Expression and light-triggered movement of rhodopsins in the larval visual system of mosquitoes

The Aedes aegypti mosquito is the principal vector of the arboviruses causing yellow fever and dengue fever. This mosquito is known to use multiple sensory cues for efficient host-seeking behavior (McMeniman et al., 2014). We sought to characterize the Aedes visual system to understand how vision contributes to host-seeking and other behaviors underlying disease transmission. The Aedes genome contains ten predicted rhodopsins, encoding the G-protein-coupled receptors that initiate visual transduction (Nene et al., 2007). These were classified into five groups based on phylogenetic analysis. Three of these groups are labeled as long-wavelength (>500 nm), short-wavelength (400–500 nm), and UV-sensitive rhodopsin (<400 nm), and two of these groups contain uncharacterized family members. We previously characterized the patterns of rhodopsin expression in the adult Aedes compound eye (Hu et al., 2009, 2011, 2012, 2014). These studies revealed features of the visual system in Aedes that were not predicted from study of other invertebrate models.

One noteworthy feature is the daily cycle of rhodopsin movement from the cytoplasm into the photosensitive rhabdomeric membranes at dusk and then back into the cytoplasm at dawn (Hu et al., 2012). The increased rhodopsin content within the rhabdom at night will increase light sensitivity, which is probably a valuable photoreceptor adaptation for acquisition of visual information in the dim light conditions where mosquitoes are most active. In addition, a prevalent cause of photoreceptor degeneration both in vertebrates (Vihtelic and Hyde, 2000; Okano et al., 2012) and invertebrates (Stark and Carlson, 1984; Meyer-Rochow et al., 2002) is exposure to bright light. For this reason, another benefit of rhodopsin movement out of the rhabdoms during daylight may be to protect photoreceptors from bright-light-induced damage.

Here, we extend the analysis of the Aedes visual system by characterizing rhodopsin expression and light-triggered rhodopsin movement during the larval stage. Mosquito development proceeds through four aquatic larval instars and a mobile aquatic pupal stage prior to emergence of the adult (Clements, 1999). The major components of the larval visual system are the stemmata, or larval ocelli, that mediate phototaxis and diving behaviors during the larval and pupal periods (Kasap, 1977b). We show that there are two classes of photoreceptors within the stemmata. The major stemmatal photoreceptor class expresses the Aaop3 rhodopsin (see Materials and methods for an explanation of Aedes opsin nomenclature). Aaop3 exhibits extensive light-driven relocation, establishing that the light-triggered loss of ∼80% rhabdom volume documented in Aedes larval photoreceptors (White, 1968) is associated with rhodopsin movement. The minor photoreceptor class expresses the rhodopsin Aaop7. This rhodopsin possesses shorter wavelength sensitivity than Aaop3 and exhibits limited light-driven movement. We also show that the major adult stage rhodopsin, Aaop1, is expressed in the developing compound eye during the 4th larval instar. This finding is consistent with earlier results that indicated that the compound eye may contribute to larval and pupal vision (Kasap, 1977a).

From the 1st instar, the visual system of the mosquito larva consists of a set of five stemmata that are positioned laterally on each side of the head (Sato, 1953; White, 1961). Fig. 1A shows a diagram including the names of the stemmata. Fig. 1B–G show images of the mosquito head at different developmental stages. Midway through the 1st instar stage, the development of the adult compound eye initiates within epidermal cells positioned anterior and adjacent to these stemmata. In the 3rd instar larva, two distinct regions corresponding to the dorsal and ventral hemispheres show the initial pigmentation of the adult eye (Fig. 1C, arrowheads). The recruitment and differentiation of adult retinal cells occur from these two sites, such that the two distinct pigmented eye fields are visible in the early 4th instar larva (Fig. 1D, arrowheads). These two areas have merged into a single eye field by the mid-4th instar larval stage (Fig. 1E) and continue to expand anteriorly to produce the fully formed adult eye by the midpupal stage (Fig. 1F). The site of the larval stemmata remains visible at the posterior edge of the adult eye (Fig. 1G, arrow).

Aaop3 is a member of the long-wavelength rhodopsins. Peptides corresponding to amino acid sequences within the N-terminal and C-terminal ends of the Aaop3 rhodopsin were used to generate Aaop3 antisera. These two peptide sequences are poorly conserved with other rhodopsin subfamily members (Fig. 2A,B, left), and the resulting antisera were expected to uniquely recognize the Aaop3 rhodopsin. The antisera stained transgenic Drosophila expressing Aaop3 but not wild-type or transgenic Aaop7-expressing Drosophila (Fig. 2A,B, right).

In the 4th instar mosquito larval head, individual stemmata are recognized by strong actin staining (red) within the rhabdom (R) regions (Fig. 2C). Four stemmata were identified previously (White, 1967; Brown and White, 1972) and named as the dorsal anterior (DA), dorsal posterior (DP), central (C) and ventral (V) stemmata. We identified a fifth, much smaller, stemma that was located dorsally to these four stemmata and named this the satellite (S) stemma (Fig. 2C). The central stemma is composed of 20–22 photoreceptors. The dorsal anterior, dorsal posterior and central stemmata each contain 9–12 photoreceptors. The satellite stemma contains six photoreceptors. In each of the stemmata, the rhabdomeres of all photoreceptors fuse into a common rhabdom structure (Fig. 2C,E).

Fig. 2C,D also shows that Aaop3 (green) is detected within all five stemmata structures. In tissue prepared from a light-treated larva, Aaop3 is largely excluded from the rhabdom structures and found dispersed within the cell body of the majority of stemmatal photoreceptors. There are a small number of photoreceptors in the central and the satellite stemmata lacking detectable Aaop3 expression (Fig. 2C, asterisks).

The identification of Aaop3 as a stemmatal rhodopsin allowed us to test whether light exposure triggers an extensive movement of this rhodopsin. Indeed, in retinas of larvae fixed and dissected in dark conditions, Aaop3 is sequestered within the rhabdoms and found at very low levels within the photoreceptor cell bodies (Fig. 2F–H). The same region within the rhabdom of the central stemmata is not labeled by Aaop3 (Fig. 2F, asterisk) as seen in the light-treated retina (Fig. 2C, asterisk). The comparable views of light-treated and dark-treated stemmata presented in Fig. 2C–E and Fig. 2F–H respectively, show that there is a robust movement of Aaop3 upon transitions between light and dark conditions, with Aaop3 sequestered within the rhabdomeric compartment when larvae are placed in dark conditions.

Antibodies against Aaop7, another member of the long-wavelength rhodopsin group, were created using the same strategy described to generate antisera against Aaop3. Aaop7 shares limited sequence identity with other long-wavelength rhodopsins within the N-terminal region, making it likely that antibodies recognizing this peptide sequence are specific to Aaop7 (Fig. 3A, left). In agreement with this expectation, the resulting antiserum successfully labeled transgenic Drosophila retina expressing Aaop7 (Fig. 3A, middle) but failed to label the wild-type or Aaop3-expressing transgenic Drosophila retina (Fig. 3A, right).

To determine whether Aaop7 is expressed in the mosquito larval stemmata, similar experiments to those already described for Aaop3 were carried out with the Aaop7 antiserum. This approach revealed that Aaop7 (green) is expressed in small groups of photoreceptors within the central stemmata and the satellite stemmata (Fig. 3B, arrows). The location of these groups is consistent with the view that the Aaop3 and Aaop7 rhodopsins are expressed in nonoverlapping sets of photoreceptors of the central and the satellite stemmata. In many preparations, the Aaop3 and Aaop7 groups were composed of three photoreceptors each in the satellite stemma, and the Aaop7 group was composed of three photoreceptors in the central stemma. However, as it was not always possible to reliably count the cell bodies, some animals may not have exactly these numbers of Aaop3- and Aaop7-expressing photoreceptors.

We then examined whether Aaop7 also shows light-driven motility by examining the satellite and central stemmata in light- and dark-treated mosquitoes. In z-series confocal reconstructions of the satellite stemma, Aaop7 (green) was detected only within the actin-rich (red) rhabdoms of dark-treated photoreceptors (arrow, Fig. 3C). Upon light treatment (Fig. 3D), some Aaop7 was detected within cytoplasmic vesicles (arrow) of the photoreceptor cell bodies (CB), but the majority of Aaop7 remained within the rhabdom. In a similar analysis of the central stemma, Aaop7 was localized to the rhabdom in dark conditions (Fig. 3E), and showed limited movement to the photoreceptor cell bodies (Fig. 3F) when treated by light. These results indicate that Aaop7 exhibits less light-triggered movement than Aaop3.

Comparative sequence analysis places Aaop3 and Aaop7 in the group of long-wavelength rhodopsins (Nene et al., 2007). To experimentally validate this characterization, we performed an electroretinogram (ERG) analysis on transgenic Drosophila expressing these genes. Aaop3 transgenic Drosophila failed to give an ERG response, probably because of poor Aaop3 expression and an inability of Aaop3 to localize to Drosophila photoreceptor rhabdomeres (Fig. 2A,B). By contrast, ERG responses were consistently recorded from the Aaop7 transgenics. Fig. 4A shows a representative ERG trace from the Aaop7 transgenic flies, and Fig. 4B shows a compilation of the data obtained from four individual flies. The results show that the ERG response to photon-matched stimuli was greatest at a wavelength of 450 nm. However, the Aaop7 transgenic fly retains high sensitivity to shorter wavelengths (400 and 350 nm) and does not exhibit the second peak response in the UV observed for Drosophila Rh1 (Fig. 4C,D) and Aaop2, the only other long-wavelength Aedes opsin that has been analyzed in this way (Hu et al., 2011).

The mosquito adult compound eye develops from a placode immediately anterior to the larval stemmata. This developmental process is already in progress during the 1st instar larva, and the 1st ommatidial units are fully differentiated by the late 3rd instar (White, 1961). New ommatidial units continue to be added until the pupa is 24 h old (Sato, 1953). A larval whole-mount preparation allowed us to investigate the timing of rhodopsin expression in the developing compound eye.

Fig. 5A shows the retinal region of a 4th instar larva. The expansion of the developing adult eye in the anterior direction is marked by the line of three arrows. Aaop7-expressing ommatidia (green) are present near the anterior edge of the developing compound eye (Fig. 5A, asterisk). In this preparation, we were not able to determine whether Aaop7 is restricted to a particular class of photoreceptors within the developing ommatidium. However, we also detected Aaop7 at lower levels in mature ommatidia bordering the larval stemmata (Fig. 5B). In these mature ommatidia, Aaop7 expression is restricted to the R8 photoreceptor positioned centrally within the fused rhabdom structure. Aaop7 is also detected within the axonal projections from photoreceptors of the developing compound eye and larval stemmata (arrows, Fig. 5C).

Aaop1 is the major rhodopsin of the adult compound eye because it is expressed in all R1–R6 photoreceptors and most R8 photoreceptors (Hu et al., 2012). We made use of a larval whole-mount preparation to determine the onset of Aaop1 rhodopsin expression in the developing compound eye. In 4th instar larva, Aaop1 expression is readily detected in the posterior mature ommatidia of the developing compound eye (Fig. 5D). The magnified image (Fig. 5E) shows Aaop1 expression within the fused rhabdom and also within the cell bodies surrounding the fused rhabdom (arrow). Given that the R1–R6 photoreceptors are positioned in this peripheral location (Hu et al., 2012), these results indicate that the initiation of Aaop1 expression within the R1–R6 photoreceptors occurs during the 4th instar larval stage.

In this report, we characterized the expression pattern and light-triggered movement of rhodopsins in Aedes larval photoreceptors. These photoreceptors are positioned laterally on the head and organized into five stemmata. The majority of stemmatal photoreceptor cells express the long-wavelength Aaop3 rhodopsin. In a lighted environment, Aaop3 localizes to cytoplasmic vesicles in the photoreceptor cell body, whereas under dark conditions, Aaop3 moves into the rhabdom. This extensive light–dark translocation of a rhodopsin was previously observed for Aaop1 in adult Aedes mosquitoes and it is proposed to be an adaptation mechanism for maximizing the light sensitivity range of photoreceptors (Hu et al., 2012).

The pioneering research of White (1967) used the Aedes larval visual system to identify a daily cycle in which rhabdoms are lost at dawn and regenerated at dusk. Membrane loss during the dawn transition reduces the rhabdom surface area by 50–70%. This reduction is accompanied by the formation of large numbers of cytoplasmic multivesicular bodies (White, 1968). Subsequently, similar light-triggered loss of rhabdoms or rhabdomeres was documented in adult photoreceptors of other invertebrates, including crabs (Nässel and Waterman, 1979), spiders (Blest et al., 1978) and tipilid flies (Williams and Blest, 1980). More recent studies revealed an extensive redistribution of rhodopsin during the membrane cycling process. In the horseshoe crab Limulus polyphemus, the 10% loss of membrane volume at dawn (Sacunas et al., 2002) is accompanied by a 50% loss of rhodopsin content (Katti et al., 2010). The adult Aedes photoreceptors exhibit greater changes at dawn, where a 50% loss of membrane volume (Brammer et al., 1978) is accompanied by an almost 100% loss of rhodopsin content (Hu et al., 2012). Thus, it is likely that rhodopsin movement is the trigger for membrane renewal and that this process occurs to different extents in different species. The identification of Aaop3 as the larval stemmatal rhodopsin involved in the extensive light-driven membrane renewal makes the Aedes larva an excellent experimental system to further analyze this process.

Two separate groups of three photoreceptors within the central and satellite stemmata express the Aaop7 rhodopsin instead of Aaop3. Direct ERG recordings of the Aedes stemmata established that its peak spectral sensitivity is at 520 nm (Seldin et al., 1972). Microspectrophotometric analysis further showed that the rhodopsins of the stemmata together have an absorbance spectral peak at 515 nm (Brown and White, 1972). Both values are consistent with what would be expected upon expression of Aaop3 and Aaop7, as sequence comparisons classify these as long-wavelength rhodopsins (Nene et al., 2007). We attempted to individually assess the properties of Aaop7 and Aaop3 by expressing each of these rhodopsins in transgenic Drosophila. We found that Aaop7 showed a peak sensitivity near 450 nm. However, the major stemmatal rhodopsin Aaop3 was not sufficiently expressed in transgenic Drosophila to be analyzed in this fashion. We reasoned that because Aaop3 is the major larval rhodopsin, the stemmatal 520 nm peak sensitivity observed in the earlier studies would largely be the result of the Aaop3 spectral properties. Aaop7, then, is shifted towards shorter wavelengths. Therefore, the presence of two classes of photoreceptors expressing either Aaop3 or Aaop7 might provide some color discrimination within the blue–green region of the visible light spectrum.

The existence of two classes of Aedes stemmatal photoreceptors distinguished by rhodopsin expression is similar to the situation in Drosophila. In the Drosophila larval visual system, known as Bolwig's organ, one class consists of three or four photoreceptors that arise from primary precursor ‘founder cells’ (Suzuki and Saigo, 2000). These photoreceptors express the rhodopsin Rh5 (Sprecher et al., 2007). During embryogenesis, these founder cells recruit secondary precursors that develop into a second class of photoreceptors expressing the rhodopsin Rh6 (Sprecher et al., 2007). In Aedes, it is possible that there is a conserved developmental pathway in which Aaop7 precursor cells recruit the Aaop3-expressing photoreceptor cells. However, Aedes developmental processes ultimately result in five stemmatal units of different sizes and only two of these contain Aaop7-expressing cells. These considerations indicate that the developmental steps specifying the mosquito larval eye are more complex than the Drosophila larval eye.

The adult compound eye rudiment is first detected during the 1st instar larval stage. It is evident as a thickening of the epidermis, referred to as the optical placode, immediately anterior to the stemmata (Sato, 1953; White, 1961). Photoreceptors are born during a wave of differentiation starting at the posterior edge, closest to the larval stemmata, and then moving anteriorly. As a result, the posterior regions contain more mature ommatidial units than the anterior regions. The process appears similar to the extensively studied morphogenesis of the adult Drosophila eye (Treisman, 2013). During ommatidial differentiation, Aaop7 is briefly expressed in ommatidial units at the anterior edge of developing compound eye. Earlier in the discussion, we hypothesized that the Aaop7-expressing cells may be the founder cells for construction of stemmatal units. The transient expression of Aaop7 in adult ommatidial units could also play a developmental role. In this regard, it is noteworthy that Aaop7 appears to localize to the axons of both founder stemmatal cells and founder R8 cells of the developing adult ommatidia. The axons of these photoreceptors track together into the optic lobe, raising a possibility that Aaop7 expression plays a role in axonal pathfinding or related developmental process.

In the Aedes compound eye, differentiated photoreceptors are present in the most-posterior ommatidia by the middle of the 4th instar larva. The anterior edge of the eye reaches maturity at the 24-h-old pupal stage (Sato, 1953). Axonal projections from the adult photoreceptors during this same time period are accompanied by development and organization of the lamina and medulla optic lobe neuropil (Mysore et al., 2014). In the swimming larva and pupa of Aedes and other mosquitoes, a shadow moving across the visual field triggers an avoidance response consisting of a downward movement from the surface of the water (Kasap, 1978). To distinguish between the role of the larval stemmata and the adult compound eye in this behavior, Kasap (1977a) used black paint to cover the stemmata, the compound eye or both. In Aedes 4th instar larvae and pupae, only painting over both the stemmata and the compound eye resulted in statistically significant defects in the light-triggered behavior. These results indicate that photoreceptors of the compound eye may begin to augment the light-sensing capabilities of the larval stemmata during the 4th instar stage. Our observation that Aaop1 is already expressed in the R1–R6 photoreceptors at this time is consistent with this view.

The Liverpool strain of Aedes aegypti was used. Aedes aegypti rhodopsins referred to as Aaop1, Aaop3 and Aaop7 correspond to GPROP1 (AAEL006498), GPROP3 (AAEL006484) and GPROP7 (AAEL007389) of the Aedes aegypti genome project (Nene et al., 2007, https://www.vectorbase.org). To create the Aaop3- and Aaop7-expressing Drosophila transgenic lines, their open reading frames were retrieved from genomic DNA by PCR, P-element transformation plasmids were created, and germline transformation was carried out as previously described for other mosquito rhodopsins (Hu et al., 2009, 2011).

Larval heads were removed and fixed for 4–6 h at 4°C in 2% paraformaldehyde (Electron Microscopy Sciences, stock 16% solution) in phosphate-buffered saline (PBS). Collection of heads in the dark was performed under dim red light using a 650 nm long-pass filter. Heads were then transferred to PBS and kept at 4°C for up to 5 days. Larval stemmata were dissected in PBS and incubated in primary antibody [anti-Aaop1, anti-Aaop3 C-terminus, anti-Aaop3 N-terminus, or anti-Aaop7 N-terminus antisera, generated as described previously (Hu et al., 2011)] diluted 1:100 in BNT (1× PBS, 0.1% BSA, 0.1% Tween-20 and 250 mmol l⁻¹ NaCl) overnight at 4°C. After three 10 min washes in PBS, the stemmata were incubated in secondary antibody (Alexa-Fluor-488-conjugated goat anti-rabbit IgG, 1:400 in BNT) and Alexa-Fluor-594-conjugated phalloidin (1:40 in BNT) for 4 h at room temperature. After three 10 min washes in PBS, stemmata were mounted in 10 µl of Vectashield (Vector Laboratories, Burlingame, CA). Tape strips mounted on glass ring slides were used to hold the coverslip above the stemmatal tissue to prevent damage to the tissue. Stemmata were visualized with a Nikon A1R confocal microscope, and brightness and contrast on the resulting photomicrographs were uniformly adjusted using Photoshop CS5 software.

White-eyed Drosophila heads were removed and placed in fixative (4% paraformaldehyde with 5% sucrose) overnight at 4°C. The heads were then washed three times for 10 min each time in a solution of 5% sucrose in 1× PBS and incubated overnight at 4°C. They were then incubated in 30% sucrose in 1× PBS solution overnight at 4°C, and then placed in 30% sucrose in 1× PBS and Tissue Freezing Medium (TFM, Triangle Biomedical Sciences, Cincinnati, OH, USA) mixed 1:1. The heads were embedded in TFM and sectioned at 10 µm at −27°C. The sections were washed in 1× PBS for 20 min and incubated in blocking buffer (1× PBS, 5% normal goat serum, 0.3% Triton X-100 and 1% dimethylsulfoxide) for 1 h at room temperature. Sections were incubated overnight at 4°C in primary antibody against Aaop1, Aaop3 or Aaop7 diluted 1:100 in blocking buffer. After three 10 min washes in 1× PBS with 0.1% Tween-20, the sections were incubated in secondary antibody (Alexa-Fluor-488-conjugated goat anti-rabbit IgG, 1:400 in BNT) and Alexa-Fluor-594-conjugated phalloidin (1:40 in BNT) for 1 h at room temperature. After three 10 min washes in 1× PBS with 0.1% Tween-20 and one wash in PBS for 5 min, sections were mounted in Vectashield (Molecular Probes). Cryosections were visualized using a Leica DM5000 fluorescence microscope. The photomicrographs were uniformly processed with Photoshop CS5 for brightness and contrast.

Aaop3 and Aaop7 transgenic Drosophila were placed in genetic backgrounds to create white-eyed flies with only the Aaop7 or the Drosophila Rh1 rhodopsin expressed in the R1–R6 photoreceptors and with no responses from the R7 and R8 photoreceptors. Therefore only the activity of Aaop7 or the Rh1 control rhodopsin was represented in the ERG response (Ahmad et al., 2006). The Aaop7- and the Rh1-expressing Drosophila were then subjected to ERG analysis as described previously (Hu et al., 2014). Narrow band pass and neutral density filters (Oriel, Stratford, CT) were used to attenuate a 1000 W tungsten light source (Oriel, Irvine, CA) so that the flies could be exposed to light stimuli of comparable photon content (20 μE m⁻² s⁻¹) at 600 nm, 550 nm, 500 nm, 450 nm, 400 nm and 350 nm wavelength.

Confocal microscopy was conducted within the Notre Dame Integrated Imaging Facility.