Species-specific differences in mosquito behavioral response to repellents
We previously showed that DEET does not directly activate odorant receptors (ORs) in An.coluzzii mosquitoes, and does not directly repel them [13]. This is in contrast to what has been reported for Ae. aegypti [6-8] and Cx. quinquefasciatus [4, 9] mosquitoes. Therefore, we tested the direct effect of DEET (not in contact with other odorants) in the close proximity response assay on these two mosquito species (Fig. 2). In addition, we tested the effect of other commonly used synthetic repellents (IR3535 and picaridin) as well as three natural repellents (lemongrass oil, eugenol, and PMD) on all three mosquito species. Consistent with previous findings [4, 6, 8, 9, 13, 18, 20], DEET did not repel An. coluzzii mosquitoes, but was mildly repulsive to both Ae. aegypti and Cx. quinquefasciatus mosquitoes (Fig. 2). The synthetic repellents IR3535 and picaridin were not repulsive to any of the mosquito species tested. The greatest behavioral differences across the mosquito species were in their responses to natural repellents. An.coluzzii mosquitoes were repelled by lemongrass oil and PMD only (Fig. 2a). Eugenol showed a weak repellent effect to Anopheles mosquitoes, but it was not significantly different than paraffin oil (P = 0.08, Fig. 2a). In contrast, Ae. aegypti and Cx. quinquefasciatus mosquitoes were repelled by lemongrass oil, PMD, and eugenol (Fig. 2b, c). PMD was more repellent to Anopheles and Culex mosquitoes than to Aedes mosquitoes, whereas eugenol was more repellent to Aedes mosquitoes than to Anopheles or Culex mosquitoes. Differences in repellencies might reflect species-specific differences in their olfactory receptor neurons to respond to these odors.
Olfactory behavioral responses to human odorants measured by the close proximity assay
A better understanding of how odors repel mosquitoes could lead to the generation of improved spatial repellents. Work in Drosophila has suggested that even attractive odorants can become repellent at high concentrations [24, 25]. For example, the strong attractant apple cider vinegar became less attractive at a higher concentration due to the additional activation of an olfactory receptor neuron at the high odor concentration [24]. We questioned if human odorants (which are often attractants) might also become repellent to mosquitoes at high concentrations. To address this, we first used the close proximity assay to test the behavioral response of An. coluzzii mosquitoes towards two human skin odorants 1-octen-3-ol and benzaldehyde at a range of concentrations (0.1%, 1%, 10%, and 100%). At 0.1% concentrations for both odorants, Anopheles mosquitoes did not respond in this assay (Fig. 3a). In contrast, 1-octen-3-ol caused mosquitoes to fly away at 1%, 10%, and 100% concentrations, while benzaldehyde caused mosquitoes to fly away only at 10% and 100% concentrations (Fig. 3a). These data suggest that Anopheles mosquitoes maybe attracted to a narrow range of host-odorant concentrations, and concentrations higher than that could be behaviorally repellent. We next aimed to address if the higher odor concentrations were leading to changes in olfactory neuron responses that reflected changes in olfactory behaviors.
Higher odorant concentrations recruit additional olfactory neurons
The concentration of 1-octen-3-ol in human sweat is typically 0.49 µg/ml [26]. Although it is difficult to directly compare odor stimulations, the repellent concentration of 1% 1-octen-3-ol is likely high in comparison with the concentration of 1-octen-3-ol found in human sweat [26]. To begin to address how high concentrations of non-repellent odorants might drive mosquito repulsion, we used calcium imaging to examine the Anopheles mosquito antennal response towards 1-octen-3-ol and benzaldehyde at 0.1%, 1%, 10%, and 100% concentrations. We used transgenic mosquitoes in which the calcium indicator GCaMP6f was expressed in all neurons that express the odorant receptor coreceptor Orco (genotype: Orco-QF2, QUAS-GCaMP6f [13]). This enabled us to directly monitor responses in olfactory neurons (Orco+ neurons) towards the test odorants. Low concentration of both odorants elicited specific patterns of olfactory neuron activities, which presumably reflect olfactory neurons expressing odorant receptors most sensitive to these odorants. Higher concentrations of both odorants elicited stronger responses in the same antennal olfactory neurons (Fig. 3b). Interestingly, at higher odorant concentration, more neurons responded than at low concentrations (Fig. 3b). This suggests that higher concentrations of non-repellent odorants not only more robustly activate highly sensitive olfactory neurons, but might recruit low-sensitivity, and potentially repellent-activated, neurons. It is also possible that widespread activation of many olfactory neurons might be interpreted by the mosquito olfactory system as a repellent signal [27]. Overall, these data highlight olfactory neuron activity patterns that might be linked to repellent responses.
Repellent mixing modulates the odor potency of repellent components
By reducing the volatility of odorants with which they are mixed, synthetic repellents like DEET can function to ‘hide’ human odors from host-seeking mosquitoes [13]. However, this also suggests that spatial repellents, when mixed with DEET, might not be as effective at repelling mosquitoes than the spatial repellents alone. We previously reported that DEET prevented eugenol from strongly activating olfactory neurons but DEET did not affect the response to lemongrass oil in calcium imaging experiments [13], suggesting that lemongrass oil may still function as a spatial repellent in mixtures with DEET. To address this, we examined the behavioral effect of mixing DEET with the natural repellents lemongrass oil and PMD (Fig. 4). DEET did not change the response of An.coluzzii mosquitoes towards lemongrass oil; all An.coluzzii mosquitoes were similarly repelled by lemongrass oil (30%) and a mixture of lemongrass oil (30%) and DEET (30%) (Fig. 4a). On the other hand, adding DEET to PMD significantly decreased the mosquito repulsion mediated by PMD alone, although the response to the PMD+DEET mixture was still significantly different from paraffin oil (P = 0.04, Fig. 4a). Given that DEET can function to reduce odor signaling we might be able to use calcium imaging of olfactory neurons as a way to correlate behavioral changes to olfactory neurons responses. To test this, we performed calcium imaging of olfactory neurons to test antennal responses to PMD, lemongrass, and their mixtures with DEET. Lemongrass oil and eugenol strongly activated ~2-5 olfactory neurons in antennal segment 11. When mixed with DEET, lemongrass oil was still able to robustly activate the same olfactory neurons (Fig. 4b). On the other hand, when PMD was mixed with DEET, the olfactory responses to PMD were strongly decreased, with less olfactory neurons being strongly activated (Fig. 4b). These results confirm our previous findings with testing the human odorant, and demonstrate that changes in behavioral responses to repellent mixtures are also reflected by observed changes in neuronal activity.