Aedes aegypti (Ae. aegypti) mosquitoes are major vectors of dengue, chikungunya and other arboviral diseases. Ae. aegypti’s capacity to reproduce and to spread disease depends on the female mosquitoes’ ability to obtain blood meals and find water-filled containers in which to lay eggs (oviposit). While humidity sensation (hygrosensation) has been implicated in these behaviors, the specific hygrosensory pathways involved have been unclear. Here we establish the distinct molecular requirements and anatomical locations of Ae. aegypti Dry Cells and Moist Cells, and examine their contributions to behavior. We show that Dry Cell and Moist Cell responses to humidity involve different ionotropic receptor (IR) family sensory receptors, with dry air-activated Dry Cells reliant upon the ionotropic receptor Ir40a, and humid air-activated Moist Cells upon Ir68a. Both classes of hygrosensors innervate multiple antennal sensilla, including sensilla ampullacea near the antennal base as well as two classes of coeloconic sensilla near the tip. Dry Cells and Moist Cells each support behaviors linked to mosquito reproduction, but contribute differently: Ir40a-dependent Dry Cells act in parallel with Ir68a-dependent Moist Cells to promote blood feeding, while oviposition site seeking is driven specifically by Ir68a-dependent Moist Cells. Together these findings reveal the importance of distinct hygrosensory pathways in blood feeding and oviposition site seeking, and suggest Ir40a-dependent Dry Cells and Ir68a-dependent Moist Cells as potential targets for vector control strategies.   

Calcium imaging:

To monitor intracellular levels of intracellular calcium in Ir40a-positive cells, QUAS-GCaMP7s (marked by 3xP3-CFP) mosquitoes were crossed to the Ir40a^RFP and Ir40a^EYFP knock-ins (which express QF2 under the control of endogenous Ir40a regulatory sequences). Ir40a^RFP/+ ; QUAS-GCaMP7s/+ and Ir40a^RFP/Ir40a^EFYP; QUAS-GCaMP7s/+ animals were used to assess responses in control animals and Ir40a mutants, respectively. As Ir68a QF2 knock-ins drove weaker GCaMP expression, imaging of Ir68a-positive cells was performed in a yellow mutant background, whose lighter cuticle facilitates imaging. Ir68a^RFP/+ ; QUAS-GCaMP7s/+ animals were crossed with a yellow mutant strain for two generations, and yellow/ yellow; Ir68a^RFP/+ ; QUAS-GCaMP7s/+ animals were used for control imaging. These mosquitoes were crossed to yellow/ yellow; Ir68a^RFP /+ individuals and descendants were molecularly genotyped to create a stable stock of yellow/ yellow; Ir68a^RFP/Ir68a^RFP; QUAS-GCaMP7s/+ for Ir68a mutant imaging.

For imaging, female mosquitoes (1-4 days old) were anesthetized on ice. Double-sided tape was applied to a microscope slide, which was then covered with a plastic coverslip (Fisherbrand, #12-547). To image coeloconic sensilla, mosquitoes were positioned laterally with their heads affixed to the double-sided tape, allowing the antenna to extend across the plastic coverslip. Adjacent to the mosquito head, another strip of double-sided tape was applied at the base of the antenna to secure its position. To minimize movement, a narrow strip of double-sided tape was carefully placed on segment 12 for tip imaging or on segment 13 for imaging near tip segments. For sensilla ampullacea imaging, a thin clay layer under the coverslip elevated the basal antennal segments, facilitating a horizontal position of the basal antennal segments. A narrow strip of double-sided tape was placed on segment 3 to restrict movement.

Stimuli were delivered using a micromanipulator station holding two tubes that pump dry air (Middlesex Gases), each at 1 liter per minute, controlled by a flowmeter (Middlesex Gases, #7965B-J03G). For humidity response assays, one tube passed through a water-filled flask to produce moist air at ~90% relative humidity (RH), while the other tube, passing through an empty flask, delivered dry air at ~7% RH. Temperature responses relied on an ~ 5°C differential between the two airstreams, created by warming one tube with a heating ring. Tube positions were alternated between animals.

Images were acquired using an Olympus BX51WI microscope fitted with an Olympus SLMPlan 50x/0.45 objective and a Hamamatsu OrcaR2 camera recording at 10 frames/sec. During imaging, the first stimulus was applied for 10 seconds, then switched to the other stimulus for 20 seconds, and then switched back for 10 seconds (400 frames in total). Temperature at the antenna was logged at 4.25 frames/ second using DAQExpress (Version 5.1.10) coupled with a temperature microprobe (Physitemp, #IT23).  Recordings were processed in Fiji (RRID:SCR_002285) using Stackreg to correct for movement. A ROI tool was used to circle cell regions and three background regions, and their intensity was calculated among 400 frames. Baseline (F₀) was calculated by the average intensity of cells at 5s to 1s prior to the initial stimulus shift. Cell responses were quantified as the average ∆F/ F₀ from 2 to 5 sec after the initial stimulus shift.