Science in Society Archive

Non-transgenic Mosquitoes to Combat Dengue

New research shows how a common symbiotic bacterium can stop the dengue virus multiplying in the mosquito host, making hazardous and inefficient transgenic mosquitoes obsolete Dr. Mae-Wan Ho

A team of researchers at several universities in Australia have made a breakthrough in controlling the spread of dengue fever. They infected the mosquito vector with a symbiotic bacterium that stops the dengue virus from multiplying in the mosquito host thereby effectively blocking the transmission of the virus from mosquitoes to humans.

This makes obsolete all attempts at controlling natural disease vectors through transgenic mosquitoes, which are by comparison ineffective, inefficient, costly, and hazardous to varying degrees (see [1] Transgenic Mosquitoes Not a Solution, SiS 54).

Wolbachia-infected mosquito protects against infection

The strategy adopted by the Australian research team led by Scott O’Neill at Monash University, Victoria, is simple.  The symbiotic bacteria living inside insect cells naturally induce resistance to pathogens, so infecting natural populations of insect vectors with such bacteria could be a way of controlling the spread of the disease. The bacteria belong to the genus Wolbachia, a widespread endosymbiont of many insects, and known to induce resistance to infections by viruses and other pathogens in the host (see Box).

Wolbachia a widespread endosymbiont

Wolbachia is a genus of bacteria that infect a wide range of arthropods and some nematodes. They are found in 17 to 76 % of surveyed arthropods and estimated to be present in 66 % of all arthropod species. This makes Wolbachia one of the most widespread intracellular bacteria, where it has evolved a symbiotic relationship with the host. The bacterium was first identified by Marshall Hertig and Burt Wolbach in the mosquito Culex pipiens in 1924 [2].  In 1971, Janice Yen and Ralph Barr at University of California Los Angeles discovered that Culex mosquito eggs were killed by cytoplasmic incompatibility when the sperm of Wolbachia-infected males fertilized infection-free egg. This cytoplasmic incompatibility is maternally inherited, and provides a driving mechanism for infection to spread rapidly through the population.

Wolbachia has been linked to viral resistance in Drosophila melanogaster and mosquitoes. Flies infected with the bacteria are more resistant to RNA vruses such as the Drosophila C virus, Nora virus, Flock House virus, cricket paralysis virus Chkungunya virus, and West Nile virus, as well as a Plasmodium that cause bird malaria  [3-5].

In a previous study [5], the team found that a laboratory strain of Wolbachia pipientis, wMelPop-CLA, isolated from Drosophila melanogaster and introduced into the dengue mosquito vector Aedes aegypti shortened the mosquito’s lifespan by 50 %, but also inhibited infection with dengue and Chikungunya viruses as well as the avian malaria parasite, Plasmodium gallinaceum.   

Population replacement as opposed to reduction

Shortening the lifespan of the insect vector host was initially deemed desirable as a means of reducing the potential for transmitting dengue by making the wild populations less long-lived, thereby effectively reducing the natural population.

However, the discovery that Wolbachia infection can make the mosquito vector resistant to the pathogen is more important than any effect on the host lifespan. In terms of population replacement - the option now favoured by most researchers - an infected host with normal lifespan would be more effective in controlling disease, as they will invade and infect natural populations more readily.

In the new study [6, 7], a different strain (wMel) isolated from Drosophila melanogaster that did not shorten host lifespan was introduced into A. aegypti. The infected mosquitoes were first tested in caged populations under semi-field conditions before they were released and successfully invaded two natural populations, reaching near-fixation in a few months.

Traditional control measures focus on reducing populations of the disease transmission vector, but these have largely failed to slow the current dengue pandemic, with more than 50 million estimated to be affected each year. Wolbachia pipientis can spread rapidly into uninfected host populations by inducing cytoplasmic incompatibility. This causes embryos from Wolbachia-uninfected females to die when they are mated with infected males, whereas the embryos of eggs from infected females are unaffected. Because Wolbachia is maternally inherited, this effect provides a transmission advantage for the symbiont, resulting in rapid invasion of insect host populations.

Creating new mosquito strains infected with endosymbiont Wolbachia

To facilitate the infection of A. aegypti, the wMel strain was first transferred from D. melanogaster embryos into a mosquito cell line and grown for approximately two years to allow the bacteria to adapt to the mosquito intracellular environment. Wolbachia bacteria purified from the cell line were then injected into A aegypti embryos, and stably infected mosquito lines were established in 2 to 8 generations. One of them, MGYP2, was selected for further experiments.

MGYP2 mosquitoes were grown in the presence of tetracycline, and after 8 generations, a ‘tetracycine cured’ line MGYP2tet was established that no longer has Wolbachia in it. An outbred line MGYP2OUT was created by backcrossing the MGYP2 line for three generations to the F1 progeny of wild-caught A. aegypti eggs from Cairns, Australia. The F1 progeny from wild-caught eggs were used as control, non-infected ‘wild-type’, to be compared with wMel-infected MGY2OUT mosquitoes.

The density of the wMel Wolbachia in all infected lines was about 10 bacteria to each mosquito cell, about 3-fold lower than mosquitoes infected with the life-shortening wMelPop-CLA. The tissue distribution of the bacteria in the adult female mosquitoes was visualized with fluorescence in situ hybridization, and shown to be widespread, with infection levels in ovaries and salivary glands similar to those infected with wMelPop-CLA. The heavy infection in ovaries would support high levels of maternal transmission, while the absence in saliva ensures that it will not be passed onto humans [8].  However, in contrast to the wMelPop-CLA infection, wMel is not present at high levels in Malphigian tubules and fat bodies, thoracic ganglia or brain tissue, which may account for the non-pathogenic nature of wMel.

Mass reciprocal crossing experiments between MGYP2 and MGYP2tet showed that the wMel bacteria induce strong cytoplasmic incompatibility in that no hatched eggs result from the crosses between infected males and uninfected females. In contrast, wMel-infected females mated to uninfected and wMel-infected males resulted in hatch rates of approximately 90 %.

Under semi-field conditions - consisting of cages that provide environments simulating the natural habitat of A. aegypti in north Queensland Australia - the fecundity and viability of wMel-infected MGYP2OUT were not different from wild-type mosquitoes. But wMelPop-CLA infected female mosquitoes showed drastically reduced fecundity, and its eggs were also much less viable.

The invasion potential of Walbachia in mosquito lines infected with wMel and wMelPop-CLA respectively were then tested in the same semi-field facility. Starting with a frequency of infected mosquitoes at 0.65, the wMel infection increased rapidly and reached fixation within 30 days in cage A and 80 days in cage B, while the wMelPop-CLA reached fixation after 40 days in cage A, but reached only 80 % replacement in cage B, possibly due to predation by two geckos found in cage B.

To test whether the infected mosquitoes were resistant to the dengue virus, a dengue virus infected blood meal was fed to females infected with wMel, and the levels of virus in whole bodies, legs and saliva were assessed after 14 days. The wMel mosquitoes had approximately 1  500-fold fewer virus compared with uninfected controls. Levels of virus in wMelPop-CLA infected mosquitoes were even lower, by 10  000-fold.  Infectious virus was present in 29 out of 36 saliva samples (80.2 %) from uninfected control mosquitoes, whereas only 2 out of 48 saliva samples from wMEL-infected mosquitoes contained low levels of virus. It turns out that the positive samples were from an individual that escaped infection through imperfect maternal transmission. Saliva from wMelPop-CLA infected mosquitoes were completely free of virus. Thus, there was practically complete blockage of dengue virus transmission by Wolbachia-infected mosquitoes. 

Deliberate release after extensive consultation at multiple sites and regulatory approval

A deliberate release of wMel-infected A. aepgypti was carried out in early January 2011 during the wet season. Adult male and female mosquitoes were released near Cairns in north-eastern Australia, a total of 141 600 at 184 sites in Yorkeys Knob and 157 300 at 190 sites in Gordonvale over a period of 10 to 11 weeks (one release per week). The open releases were approved by the Australian Pesticides and Veterinary Medicines Authority and were preceded by an extensive period of community engagement and subsequent strong community support. Following the initial release, Wobachia frequencies were monitored every 2 weeks and monitoring continued after releases were terminated.

The frequency of Wolbachia-infected A. aegypti increased rapidly to more than 80 to 95 % by the end of April.

O’Neill and colleagues point out that the Wolbachia-infected mosquito populations can act as nursery areas for future collection and dispersal of mosquitoes without having to rear additional Wolbachia-infected mosquitoes in special insectariums. Because the resistance to pathogens depends on a complex interaction between the mosquito host and endosymbiont bacteria [5], it is not easy for the pathogens to overcome the resistance by mutation or recombination.

The next step is to test whether the infected mosquitoes are effective for controlling dengue, and further ahead, whether other mosquito-borne diseases such as malaria could be similarly controlled.

One remaining question is whether Wolbachia itself can be passed onto human hosts, or become a pathogen for human beings. A Fact Sheet published by O’Neill’s group states [8]:

“In our research Wolbachia infected insects are feeding on our researchers all the time and there is no sign of any human illness associated with insect Wolbachia.

Wolbachia is an insect bacterium that has not been detected living inside humans or any other vertebrates. It can be made to infect human tissue culture cells in the laboratory but these laboratory systems are very artificial and do not predict the actual ability of Wolbachia to infect an actual human being.”

To conclude

The safest and most effective way to control the transmission of mosquito-borne infectious diseases such as dengue may be by infecting mosquito vectors with a common natural endosymbiont bacterium Wolbachia. The bacterium spreads readily by maternally inherited cytoplasmic incompatibility that kills the offspring of uninfected females fertilized by sperms from infected males. It also induces high levels of viral resistance in the mosquito host, thus blocking disease transmission. This approach leads to rapid replacement of natural populations of disease vectors with disease-resistant insects that remain stable in the wild, with minimum disturbance to the ecological niche, and without the need for continual replenishment with disease-resistant strains.  

It makes all attempts at engineering transgenic mosquitoes for reducing or replacing natural populations obsolete.       

Article first published 12/03/12


  1. Ho MW. Transgenic mosquitoes not a solution. Science in Society 54 (to appear) 2012.
  2. Wolbachia. Wikipedia, 28 January 2012,
  3. Teixeira L, Ferreira A and Ashburner M. The bacterial symbiont Wolbachia induces resistance to RNA viral infections in Drosophila melanogaster. PLoS Biol 2008, 6, e10000002.
  4. Hedges LM, Brownlie JC, O’Neill SL and Johnson KN. Wolbachia and virus protection in insects. Science 2008, 322, 702.
  5. Moreira LA, Iturbe-Ormaetxe I, Jeffery JA et al and O’Neill S. A Wolbachia symbion in Aedes aegypti limits infection with Dengue, Chikungunya, and Plasmodium. Cell 2009, 139, 1268-78.
  6. Walker T, Johnson H, Moreira LA, Iturbe-Ormaetze I, et al and Hoffmann AA. The wMel Wolbachia strain blocks dengue and invades caged Aedes aegypti populations. Nature 2011, 476, 450-54.
  7. Hoffmann AA, Montgomery BL, Popovici J et al and O’Neill SL. Successful establishment of Wolbachia in Aedes populations to suppress dengue transmission. Nature 2011, 476, 455-9.
  8. What is Wolbachia? Eliminate dengue our challenge, fact sheet 6(b), October 2010,

Got something to say about this page? Comment

Comment on this article

Comments may be published. All comments are moderated. Name and email details are required.

Email address:
Your comments:
Anti spam question:
How many legs on a cat?

There are 4 comments on this article so far. Add your comment above.

Rory Short Comment left 13th March 2012 01:01:43
The contents of this article are music to my soul. As creatures imbued with consciousness we should be working in cooperation with Nature, not against it, and this article gives an excellent example of doing so.

John Farion Comment left 14th March 2012 06:06:06
Cheers for this. Now let's work on bedbugs! John Farion

Todd Millions Comment left 21st March 2012 16:04:01
Message for John Farion-Bed bugs furthertherunto Mr Farion,The matter is well in hand-Ants.Find a local mound of predetory ants,and leave the clothes ,bedding ect you suspect harbouring bed bugs over it for a couple days.Old russian trick from the steppes.It also works well with politicians,clergy and other advertizers and bankers-if you stake em down firmly and leave for a week.Cheers

chewaik Comment left 15th December 2013 20:08:29
everyone is desperate to get rid of dengue burden soonest possible, how come this non-gm to combat dengue is so slow to implement in Asia????