This report is the first comprehensive study on AQP genes and their expression in the adult yellow fever mosquito, Aedes aegypti. We specifically focused on AQPs expressed in the MTs, an insect organ, specialized in water and waste excretion.
An interesting result of our phylogenetic analysis (see ) is that we were unable to identify any typical aquaglyceroporin in dipteran insects. Human aquaglyceroporins, the Homo sapiens
AQPs 3, 7, 9, and 10, form a separate clade with six Plasmodium
, one Leishmania
, one fungal and one louse AQP. A multiple sequence alignment of all AQPs used for our analysis (Figure S1
) revealed that all aquaglyceroporins share a cysteine residue at position six, C-terminal to the NPA motif in the B-loop. None of the dipteran AQPs in our analysis had a cysteine at this position. The fact that dipteran insects do not appear to possess an AQP-like glycerol transporter raises the question as to how they transport glycerol over cell membranes. Glycerol plays an important role in insect cold tolerance and diapause and is found in high concentration in the hemolymph of diapausing insects 
The microarray expression data was created with RNA isolated from total mosquitoes () and shows that AQP expression was generally down regulated after a blood meal, even at 3 h PBM. After taking a blood meal mosquitoes seek out a resting place to digest the blood and perform vitellogenesis and egg development. This resting phase extends over a period of approximately two days. During this time mosquitoes don't take up water, therefore high AQP expression levels are not necessary.
AQP expression show specific patterns in distinct organs/body parts (). All examined organs/body parts differ in their AQP expression patterns. We were specifically interested in patterns from midgut and MTs since blood meal-derived water has to cross the midgut epithelium and subsequently the MT epithelium for excretion. AQPs 1, 2, 4, and 5 were expressed in the midgut. Interestingly, three AQPs expressed in the midgut were down regulated at the 3 h PBM time point, corresponding to the time at which the bulk of blood meal-derived water has already been excreted. Our data suggests that AQPs 1, 4, and 5 are the principal AQPs in the MTs of adult females.
We developed an in vivo
diuresis assay for adult yellow fever mosquito females based on PBS injection. PBS is a non-toxic isotonic buffer containing sodium chloride, sodium phosphate, potassium chloride and potassium phosphate. This novel assay has several advantages over classical blood meal-based assays. Firstly, mosquitoes receive a standardized amount of PBS, while blood meal sizes can vary significantly. Secondly, since Ae. aegypti
mosquitoes start secreting urine about 30 s after starting a blood meal 
, the determination of blood meal sizes is difficult. Applying our method, mosquitoes injected with PBS started secreting urine after 2 min, which allowed accurate measurement of their weight after injection. As mentioned above, mosquito diuresis is controlled by neuropeptide hormones that are secreted from the central nervous system in response to unknown stimuli associated with blood ingestion. PBS injection triggers this response efficiently. The rapid onset of diuresis after injection suggests that AQP activity in the MT is regulated by trafficking of AQP-containing vesicles to the plasma membrane, analogous to the processes described for the renal collection duct in human kidney 
RNAi is a powerful tool used to study gene function in mosquitoes. We successfully employed RNAi-mediated knockdown for the analysis of AQP function in the MT. Knockdown rates, as determined by real time PCR, were between 60 and 95% and therefore well inside the range that can be achieved by dsRNA transfection in cell culture 
. Similar knockdown efficiencies have been found in whole mosquitoes 
A point of concern is that the AQP dsRNA injections likely resulted in AQP knockdown in tissues other than the MTs. However, excreted watery liquids have been shown to always pass through the MT in adult mosquitoes 
. Therefore we expect the effects of AQP knockdown in other tissues to be negligible for the outcomes of our in vivo
As a positive control for successful AQP inhibition we used mercury ions (Hg2+
), which are capable of binding with a cysteine and an alanine within the AQP pore, thus obstructing water transport through the channel. Mosquitoes injected with 200 uM HgCl2
were still able to excrete 10% of the injected fluid in one hour. This might be due to paracellular permeability, water transport through the space between the cells, which has been described in insect MTs and can be enhanced by kinins 
Using RNAi knockdown mosquitoes we confirmed the function of three different AQPs in the MTs of Ae. aegypti. While RNAi control mosquitoes were able to excrete about 40% of the injected fluid in one hour, knockdown of single AQPs (AaAQP1, 4, or 5) resulted in a significant decrease in excretion. Simultanous knockdown of all four MT-expressed AQPs reduced excretion down to 18% of the injected fluid in one hour, indicating that a combination of AaAQPs 1, 4, and 5 performs water transport in the MTs of female Ae. aegypti.
The redundancy of function we have observed here has also been noted in human kidney where seven different AQP proteins are expressed 
. One plausible explanation would be that several AQP genes with different promoters allow the fine regulation of AQP expression in stage- and tissue specific manner or after a range of stimuli.
There is a great need for the development of novel, effective insecticides to fight insect vectors, since the public health insecticides currently in use are based on only a limited number of active compounds 
. Because of their vital importance in insect larvae and adult homeostasis, insect AQPs could become targets for the development of novel insecticides. The study presented here has identified six genes encoding putative AQP membrane transporters in Ae. aegypti
and demonstrated the functional role of three of them in regulation of water transport. Further analysis of these AQPs and their regulation has the potential to contribute to the future development of novel anti-vector strategies.