Team:BGU Israel/Demonstrate
The Bacillus thuringiensis subsp. israelensis (Bti) bacteria, which expresses the pBtoxis, is one of the most common means of biocontrol of mosquitoes today. "Baktush", a Bti based reagent (comes from the Latin-Hebrew combination acronym:BACteria-yaTUSH, where Yatush is the Hebrew word for mosquito). Today, Bti is being used world-wide with a bacterium that was found in our Life Sciences Department at Ben-Gurion University of the Negev by the late Prof. Yoel Margalith. Bti produces insecticidal toxins and virulence factors that selectively target the larval stages of mosquitoes, black flies and fungus gnats[1]. This toxin is being spread in large water sources around the world in hope to destroy the larvae that infest the water sources. However, it is only partially effective as the toxin is very diluted in large bodies of water. Also, there is no treatment directed towards small water sources (an important source of mosquito infestations in urban areas), as mosquitoes can lay eggs even in an ashtray left outside or in a plant pot with a little bit of water left in it.
Our goal was to find a better delivery targeted system for biological pest control. We proposed the following question: Who knows best where mosquito eggs are being laid? The obvious answer was the female mosquito that lays the eggs. Hence, we set out with our ‘FlyGEM - The Trojan Mosquito’ project, where we aimed to genetically engineer a bacterium that will harbor a polycistronic plasmid that we would construct in our lab with the Bti toxin subunits. The bacterium will colonize the mosquito's midgut microbiome, and that mosquito will act as our guided missile and will lay infected eggs. These eggs that contain bacteria expressing the toxin subunits of Bti, which will be lethal only to the larvae and not the adult mosquito harboring the bacteria in her gut.
Below is an illustration mapping our plan.
We designed an experiment to see if feeding mosquitoes with Serratia containing a GFP marker will lead to the transferring of the bacteria from the female mosquito to her laid eggs (Figure 1). We relied on recent literature, where it was reported that the mosquito's gut microbiome transfers from the female mosquito to her offspring [3]. After feeding the mosquitoes with our engineered bacteria, we allowed a blood meal feeding and the laying of eggs. We made a bacterial culture from the eggs and incubated the culture overnight. We checked for a fluorescent signal of GFP using a fluorescent plate reader (Figure 2). Using a t test, a significant difference was viewed between the untreated eggs and the infected eggs.
We observed a strong fluorescent signal from eggs that were laid by infected female mosquitoes, compared to eggs laid by female mosquitoes that were untreated (that gave only a background fluorescence originating from mosquito eggs autofluorescence).
Figure 3: Confocal microscope images of eggs that were laid from control female mosquitoes that were not fed with engineered bacteria (both panels). It can be seen from these images that there is still autofluorescence from a source that is unknown to us in the untreated eggs. However, the treated eggs display a more significant fluorescence that manifested itself in the cultured bacteria originating from these eggs that were highly fluorescent and displayed resistance to the antibiotics, indicating the presence of the plasmid in the bacteria, that has propagated from the mother to its eggs and that was not present in the untreated control experiment eggs.
After the success of the experiment, we met with Dr. Philipos Papathanos from the Hebrew University who specializes in engineering insect biology. He noted that according to regulation laws in most countries, it is illegal to release female mosquitoes because they are the disease-causing agents. At this point, we understood that we needed to change our strategy for bacterial distribution (For more information, read our "Integrated Human Practices"). We designed a similar experiment to that of the vertical transfer experiment (transfer of bacteria from female mosquitoes to their offspring), only this time we separated males and females at the pupal stage, as Dr. Papathanos taught us, and let the males feed on bacteria, while the females were not treated with bacteria.
After feeding, we transferred the non-infected female mosquitoes to the male cages, divided in equal numbers. We let them mate for a day, and then feed on a blood meal and lay eggs. We made a bacterial culture from the eggs and let it incubate overnight. We checked for a fluorescence signal using a plate reader (Figure 5). Using a t test, a significant difference was viewed between the untreated eggs and the infected eggs.
We found a strong normalized fluorescence signal from the eggs that were laid by female mosquitoes who mated with infected male mosquitoes, compared to the eggs laid by female mosquitoes who mated with untreated male mosquitoes, that were fed on sucrose only.
We also looked at male mosquitoes under a confocal microscope (x10) to see if there is fluorescence in their gut. The results were fascinating.
We have found a better delivery system and established its efficiency!
Once we understood our designed plan was based on releasing infected male mosquitoes, we needed to answer the obvious question that came next: ‘How many infected male mosquitoes we would need to release in order to reduce or eliminate the female population in a defined area or location.’ We demonstrated that a single release will not be enough and will not affect dramatically the female population. Therefore, we changed the method of releasing male mosquitoes from a single release to a release in intervals at a steady release rate. The model was implemented in MATLAB, in a collaboration with the help of Technion Institute of Technology iGEM team. This model gave us an insight about the chance of success of our project, and the ability to achieve our goal.
In order to begin our plasmid construction, we first researched what are the most toxic Bti subunits to the Aedes aegypti mosquito[4]. Our plasmid was planned to be polycistronic, expressing 5 genes altogether (three toxins, one chaperone and one marker). It was planned on a pBEST plasmid backbone with a few design adjustments.
In order to create our final plasmid, we planned and constructed three initial "minimal" toxic plasmids using Gibson Assembly in our lab (Maps of those plasmids are shown in Figure 10). During the creation of these plasmids we added common tags (HA, Strep and HIS) to each of the subunits, in order to show later, in a Western blot analysis, the expression of the toxin subunits.
After we constructed our initial plasmids (including the subunits Cry4Ba, Cry11Aa, and Cry11Aa + p20), we needed to check the toxicity of the plasmids on live larvae to see if our constructs express well and are toxic to the larval stage of the Aedes aegypti mosquito.
We decided to separate between fifty to hundred, four-days-old larvae into different containers (triplicates for robustness) and infect them with our genetically engineered Serratia to check how they react and if they die. The results show that our plasmids indeed cause the larvae to die, some better than others. The results were confirmed in an ANOVA test, with a p value < 0.001. Reminder: we have not constructed our final plasmid yet. We predict that once we do, we would see a greater integrative effect.
When we first started the first toxicity assays, we noticed a mysterious phenomena. Every day we came to count the infected dead larvae, we counted fewer dead larvae than the day before. We did not understand how such a thing could occur, so we decided to test the problem at hand.
We placed a camera to see what happened and found that the live larvae eat the infected dead larvae! We looked at a larva that ate an infected dead larva under a confocal microscope and detected a fluorescent signal in its gut. We realized that we might be looking at an even better distribution than we thought. Not only will the larvae that will hatch from infected eggs get infected and die, but also larvae that hatched from non-infected eggs but were laid in the same small water source, can eat those infected dead larvae and die as well!
Conclusions:
* Delivery of Serratia bacteria from male to female mosquitoes and onto their eggs succeeded!! - We have found a BETTER delivery system that specifically targets the larvae of specific mosquitoes.
* Constructs of toxin plasmids were successfully prepared.
* Toxicity assays have proved that even a single toxin subunit can kill larvae.
* BONUS- bacterial propagation at the target point: IMPROVED DISTRIBUTION by live larvae feeding on infected dead larvae!
Future Plans:
* Patent submission- provisional patent application was submitted, PCT phase will follow.
* Manuscript will be prepared- we would like to share our success with the scientific community, upon several more experiments we intend to publish our study in a scientific journal.
* Construct of the final polycistronic plasmid- this is underway.
* Full transfer experiment of the polycistronic plasmid in bacteria: From males to females to eggs and then hatched larvae.
* Study of the mechanism of transfer- underway.
* Pilot field studies with local authorities.
References:
[1] United States Environmental Protection Agency.
[2] Gusmão, Desiely S., et al. "Culture-dependent and culture-independent characterization of microorganisms associated with Aedes aegypti (Diptera: Culicidae)(L.) and dynamics of bacterial colonization in the midgut." Acta tropica 115.3 (2010): 275-281.
[3] Boyer, Sebastien, et al. "Sexual performance of male mosquito Aedes albopictus." Medical and Veterinary Entomology 25.4 (2011): 454-459.
[4] Wang, Sibao, et al. "Driving mosquito refractoriness to Plasmodium falciparum with engineered symbiotic bacteria." Science 357.6358 (2017): 1399-1402.