Ticks Home in on Body Heat: A New Understanding of Ectoparasite Host-Seeking and Repellent Action

Ticks are second only to mosquitoes as vectors of disease to humans and animals. Commercial insect repellents reduce or prevent potentially infectious tick bites by disrupting tick host-seeking behavior. Tick host-seeking is mainly ascribed to the Haller’s organ, a complex sensory structure on the tick foreleg that detects odors, carbon dioxide and heat 4–7, but these host-seeking mechanisms and the mechanism of their disruption by repellents are not well understood1,2. There is anecdotal evidence that ticks and other ectoparasites are attracted to heat, but it has never been demonstrated that they use radiant heat to detect hosts at a distance. In fact, previous attempts to do this have concluded that radiant heat is not used by ticks. Here we show that Amblyomma americanum and Dermacentor variabilis ticks can sense and home in on a human from several meters away, guided by radiant heat sensed by the Haller’s organ, and specifically the capsule, a covered spherical pit organ. An aperture in the capsule cover confers directionality and highly reflective interior surfaces of the capsule provide high sensitivity. Low concentrations of the insect repellents DEET, picaridin, 2-undecanone, citronellal and nootkatone eliminate thermotaxis without affecting olfactory-stimulated host-seeking behavior. Our results demonstrate that the tick Haller’s organ capsule is a radiant heat sensor used in host-finding and that repellents disrupt this sense at concentrations that do not disrupt olfaction. We anticipate this discovery to significantly aid insect repellent research and development of innovative strategies for protection against ectoparasites and vector-borne disease.

Main Text: Ticks transmit the widest array of pathogens of any arthropod vector, including viruses, bacteria, fungi, protozoa and helminths, and are second in importance only to mosquitoes as vectors of disease to humans and animals 1 . The American dog tick Dermacentor variabilis (Acari: Ixodidae) and the lone star tick Amblyomma americanum (Acari: Ixodidae) are the two most abundant ticks in the US 2 ; both have the capacity to transmit ehrlichiosis, tularemia and Rocky Mountain spotted fever. Also of public health concern is the increasingly prevalent galactose-alpha-1,3-galactose red meat allergy associated with A. americanum bites 2,3 .
Ticks find hosts using the Haller's organ on the foreleg to detect odors, carbon dioxide and heat [4][5][6][7] , but the only study to investigate the role of heat in tick host-seeking concluded that radiant heat is not used 4 . The part of the Haller's organ known as the capsule is thought to serve an olfactory function 4,8-10 , but its camera-lucida-like structure as a covered, sensilla-containing spherical pit with an aperture 8,11 seems better suited to detect radiation. Although the aperture would severely hinder odorant diffusion, it would confer directionality for radiation-sensing by restricting field of view. We show that previous studies on heat sensing by ticks were flawed and that the Haller's organ capsule is indeed an exquisitely sensitive radiant heat sensor that guides host-finding. We further show that exposure to common tick repellents disrupts this thermotaxic behavior, which we propose to be a novel mode of action of repellents.
Questing unfed virgin adult A. americanum females placed on the floor in the center of a 10 cm wide X 10 cm high X 1m long thermotaxis arena ( Fig. 1) were strongly attracted to a warm surface (target) at one end of the arena, 50 cm away, when it was at 37 °C or warmer.
Most ticks moved toward the warm target almost immediately, and of the few that initially moved toward the opposite (cold) end of the arena, most reversed direction, so that after 5 min only one tick of 48 remained within 10 cm of the cold wall, while 40 were within 10 cm of the 3 37 °C target. Seven ticks ended the trial between these end regions (Fig. S1, Movie S1). With the target at 22 or 30 °C, many ticks did not move far from the origin, and an approximately equal number moved toward either end, with a strong tendency to continue moving in the initial direction. Results were similar for A. americanum males (Fig. S1) and both sexes of D. variabilis (Fig. S2).
Net movement and average initial velocity toward the target began immediately when it was at 37, 40 or 44 °C, and except in the case of A. americanum females, also at 50 °C (Fig. 1B, D), indicating that ticks easily detected the warm target from 50 cm away. Initial speed did not vary consistently with temperature ( Fig. 1C), indicating that temperature guides but does not stimulate movement. Initial velocities declined at target temperatures above 40 °C (Fig. 1D).
All ticks studied showed a significant warm/cold preference at 37 °C and above for final choice and initial movement ( Fig. 2A, B). The fraction choosing an end region was always higher for initial movement compared to final choice, because some ticks stopped moving before reaching either end. A preference index (PI) ranging from -1 to +1, with 0 indicating no preference, quantifies warm preference (Fig. 2C, D). PI was near zero at 22 and 30 °C. Initial PI at 37 °C was between 0.7 and 0.95, indicating that at least 70 % of ticks detect the 37 °C target at 50 cm. Except for A. americanum males, initial PI showed a sharper optimum than final PI, indicating that temperatures above 40 °C may be initially inhibitory.
To better assess tick heat-sensing range, we started A. americanum females at the far end of the arena, 100 cm from the target. With the target at 22 °C, many ticks immediately began moving toward it, ending near the center of the arena, at -55 ± 4 cm (mean ± SEM; Fig.3A).
With the target at 40 °C to radiate energy at the same rate as human skin at 37 °C 12 , average movement toward it was much faster from the start and the ticks reached an average distance of only 16.5 ± 4.5 cm from it (Fig. 3A). From the distribution of initial velocities (Fig. 3B), 45 of the 88 ticks showed initial movement toward the target when at 22 °C, and can be considered to be moving spontaneously, whereas the remaining 43 were not. With the target at 40 °C, 76 ticks began moving toward it immediately, representing 45 that were moving spontaneously and 31 that were thermotaxing. Since 31 of the 43 ticks that were not moving spontaneously, or 72%, began thermotaxing immediately from 1m away, we can conclude that at least 72% of adult A. americanum females detect the 10 X 10 cm target at a distance of 1 m, just as many as at 0.5 m.
A longer arena or smaller target is needed to assess the range of tick heat detection, but we can conclude that since an adult human torso is approximately 4 times the width of our target, a tick could detect an adult human at a distance of < 4 meters under ideal conditions.
The fastest walking speed of A. americanum females in Figure 3B was 1.65 cm•s -1 , equal to the speed observed for Hyalomma ticks hunting in the field 5 . The fastest speeds we observed for A. americanum males, and D. variabilis females and males were 1.5, 1.2 and 1.1 cm•s -1, respectively.
We propose that the tick Haller's organ capsule is a radiant heat sensor. Conduction of heat along the floor and sides of the arena was insignificant 12 and elimination of the possibility of convective heat transfer from the target to the ticks by interposing an IR-transparent plastic barrier had no significant impact on the percentage of ticks choosing the 40 °C target (Fig. 3C).
On the other hand, when both arena end walls were clad with aluminum foil to eliminate radiant heat emission without affecting convection, or when the Haller's organ capsule apertures were waxed over, the ticks no longer preferred the 40 °C target (Fig. 3C), confirming that the capsule is a radiant heat sensor that guides tick thermotaxis.
We reasoned that for the capsule sensillae to sense temperature of a distant source, the capsule walls and the interior surface of the cover must be reflective. Viewed with epiillumination, the aperture of the intact capsule appeared to emit reflected light (Fig. 3D). When the cover was removed, the walls of the capsule and the inside faces of cover fragments scattered on the surrounding cuticle appeared highly reflective (Fig. 3E).
Commercial insect repellents dose-dependently reduced tick thermotaxis (Figs. 4A, B, S3). Exposure to 50 ng•cm -3 DEET, 54 ng•cm -3 picaridin and 42 ng•cm -3 2-undecanone completely and reversibly abolished A. americanum female thermotaxis, with 50 % inhibitory concentrations of 26, 31 and 17 ng•cm -3 , respectively (Fig. S3B). Thermotaxis was also disrupted by citronellal and nootkatone (Fig. 4B), but fewer trials were conducted because their residues were difficult to remove from the arena. Individual ticks exposed to any of the repellents behaved as if there were no warm target, beginning to move immediately and continuing in the direction in which they started, but showing no clear end wall preference (Fig. S3A, Movie S1).
Number of ticks making a choice (Fig. S3B) and initial speed (Fig. S3C) did not decline with increasing repellent concentration, indicating that repellents affect heat detection but not hostseeking behavior. We next examined the effect of DEET on the olfactory response to CO2, a necessary and sufficient stimulant of host-seeking behavior 6,13 . In control Y-tube bioassays testing medical grade air in both arms (Fig. S4), 32% of ticks made a choice, with approximately equal numbers going to either arm (Fig. 4C, sham CO2). Introducing 4% CO2 into one arm, the concentration in human breath, resulted in 52 % of ticks making a choice -primarily the arm with CO2. Exposure to 50 ng•cm -3 DEET had no effect on tick detection of CO2 in Y-tube bioassays (Fig. 4C), confirming that DEET at this concentration disrupts thermotaxis but not host-seeking behavior.
We have shown here for the first time that ticks thermotax guided by a radiant heat sense in the Haller's organ capsule sensitive enough to locate a human body 4 meters away. Radiant body heat includes wavelengths between 5 and 20 µm 14 , well outside the spectral sensitivity of rhodopsins, and biological radiant heat detectors are normally thermosensors arranged in a pit such that they are warmed by radiant heat from the target in preference to that from the self 15 .
We have demonstrated that the Haller's organ capsule is in fact such a "pit organ", with an aperture and a reflective interior lining for high directionality and sensitivity. The ability of A.
americanum females to detect a 10 cm wide target at 1 m indicated a field of view of 6 degrees or less. Recent findings have shown that the tick Haller's organ, but not specifically the capsule, is required for attraction to 880 nm near-infrared light 16 . While well outside the range of thermal IR, it is likely that the tick radiant heat sensors were warmed enough by this light to stimulate thermotaxis. However, we demonstrate for the first time that the capsule is a radiant heat sensor that plays a critical role in tick host-seeking.
The mechanism of action of arthropod repellents is not well understood. While they are generally thought to repel or to cause directed movements away from the source 17 ,we have shown that several of the most commonly used personal repellents specifically disrupt tick thermotaxis at concentrations that do not affect CO2-stimulated host-seeking behavior.
Repellents have been studied extensively on mosquitoes, which can detect them by olfaction and gustation 18,19 , but it is not clear whether this detection ability is the mechanism of repellency in the field. In light of the findings presented here, the roles of olfaction and heat sensing in repellent action must be re-evaluated.
Captions for Movie S1 Other Supplementary Materials for this manuscript include the following: Movie S1 Table S1 -all choice data and statistical analyses.    DEET: air in one tube contained 4% CO2 and ticks were exposed to 50 ng•cm -3 DEET. Significance of warm/cold or CO2/air difference: **: p<0.001; *: p<0.05. Thermotaxis Assays: We used a thermotaxis assay to investigate the detection range, temperature dependence and repellent sensitivity of heat perception in ticks and to identify the sensory organ responsible for this sense. A warm surface rather than an electronic infrared radiation (IR) source was used as the target for thermotaxis since the radiant heat emitted by the warm surface would closely mimic the spectrum and intensity of thermal radiation emitted by host animals, allowing the determination of the temperature sensitivity of the radiant heat sense.

Ticks
Furthermore, it is important to use a large enough surface as a radiant heat source because, while the intensity of radiation from each point on the surface decreases as the square of distance, the surface area within the field of view of the radiant heat sensor increases by the same factor, so that the intensity of radiant heat entering the sensor is independent of distance as long as the target fills the entire field of view. The key parameters of a radiant heat sensor are therefore temperature sensitivity and field of view.
Thermotaxis assays were conducted in an arena consisting of a long box with interior dimensions of 10 cm x 10 cm x 100 cm (Fig. 1)  Both TEC plates were always temperature-controlled, even when at ambient temperature. Robust tick response within the thermotaxis arena required that ticks be actively displaying the hostseeking behavior known as questing, which is stimulated by CO2 in the breath of hosts 10,21 .
Questing of tick test subjects was induced by brief agitation of the holding vial and by exposure to the breath of the experimenter. All ticks were questing prior to placement into the arena for thermotaxis bioassays. With the exception of temperature sensitivity trials, the temperature of the target was set to 40 °C, the temperature at which, according to the Stefan-Boltzmann law, a surface with an emissivity of 0.94, as measured for the blue anodized TEC plate surface, would radiate energy at the same intensity as human skin does at 37 °C, which has an emissivity of 0.98 14 .
Previous investigations that failed to find that ticks 4 or other hematophagous insects 23 were attracted by radiant heat used a test tube covered with aluminum foil as the non-emitting source and a similar tube with a layer of cellophane over the aluminum foil as the IR-emitting source. We were skeptical that the thin, transparent layer of cellophane would emit sufficient radiant heat, so we wrapped a section of a 1l graduated cylinder with aluminum foil and another section of the same cylinder with aluminum foil covered with cellophane, and filled the cylinder with 37 °C water. The thermal imaging camera measured the temperature of the uncovered glass as 37 °C and that of the aluminum foil section as 21 °C, because the foil emits almost no thermal IR itself, but rather reflects that from the room temperature surroundings. The cellophane over aluminum foil section, which was intended in the cited studies to provide radiant heat, appeared to the thermal imaging camera to be at 24.6 °C (Fig. S2) indicating that the cellophane radiated very little heat and mostly transmitted ambient thermal radiation reflected by the underlying aluminum foil. We can conclude that previous studies failed to show that ticks were attracted by radiant heat because the radiant heat stimulus used was much less intense than intended. Another flaw with this method is that if there had been a warm radiating surface in the arena, its radiation To confirm that the Haller's organ posterior capsule is the IR sense organ, thermotaxis trials were conducted with unfed virgin adult A. americanum females before and after application of low-melting dental wax (Kerr, Orange, CA, USA) onto the posterior capsules of both front legs using a hot 0.25 mm diameter silver wire attached to the tip of a soldering iron (Weller, Apex, NC). Care was taken to exclusively place the wax onto the posterior capsule; any wax accidentally applied to the anterior pit sensillae was removed using sterile forceps and a minute piece of kimwipe (Kimberly-Clark, Irving, TX). Following wax application, ticks were allowed to recover for 10 min prior to being tested in thermotaxis trials. We also attempted to localize the IR sense to the forelegs and the Haller's organ by means of amputation, but the questing behavior of amputated test subjects was greatly reduced. the total flow through the olfactometer and exhausted them out of the test area. Ticks were preexposed to repellent vapors for up to 5 min, as previously described, prior to placement into the Y-tube, with no additional repellent exposure during Y-tube bioassays. 50ng/cm 3 and 100ng/cm 3 DEET, which are 1X and 2X, respectively, the concentration found to completely abolish thermotaxis, were evaluated in Y-tube bioassays. After repellent pre-exposure, ticks were immediately placed at the Y-tube starting mark and their movements were filmed for 5 min.

Repellent Thermotaxis
Olfactometer bioassays consisted of six runs of eight unfed virgin adult A. americanum females randomly selected from a total population of 250 and acclimated to experimental conditions for 20 min prior to repellent pre-exposure. To eliminate contamination and positional bias, all Ytube components were washed thoroughly with 70% ethanol between replicates and the CO2 gas was alternated between the two arms of the Y-tube. Ticks were recorded as positive responders if they made a choice in Y-tube bioassays, moving 1 cm past the choice point (Fig. S4). Trials were filmed and scored as previously described and those data were analyzed using the Generalized Linear Model in R, as described above.