Female mosquitoes use a combination of cues to find their vertebrate hosts and blood-feed. Their feeding behavior not only annoys us, but creates a potent pathway for disease transmission. For example, Aedes aegypti are vectors of viral diseases such as yellow fever, dengue, chikungunya and Zika [1, 2]. Certain mosquito species like Ae. aegypti and the malaria vector, Anopheles gambiae, have evolved a preference for humans, which makes them efficient vectors for disease transmission [3, 4]. Ae. aegypti mosquitoes have evolved a preference for human hosts (Anthropophily) from an ancestral subspecies that does not prefer humans (Zoophily) [5, 6*]. Host discrimination requires olfactory receptor function and has been linked to specific receptors that have increased expression and odor sensitivity in anthropophilic Ae. aegypti [6*, 7]. From these studies, it is clear that Ae. aegypti uses olfaction to find their human hosts. Along with olfaction, other sensory pathways are also likely to participate in the detection of humans by mosquitoes. A comprehensive understanding of the cues that attract mosquitoes to humans, the receptors that detect them, and the neural circuits they activate will provide the necessary insight to develop new strategies to disrupt host-seeking behavior. To achieve this goal, genetic tools are now available.

Our ability to understand the molecular basis of mosquito behavior has been enhanced by the recent development of genome editing tools such as CRISPR-Cas9 RNA-guided nucleases, TALE-effector Nucleases (TALENS) and Zinc Finger Nucleases (ZFNs) in Anopheles and Aedes [8]. These approaches can be employed to facilitate targeted mutagenesis at any gene of interest to determine their contribution to host detection and blood feeding. These techniques could also be used to integrate genetic tools to map the neural circuits that enable these behaviors. Targeted mutagenesis of the olfactory receptors has been successfully performed in Ae. aegypti and has been linked to a reduction in host attraction [7, 9]. These techniques have already identified multiple genetic pathways that mosquitoes employ to detect their hosts, but many questions remain. This review seeks to survey the progress made in understanding the molecular mechanism of mosquito host detection.

Odor is a critical cue that signals the presence of a host to mosquitoes [10]. Human odor is a complex blend of chemicals [11]. Skin microbiota plays a large role in generating volatile compounds that attract mosquitoes [12]. In Ae. aegypti and An. gambiae, odors that elicit both electrophysiological and behavioral responses have been found. Among these compounds are lactic acid, ammonia, ketones, sulfides [13–15], 1-octen-3-ol [16], and carboxylic acids [17]. The odors emanating from a host are sensed via olfactory receptors, which can be found on the mosquito antennae, maxillary palps, and proboscis (Figure 1A, [18]). Exposing female mosquitoes to CO₂ induces flight takeoff and sustained flight [19]. CO₂ is detected by gustatory receptors that are expressed in the capitate peg sensilla of the maxillary palp [20]. Identifying which components of the diverse set of human odor-ligands are the most salient is key step in understanding how mosquitoes detect humans.

Heat attracts mosquitoes to their hosts at close range [21]. Mosquitoes will land on inanimate objects set at human body temperature in the presence of CO₂ [9*, 22]. Electrophysiological studies showed that there is an antagonistic pair of thermosensitive neurons within the coeloconic sensilla of the Ae. aegypti antennal tip where one sensillum is tuned to temperature rise and the other is sensitive to cold [23]. The integration of the responses from these two sensilla has been proposed to allow mosquitoes to respond to temperature changes and host thermal cues. The response to thermal cues may depend on the background ambient temperature, which would necessitate that mosquitoes possess a mechanism for sensing thermal contrast. The TRPA1 receptor allows mosquitoes to avoid warm objects that exceed host body temperature aiding the detection of thermal cues [24*]. The sensor(s) that allow mosquitoes to detect attractive heat cues are still unknown. Ionotropic Receptors (IRs) that are temperature responsive have been found in Drosophila [25, 26]. Further studies are needed to identify whether these receptors or others are important for mosquitoes to detect the temperature of their hosts.

Mosquitoes are also guided by visual cues to fly towards their hosts [27]. Adult mosquitoes possess compound eyes that are sensitive to varying light intensity [27]. It has been documented that photoreceptors in night-biting mosquitoes, An. gambiae adjust to varying light intensity by regulating the rhodopsin levels. This could enhance visual sensitivity to a potential host in low light conditions [28]. Unlike Anopheles, visual cues are proposed to be more crucial for day-biting mosquitoes including hematophagous Aedes and Culex, but little is known about this. Visual cues likely play an intermediate role in host detection by integrating long-range odor plume tracking with shorter-range cues [29*]. For instance, CO₂ or human odor can increase the ability of mosquitoes to pay attention to visual cues by enhancing visual flight navigation to the host [30, 31]. Understanding the connection between olfactory sensitivity, flight navigation, and visual target selection will help the field identify the behavioral neural circuits that enable mosquito host-seeking.

While female mosquitoes are guided by other cues to fly towards their hosts (Figure 1B), the tastants on the skin likely promote blood feeding once they land. After landing, they soon pierce the skin and draw blood from small blood vessels [31]. The mouthparts of the female mosquito are highly specialized for blood feeding and contain sensory hair cells which help locate blood under the skin [32*]. The An. stephensi proboscis does not only respond to taste but also detects thermal cues [22]. The transcriptome of the Ae. aegypti proboscis has been recently identified [33]. Genetically manipulating the chemoreceptors expressed in the proboscis could provide insight into mosquito biting behavior and possibly provide evidence for a role in host detection during flight. The contact cues on human skin and the receptors that sense them remain for the most part elusive.

Genetic analysis has demonstrated that mosquitoes integrate multiple stimuli to find their hosts. One of the most striking examples is the gating of multiple cues by CO₂ [9*]. Ae. aegypti mosquito attraction requires at least two cues. For example, neither thermal cues, nor lactic acid are attractive on their own. If CO₂ is combined with either of these cues, mosquito attraction is greatly enhanced. If sensitivity to CO₂ is lost, as in Gr3 mutants, this synergistic effect of CO₂ does not occur. Host odor may also gate mosquito responses to hosts [7]. Mutant mosquitoes lacking the olfactory receptor co-receptor (orco) gene have reduced attraction to human odor, but in the presence of CO₂, there was no difference between the wild-type and orco mutant mosquitoes [7]. In addition, both Gr3 and orco mutants respond normally to human arm in olfactometer assays [7, 9*]. Taken together, these results demonstrate that mosquitoes have robust and redundant mechanisms to detect human hosts. The disruption of one pathway is unlikely to eliminate mosquito host-seeking. Employing multiple sensory pathways may increase the chance of a mosquito successfully targeting a host.

Mosquitoes may also integrate odor with taste cues in response to hosts. Riabinina et al., [34] recently documented that olfactory receptor neurons (ORN) expressing the orco gene project from the labella on the proboscis to the suboesophageal zone (SEZ) of the brain, and suggests that the SEZ may integrate odor and taste cues during blood feeding. This region of the insect brain has been shown to be critical for taste integration in Drosophila [35]. In Ae. aegypti, in vivo calcium imaging revealed the activation of some ORs by some volatile compounds in the blood, and removing the function of the ORs in the stylet impairs blood feeding [31]. Clearly, the ability to integrate host sensory cues represents an essential mechanism employed by mosquitoes to guide host detection and blood-feeding behavior.

Insects have evolved complex repertoires of chemosensory receptors to respond to their environment including: odorant receptors (ORs), ionotropic receptors (IRs) and gustatory receptors (GRs) [36]. In Drosophila, the ORs are expressed in the dendrites of ORNs. The sensillar lymph surrounding the ORNs is densely packed with odorant-binding proteins (OBPs) that are hypothesized to be involved in odorant uptake and odor delivery to the ORs, but their role remains unclear [37]. The neural processing of olfactory information has been extensively studied in Drosophila [38, 39]. The axons of the ORNs project to the antennal lobe where they synapse with projection neurons (PN). In turn, PNs convey olfactory information to the mushroom body (MB) and lateral horn (LH) of the brain which subsequently leads to a behavioral response. How mosquitoes process host cue information has not been well established, but the lessons from Drosophila are likely to apply.

In Drosophila, ORs form a heteromeric complex with orco for their targeting to the cell membrane [40, 41]. The OR pathway plays a critical role in mosquito preference for human hosts and DEET repellency [7, 42]. ORs from An. gambiae have been comprehensively deorphanized using the Drosophila empty neuron system [43]. These ORs tested seemed to be narrowly tuned to several odor components that emanate from humans such as 1-octen-3-ol (present in human breath), 2,3-butanedione (by-product of metabolized sweat) and indole (human sweat volatile) [43]. The narrow tuning of these ORs may serve to improve cue salience [44].

The IRs are expressed in the dendrites of the ORNs innervating the coeloconic sensilla [45]. There are at least two IR co-receptors, Ir8a and Ir25a, and possibly a third, Ir76b [45, 46]. These co-receptors form an odor-responsive ion channel complex with odor-selective IRs. Drosophila IRs respond primarily to amines and acids while ORs respond to esters and alcohols [47]. Beyond olfaction, Drosophila IRs play a role in taste [48, 49], moisture and temperature sensing [25, 26], and possibly in the auditory system [50]. Although GRs are usually involved in taste [35], specialized GRs have been implicated in other sensory modalities in insects including light sensing [51], warmth sensing [52], and CO₂ detection [53]. Excluding the Grs that detect CO₂, it remains unknown whether the orthologues of these receptor genes respond to similar stimuli in mosquitoes [9*].

Given that olfaction is key for mosquitoes to detect their human hosts, identifying new odors that modulate mosquito olfactory receptor function is a promising approach to combatting mosquito-borne disease [54, 55] (Fig. 2). This can be accomplished using high-throughput screening assays to identify synthetic and natural compounds that activate receptors associated with repellency or inhibit receptors associated with attraction [56]. This strategy can help develop novel spatial mosquito repellents or create odor baits for traps that reduce mosquito populations. Our current understanding of mosquito host detection makes it likely that effective odor-baited traps would need to include multiple cues such as CO₂ and heat to recreate the multimodal sensory experience that drives mosquitoes to their hosts. Next generation mosquito repellents could block multiple chemosensory pathways to render the human host “invisible”. As an alternative, new repellents that overstimulate a specific chemosensory pathway could cause mosquitoes to avoid humans. To accomplish this important goal, we will need to know the specific receptors that enable mosquito attraction and repellency.

We have reviewed what is known about the multiple cues that attract mosquitoes to humans and their molecular sensors. Host-seeking is not completely abolished even when either the CO₂-sensing or OR pathways were disrupted in Ae. aegypti [7, 9*]. This clearly shows that multiple cues must be considered to understand how mosquitoes find their human hosts. Our current knowledge is not sufficient to develop new behavioral control strategies, but sets the stage for further studies. To efficiently develop these new strategies, we need to have a comprehensive understanding of how mosquitoes detect their human hosts from the perception of cues at the periphery, to the integration of the information in the central nervous system, and finally, to the motor circuits that drive the behavior. Genome editing tools such as CRISPR-Cas9 and other genetic manipulations such as using the GAL4-UAS system to mark or manipulate neural circuits could play a significant role in addressing how the mosquito’s brain responds to human cues [57]. We have learned so much from the genetic analysis of Drosophila behavior. Using similar approaches to understand mosquito behavior may provide the mechanistic insight to break the cycle of mosquito-borne disease transmission.