 lines (Figure 2). Significantly fewer copies were found in MGYP2 compared to Wolbachia-uninfected MGYP2.tet, Cairns3 and Rex-D mosquitoes (P<0.05 in all Wilcoxon rank sum tests) in challenges with YFV strain BA-55 at a titer of 103.5 TCID50/ml. However, virus levels were high in both MGYP2 and the Wolbachia-uninfected lines (medians >108 copies/body; Figure 2A). By contrast, median virus copy in PGYP1 mosquitoes was zero, an 8-log difference compared to MGYP2 (P<0.0001, Wilcoxon rank-sum test) and significantly less than PGYP1.tet, Cairns3 and Rex-D mosquitoes (P<0.0001 in all Wilcoxon rank-sum tests) (Figure 2A).

Figure 2. Effect of two Wolbachia strains on yellow fever virus replication in mosquito bodies.

Viral RNA copies assayed by TaqMan RT-PCR in wMel-infected (MGYP2), wMelPop-CLA infected (PGYP1) and Wolbachia-uninfected (MGYP2.tet, PGYP1.tet, Cairns3 and Rex-D) mosquitoes exposed to 103.5 TCID50/ml (A) and 104.5 TCID50/ml (B) of African (BA-55) and 104.0 TCID50/ml (C) of South American (Cinetrop 28) YFV strains. Box plots show median RNA copy numbers and 25 (below median) and 75 (above median) percentiles, with outer bars at the lowest and highest observations (lower outer bars not shown when data overlaps zero). Number of mosquitoes assayed per line shown above box plots. #Denotes a median of zero. *P<0.01, **P<0.001 (Wilcoxon rank sum tests). For clarity, we only denote significant differences in comparisons between Wolbachia-infected lines and their corresponding tetracycline treated counterparts or between MGYP2 and PGYP1.

In challenges with YFV strain BA-55, at the higher titer of 104.5 TCID50/ml, MGYP2 bodies still had significantly fewer virus copies than Rex-D (P<0.001, Wilcoxon rank-sum test) but not the MGYP2.tet or Cairns3 lines (Figure 2B). However, PGYP1 bodies had far less virus than either the MGYP2 or Wolbachia-uninfected lines (P<0.0001 in all Wilcoxon rank-sum tests) (Figure 2B). YFV replicated to a significantly lesser extent in MGYP2 and PGYP1 bodies following challenge with the Cinetrop 28 strain compared to control lines (P<0.0001 in all Wilcoxon rank-sum tests) (Figure 2C). YFV copy numbers were almost 5-logs lower in PGYP1 than MGYP2 bodies (P<0.0001, Wilcoxon rank-sum test) (Figure 2C).

We also explored whether Wolbachia strain and YFV inoculation titer influenced the extent to which the virus replicated in Ae. aegypti heads, using TaqMan RT-PCR quantification of viral RNA (Figure 3). In challenges with YFV strain BA-55 at the lower titer of 103.5 TCID50/ml, significantly fewer RNA copies were detected in MGYP2 heads compared to Wolbachia-uninfected lines (P<0.001 in all Wilcoxon rank-sum tests), although the number of virus copies was still high (Figure 3A). By contrast, no virus was detected in the heads of PGYP1 mosquitoes. In challenges with YFV strain BA-55 at the higher titer of 104.5 TCID50/ml, similar levels of virus disseminated to the head in MGYP2 as in Cairns3 and Rex-D, but not MGYP2.tet mosquitoes (P<0.01, Wilcoxon rank-sum tests) (Figure 3B). Median virus copy number for PGYP1 heads was six logs lower than the median for MGYP2 lines (P<0.0001, Wilcoxon rank-sum test), although virus was present in the majority of PGYP1 heads (Figure 3B). Significantly less virus had disseminated to MGYP2 and PGYP1 heads compared to control, Wolbachia-uninfected lines (P<0.0001 in all Wilcoxon rank-sum tests) when mosquitoes were challenged with the Cinetrop 28 strain (Figure 3C). Interestingly, a higher amount of YFV was found in PGYP1 heads in challenges with the Cinetrop 28 strain (median = 104.8 copies, Figure 3C) compared to BA-55 at either low (median = 0 copies; Figure 3A) or high (median = 103.1 copies; Figure 3B) inoculation titers, suggesting that the extent of YFV replication in the presence of Wolbachia may depend on virus strain.

Figure 3. Effect of two Wolbachia strains on yellow fever virus replication in mosquito heads.

Viral RNA copies assayed by TaqMan RT-PCR in wMel-infected (MGYP2), wMelPop-CLA infected (PGYP1) and Wolbachia-uninfected (MGYP2.tet, PGYP1.tet, Cairns3 and Rex-D) mosquitoes exposed to 103.5 TCID50/ml (A) and 104.5 TCID50/ml (B) of African (BA-55) and 104.0 TCID50/ml (C) of South American (Cinetrop 28) YFV strains. Box plots show median RNA copy numbers and 25 (below median) and 75 (above median) percentiles, with outer bars at the lowest and highest observations (lower outer bars not shown when data overlaps zero). Number of mosquitoes assayed per line shown above box plots. # Denotes a median of zero. *P<0.01, **P<0.001 (Wilcoxon rank sum tests). For clarity, we only denote significant differences in comparisons between Wolbachia-infected lines and their corresponding tetracycline treated counterparts or between MGYP2 and PGYP1.


Mosquito-transmitted viruses such as DENVs, CHIKV and YFV pose a significant health risk for almost half of the world's population. The release of Wolbachia-infected mosquitoes into natural populations has been proposed as a way of reducing DENV transmission while minimizing social and environmental harm [21], [22]. Although developed for dengue biocontrol, Wolbachia-infected mosquitoes may prevent the transmission of other significant arboviruses. Here, we evaluated whether mosquitoes infected with the wMel strain of Wolbachia show limited infection with CHIKV and YFV. To compare the two Wolbachia strains, we tested whether YFV was able to infect and replicate in wMelPop-CLA infected mosquitoes.

Mosquitoes infected with wMel showed significantly reduced rates of CHIKV infection and dissemination to the salivary glands compared to controls, but only in the oral exposure experiments. CHIKV also showed limited dissemination in wMelPop-CLA-infected mosquitoes following oral exposure [19], suggesting that both strains of Wolbachia may be useful candidates for release in CHIKV control programs. By contrast, YFV was much less likely to infect and disseminate in Ae. aegypti infected with wMelPop-CLA compared to wMel strains. The virus was also less likely to replicate in wMelPop-CLA infected mosquitoes, with very high virus loads detected in wMel-infected Ae. aegypti. Our experiments suggest that wMelPop-CLA infected mosquitoes may be the best candidates for YFV biocontrol programs, but were unable to determine the extent of virus replication following oral exposure rather than intrathoracic inoculation. Because virus inhibition with some Wolbachia-virus combinations does not appear to be complete, it is essential that epidemiological models be utilized to establish the threshold of virus inhibition necessary to minimize and prevent transmission in the field.

The wMel strain occurs at high levels in Ae. aegypti ovaries and salivary glands but it is not present in as many body tissues as the wMelPop-CLA strain [19], [21]. In particular, wMel does not accumulate to high levels in the fat bodies, brain and thoracic ganglia, which are involved in secondary replication of arboviruses prior to infection of the salivary glands [42]. YFV may have replicated to a sufficiently high level in these tissues in wMel-infected mosquitoes to allow dissemination to the salivary glands and potentially be transmitted to humans. The presence of wMel in the mosquito did not completely prevent the replication of CHIKV in Ae. aegypti salivary glands, as determined by cell-culture, real-time RT-PCR and IFA analysis.

Host immune pre-activation has been proposed as an explanation for interference with virus replication because Wolbachia-infected Ae. aegypti show significant immune upregulation [19], [25], [43]. Recently, Pan et al. showed that Wolbachia infection activates the Toll immune pathway in Ae. aegypti and the production of antimicrobial peptides which inhibit DENV-2 [44]. However, Wolbachia infection in Drosophila, the endosymbiont's native host, also blocks DENV-2 replication despite the absence of immune upregulation [45]. Therefore, although immune upregulation may explain some of the observed Wolbachia-mediated interference with virus replication, it may not be the whole explanation [45]. Alternatively, Wolbachia may compete with arboviruses for key cellular components [19], [46]. Fluorescence in situ hybridization has revealed that wMelPop-CLA infection is widespread throughout the body of infected Ae. aegypti, and is prevalent in tissues and organs associated with viral replication, including the fat bodies, brain and ommatidia [19]. Importantly, DENV-2 antigen did not co-localize in cells and tissues that were infected with wMelPop-CLA. As DENV-2 and YFV are closely related flaviviruses, it is possible that a similar pattern of virus exclusion is occurring with YFV in wMelPop-CLA infected cells and tissues. In vitro studies have shown that reduced virus replication correlates with the density of Wolbachia in the cell [46]. Further studies are required to determine the precise molecular mechanism underpinning Wolbachia-mediated interference with virus replication.

The discovery that Wolbachia limits arbovirus infection and replication in Ae. aegypti has the potential to fundamentally alter vector control strategies [23]. Importantly, mass release of wMel-infected Ae. aegypti in two communities near Cairns, northern Australia, resulted in >90% frequency within the population six weeks after the first release [22]. Although the Wolbachia-based control strategy was initially proposed to control DENVs, previous work and the current study have demonstrated that releases of Wolbachia-infected Ae. aegypti to control DENVs may have the added benefit of reducing CHIKV and YFV transmission. Alternatively, Wolbachia-infected Ae. aegypti could be released to specifically reduce transmission of these viruses, irrespective of whether DENVs were circulating or not in an area.

We have demonstrated that the level of virus blockage induced by Wolbachia is dependent on the strain of Wolbachia, the strain of virus and the mode of exposure to the virus. Higher YFV intrathoracic inoculation titers allowed more breakthrough of virus in wMelPop-CLA infected mosquitoes. Infection rates, accumulation and dissemination to the head also differed between BA-55 and Cinetrop 28 in the presence of wMelPop-CLA, suggesting that different virus strains may have different replication dynamics in Wolbachia-infected mosquitoes. The CHIKV experiments showed that the TaqMan RT-PCR was generally more sensitive than the CC-EIA in detecting the presence of virus. This was not unexpected, given that the TaqMan RT-PCR does not differentiate between RNA derived from live, packaged infectious virus particles, from that of dead or defective, non-infectious virus. Thus, our experiments illustrate the importance of carefully selecting virus titer, virus strain, mode of delivery and method of virus detection when assessing the potential utility of different Wolbachia strains for biocontrol.

As part of the implementation of Wolbachia-based control strategies, it will be essential to assess the virus blocking ability of different Wolbachia strains against different virus strains, and if necessary, identify additional Wolbachia strains that could be deployed in the field. Commensurate with this will be the need for ongoing monitoring of Wolbachia-infected Ae. aegypti populations to confirm that the virus-limiting phenotype is maintained in successive generations post release.


The authors wish to thank Nicole Hausser for technical support, and Nichola Kenny and Jing Huang for assistance rearing the mosquitoes. We thank Julian Druce for providing the CHIKV isolate and Roy Hall for the provision of the monoclonal antibody, B10. We also thank Robert Tesh for providing the YFV isolates.

Author Contributions

Conceived and designed the experiments: AFVDH SH SLO. Performed the experiments: AFVDH SH-M KM AD SH. Analyzed the data: FDF. Contributed reagents/materials/analysis tools: ATP. Wrote the paper: AFVDH FDF SLO.


  1. 1. Guzman MG, Halstead SB, Artsob H, Buchy P, Farrar J, et al. (2010) Dengue: a continuing global threat. Nat Rev Microbiol 8: S7–16.
  2. 2. Jentes ES, Poumerol G, Gershman MD, Hill DR, Lemarchand J, et al. (2011) The revised global yellow fever risk map and recommendations for vaccination, 2010: consensus of the Informal WHO Working Group on Geographic Risk for Yellow Fever. Lancet Infect Dis 11: 622–632.
  3. 3. Monath TP (2001) Yellow fever: an update. Lancet Infect Dis 1: 11–20.
  4. 4. Staples JE, Breiman RF, Powers AM (2009) Chikungunya fever: an epidemiological review of a re-emerging infectious disease. Clin Infect Dis 49: 942–948.
  5. 5. Akahata W, Yang ZY, Andersen H, Sun S, Holdaway HA, et al. (2010) A virus-like particle vaccine for epidemic Chikungunya virus protects nonhuman primates against infection. Nat Med 16: 334–338.
  6. 6. Schmitz J, Roehrig J, Barrett A, Hombach J (2011) Next generation dengue vaccines: a review of candidates in preclinical development. Vaccine 29: 7276–7284.
  7. 7. Kyle JL, Harris E (2008) Global spread and persistence of dengue. Annu Rev Microbiol 62: 71–92.
  8. 8. Morrison AC, Zielinski-Gutierrez E, Scott TW, Rosenberg R (2008) Defining challenges and proposing solutions for control of the virus vector Aedes aegypti. PLoS Med 5: e68.
  9. 9. Brownstein JS, Hett E, O'Neill SL (2003) The potential of virulent Wolbachia to modulate disease transmission by insects. J Invertebr Pathol 84: 24–29.
  10. 10. Sinkins SP, O'Neill SL (2000) Wolbachia as a Vehicle to Modify Insect Populations. Boca Raton, Fl: CRC Press.
  11. 11. Werren JH, Baldo L, Clark ME (2008) Wolbachia: master manipulators of invertebrate biology. Nat Rev Microbiol 6: 741–751.
  12. 12. Hilgenboecker K, Hammerstein P, Schlattmann P, Telschow A, Werren JH (2008) How many species are infected with Wolbachia? - A statistical analysis of current data. FEMS Microbiol Lett 281: 215–220.
  13. 13. Brownlie JC, Cass BN, Riegler M, Witsenburg JJ, Iturbe-Ormaetxe I, et al. (2009) Evidence for metabolic provisioning by a common invertebrate endosymbiont, Wolbachia pipientis, during periods of nutritional stress. PLoS Pathog 5: e1000368.
  14. 14. Hedges LM, Brownlie JC, O'Neill SL, Johnson KN (2008) Wolbachia and virus protection in insects. Science 322: 702.
  15. 15. Teixeira L, Ferreira A, Ashburner M (2008) The bacterial symbiont Wolbachia induces resistance to RNA viral infections in Drosophila melanogaster. PLoS Biol 6: e2.
  16. 16. McMeniman CJ, Lane RV, Cass BN, Fong AW, Sidhu M, et al. (2009) Stable introduction of a life-shortening Wolbachia infection into the mosquito Aedes aegypti. Science 323: 141–144.
  17. 17. Moreira LA, Saig E, Turley AP, Ribeiro JM, O'Neill SL, et al. (2009) Human probing behavior of Aedes aegypti when infected with a life-shortening strain of Wolbachia. PLoS Negl Trop Dis 3: e568.
  18. 18. Turley AP, Moreira LA, O'Neill SL, McGraw EA (2009) Wolbachia infection reduces blood-feeding success in the dengue fever mosquito, Aedes aegypti. PLoS Negl Trop Dis 3: e516.
  19. 19. Moreira LA, Iturbe-Ormaetxe I, Jeffery JA, Lu G, Pyke AT, et al. (2009) A Wolbachia symbiont in Aedes aegypti limits infection with dengue, Chikungunya, and Plasmodium. Cell 139: 1268–1278.
  20. 20. McMeniman CJ, O'Neill SL (2010) A virulent Wolbachia infection decreases the viability of the dengue vector Aedes aegypti during periods of embryonic quiescence. PLoS Negl Trop Dis 4: e748.
  21. 21. Walker T, Johnson PH, Moreira LA, Iturbe-Ormaetxe I, Frentiu FD, et al. (2011) The wMel Wolbachia strain blocks dengue and invades caged Aedes aegypti populations. Nature 476: 450–453.
  22. 22. Hoffmann AA, Montgomery BL, Popovici J, Iturbe-Ormaetxe I, Johnson PH, et al. (2011) Successful establishment of Wolbachia in Aedes populations to suppress dengue transmission. Nature 476: 454–457.
  23. 23. Iturbe-Ormaetxe I, Walker T, O'Neill S L (2011) Wolbachia and the biological control of mosquito-borne disease. EMBO Reports 12: 508–518.
  24. 24. Popovici J, Moreira LA, Poinsignon A, Iturbe-Ormaetxe I, McNaughton D, et al. (2010) Assessing key safety concerns of a Wolbachia-based strategy to control dengue transmission by Aedes mosquitoes. Mem Inst Oswaldo Cruz 105: 957–964.
  25. 25. Kambris Z, Cook PE, Phuc HK, Sinkins SP (2009) Immune activation by life-shortening Wolbachia and reduced filarial competence in mosquitoes. Science 326: 134–136.
  26. 26. Osborne SE, Leong YS, O'Neill SL, Johnson KN (2009) Variation in antiviral protection mediated by different Wolbachia strains in Drosophila simulans. PLoS Pathog 5: e1000656.
  27. 27. Miller BR, Mitchell CJ (1991) Genetic selection of a flavivirus-refractory strain of the yellow fever mosquito Aedes aegypti. Am J Trop Med Hyg 45: 399–407.
  28. 28. Wendell MD, Wilson TG, Higgs S, Black WC (2000) Chemical and gamma-ray mutagenesis of the white gene in Aedes aegypti. Insect Mol Biol 9: 119–125.
  29. 29. Druce JD, Johnson DF, Tran T, Richards MJ, Birch CJ (2007) Chikungunya virus infection in traveler to Australia. Emerg Infect Dis 13: 509–510.
  30. 30. Tsetsarkin KA, Vanlandingham DL, McGee CE, Higgs S (2007) A single mutation in chikungunya virus affects vector specificity and epidemic potential. PLoS Pathog 3: e201.
  31. 31. Miller BR, Monath TP, Tabachnick WJ, Ezike VI (1989) Epidemic yellow fever caused by an incompetent mosquito vector. Trop Med Parasitol 40: 396–399.
  32. 32. Rutledge LC, Ward RA, Gould DJ (1964) Studies on the feeding response of mosquitoes to nutritive solutions in a new membrane feeder. Mosq News 24.
  33. 33. Aitken THG (1977) An in vitro feeding technique for artificially demonstrating virus transmission by mosquitoes. Mosq News 37: 130–133.
  34. 34. Jennings CD, Kay BH (1999) Dissemination barriers to Ross River virus in Aedes vigilax and the effects of larval nutrition on their expression. Med Vet Entomol 13: 431–438.
  35. 35. Gould EA, Buckley A, Cammack N, Barrett AD, Clegg JC, et al. (1985) Examination of the immunological relationships between flaviviruses using yellow fever virus monoclonal antibodies. J Gen Virol 66: 1369–1382.
  36. 36. McElroy KL, Tsetsarkin KA, Vanlandingham DL, Higgs S (2005) Characterization of an infectious clone of the wild-type yellow fever virus Asibi strain that is able to infect and disseminate in mosquitoes. J Gen Virol 86: 1747–1751.
  37. 37. McGee CE, Tsetsarkin K, Vanlandingham DL, McElroy KL, Lang J, et al. (2008) Substitution of wild-type yellow fever Asibi sequences for 17D vaccine sequences in ChimeriVax-dengue 4 does not enhance infection of Aedes aegypti mosquitoes. J Infect Dis 197: 686–692.
  38. 38. Broom AK, Hall RA, Johansen CA, Oliveira N, Howard MA, et al. (1998) Identification of Australian arboviruses in inoculated cell cultures using monoclonal antibodies in ELISA. Pathology 30: 286–288.
  39. 39. van den Hurk AF, Hall-Mendelin S, Pyke AT, Smith GA, Mackenzie JS (2010) Vector competence of Australian mosquitoes for chikungunya virus. Vector-borne and Zoonotic Dis 10: 489–495.
  40. 40. van den Hurk AF, McElroy K, Pyke AT, McGee CE, Hall-Mendelin S, et al. (2011) Vector competence of Australian mosquitoes for yellow fever virus. Am J Trop Med Hyg 85: 446–451.
  41. 41. R Development Core Team (2008) R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing.
  42. 42. Black IV WC, Bennett KE, Gorrochótegui-Escalante N, Barillas-Mury CV, Fernández-Salas I, et al. (2002) Flavivirus susceptibility in Aedes aegypti. Arch Med Res 33: 379–388.
  43. 43. Kambris Z, Blagborough AM, Pinto SB, Blagrove MS, Godfray HC, et al. (2010) Wolbachia stimulates immune gene expression and inhibits Plasmodium development in Anopheles gambiae. PLoS Pathog 6: e1001143.
  44. 44. Pan X, Zhou G, Wu J, Bian G, Lu P, et al. (2011) Wolbachia induces reactive oxygen species (ROS)-dependent activation of the Toll pathway to control dengue virus in the mosquito Aedes aegypti. Proc Natl Acad Sci U S A 109: E23–31.
  45. 45. Rances E, Ye YH, Woolfit M, McGraw EA, O'Neill SL (2012) The relative importance of innate immune priming in Wolbachia-mediated dengue interference. PLoS Pathog 8: e1002548.
  46. 46. Frentiu FD, Robinson J, Young PR, McGraw EA, O'Neill SL (2010) Wolbachia-mediated resistance to dengue virus infection and death at the cellular level. PLoS One 5: e13398.