Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Jan 10.
Published in final edited form as: Curr Biol. 2011 Dec 15;22(1):12–20. doi: 10.1016/j.cub.2011.11.028

Genetic dissection reveals two separate retinal substrates for polarization vision in Drosophila

Mathias F Wernet 1, Mariel M Velez 1, Damon A Clark 1, Franziska Baumann-Klausener 2, Julian R Brown 3, Martha Klovstad 1, Thomas Labhart 2, Thomas R Clandinin 1,#
PMCID: PMC3258365  NIHMSID: NIHMS339789  PMID: 22177904

SUMMARY

Background

Linearly polarized light originates from atmospheric scattering, or surface reflections, and is perceived by, insects, spiders, cephalopods crustaceans and some vertebrates. Thus, the neural basis underlying how this fundamental quality of light is detected is of broad interest. Morphologically unique, polarization-sensitive ommatidia exist in the dorsal periphery of many insect retinas, forming the ‘Dorsal Rim Area’ (DRA). However, much less is known about the retinal substrates of behavioral responses to polarized reflections.

Summary

Drosophila exhibits polarotactic behavior, spontaneously aligning with the e-vector of linearly polarized light, when stimuli are presented either dorsally or ventrally. By combining behavioral experiments with genetic dissection and ultrastructural analyses, we show that distinct photoreceptors mediate the two behaviors: inner photoreceptors R7+R8 of DRA ommatidia are necessary and sufficient for dorsal polarotaxis, whereas ventral responses are mediated by combinations of outer and inner photoreceptors, both of which manifest previously unknown features that render them polarization-sensitive.

Conclusions

Drosophila uses separate retinal pathways for the detection of linearly polarized light emanating from the sky, or from shiny surfaces. This work establishes a behavioral paradigm that will enable genetic dissection of the circuits underlying polarization vision.

INTRODUCTION

Linearly polarized skylight created by atmospheric scattering of sunlight, is perceived by many animals [1, 2, 3], and serves as an important navigational cue [4, 5]. Sunlight reflecting off shiny surfaces, such as leaves and water, is also linearly polarized [1, 6], and represents another environmental signal [7, 8, 9]. Behavioral, electrophysiological and anatomical studies in many insects have identified specialized ommatidia in the ‘Dorsal Rim Area’ (DRA) of the compound eye as the most suitable candidate for detecting polarized skylight [10, 11, 12]. In these ommatidia, two photoreceptors maintain polarization sensitivity (PS) by failing to twist their rhabdomeres [for review: 12]. By comparison, much less is known about how insects detect polarized reflections. Behavioral studies in water bugs, dragonflies, locusts, and tabanid flies have demonstrated that polarized light can be detected by the ventral eye [7, 8, 9]. While a likely retinal substrate has been described in the backswimmer Notonecta [6], the functional relationship between specific photoreceptors and these cues have not been demonstrated. Thus, understanding the cellular and behavioral relationship between dorsal and ventral polarization signals presents an important challenge.

The Drosophila eye comprises ~800 ommatidia, each containing eight photoreceptor cells, designated R1–R8. ‘Outer photoreceptors’, R1–R6, contain a blue/green-sensitive rhodopsin Rh1, associated with a UV-sensitizing pigment that confers response to UV light [13]. Variations in inner photoreceptors create a mosaic of at least three subtypes ([14]). DRA ommatidia form a narrow band of 1–2 rows along the dorsal margin of the eye [15]. In these ommatidia, R7 and R8 have enlarged rhabdomeres, and express the UV-sensitive pigment Rh3 [16, 17]. The two remaining subtypes are named ‘pale’ (p) and ‘yellow’ (y), and are randomly distributed across the retina [14]. R7 cells each express one of two UV opsins rh3 (R7p), or rh4 (R7y), while the underlying R8 cells express either rh5 (R8p) or rh6 (R8y). Due to this chromatic heterogeneity, inner photoreceptors are thought to mediate color vision [18, 19, 20].

Genetic tools provide powerful approaches to dissecting neural circuits underlying visual behaviors in Drosophila [19, 20, 21, 22]. However, polarization vision is poorly understood in flies, since two previous studies implicated different retinal substrates. Von Philipsborn and Labhart [23] reported spontaneous turning responses of houseflies to slowly changing e-vector orientations, a behavior that was UV-specific and proposed to be mediated by DRA ommatidia. These findings agreed with electrophysiological and morphological studies demonstrating high PS in R7DRA and R8DRA photoreceptors [13, 24]. However, Wolf et al. [25], demonstrated alignment of Drosophila with the incident e-vector, a behavior that was elicited by both polarized UV and green light, even when presented ventrally, which they linked to R1–R6 photoreceptors. Here we establish a new behavioral paradigm, and use genetic tools to define the retinal substrate of polarization vision in Drosophila.

RESULTS

Drosophila manifests orientation responses to linearly polarized stimuli presented either dorsally, or ventrally

Using a custom tracking system [22], we monitored the movements of isogenic fly populations in a circular arena (Ø 7.5 cm × height 2.5 cm) illuminated from above by linearly polarized (POL) light (Figure 1A, Figure S1). Flies could freely walk on the transparent floor or ceiling of the arena, with either the dorsal or ventral eye seeing the stimulus. A polarizer, mounted on the motorized stage, rotated in 45° steps, remaining stopped for 5 seconds (Figure 1B). Flies were recorded from below using an IR video camera, and the position and orientation of each fly was correlated with e-vector orientation during the stops. Polar histograms of fly angular headings during the stopped epochs suggested that flies preferentially aligned their body axis in parallel with the e-vector (Figure 1C). Plotting these histograms on a linear axis over 360° revealed a sinusoidal modulation of orientation whose amplitude was proportional to the strength of the response, and whose phase captured its precision. To represent this polarotactic behavior in a single metric, we computed an ‘Alignment value’, A, incorporating both amplitude and phase of this distribution (see Experimental Procedures).

Figure 1. Drosophila manifests orientation responses to linearly polarized stimuli presented either dorsally, or ventrally.

Figure 1

Open in a new tab

A. Schematic of the experimental setup used to present linearly polarized UV light from above to populations of Drosophila, which were filmed in the infrared from below. A polarization filter (HN42HE) was facing the flies, with 2 sheets of diffuser paper facing the light source. IR = infrared light. UV POL = polarized UV light. Diff = Diffuser. Pol = Polarizer. B. Summary of the stimulus protocol used. A computer-controlled servomotor rotated the polarization filter in 45° increments, remaining still for 5 seconds at each position. Different motor positions are denoted with different colors. C. Polar histograms of fly angular headings are shown for dorsally stimulated flies at each motor position. D. Basic description of wild type polarotactic responses for linearly polarized UV stimulus presented dorsally. White bars symbolize UV-POL stimulation, green and blue bars stimulation with polarized light of the respective color (see methods). All error bars = +/− 1 S.E.M. *=p<0.05, **p<0.01, ***p<0.001. n.s. = not significant. E. Alignment values A, plotted as a function of dorsal UV stimulus intensity. Dashed line: intensity setting used for all subsequent experiments. Red box: UV intensity of skylight at dusk (Palo Alto, CA – see Supplemental Experimental Procedures). F. Basic description of wild type responses for linearly polarized UV stimulus presented ventrally. G. A values plotted as a function of ventral UV stimulus intensity.

Both male and female flies aligned to the e-vector of a dorsal UV stimulus (A♂=0.13±0.02 and A♀=0.15±0.0; Figure 1D), across a range of ethologically relevant intensities (see Supplemental Experimental Procedures). This response was lost in complete darkness (A = 0.00±0.01), or when the light was depolarized, by a diffuser (A = 0.02±0.01). Polarotactic responses were virtually lost (A=0.04±0.01) when a quarter wave plate (QWP) was positioned in front of the polarizer at an orientation that transformed the stimulus into circularly polarized light, which insects perceive as unpolarized (see Experimental Procedures, [26]). The responses could be restored by rotating the QWP 45° with respect to the polarizer, restoring linear polarization (A=0.13±0.01). Finally, dorsally stimulated flies did not orient to blue POL (460±10 nm) or green POL (510±10 nm) stimuli (ABlue=0.04±0.01 and AGreen=0.00±0.01). Thus, the photoreceptors that mediate dorsal POL behavior are strictly UV sensitive and detect the linearly polarized component of the stimulus.

In agreement with previous work demonstrating that Drosophila can perceive polarized light ventrally [25], both male and female flies displayed preferential alignment in parallel with the e-vector when seeing the polarized UV stimulus with the ventral half of their eyes (Figure 1F; A=0.22±0.01 and A=0.23±0.01, respectively), a response that was never detected in darkness (A=0.01±0.026). Ventral POL responses were significantly stronger than dorsal POL responses, and remained robust down to low light levels (Figure 1G). Depolarizing the stimulus strongly abrogated the response (A=0.04±0.01), as did the QWP (A=0.10±0.01), and again the response could be rescued by rotating the QWP 45 degrees (A=0.26±0.01). Robust ventral POL responses were also obtained using blue (460 nm) and green (510 nm) polarized light (A=0.19±0.02 and A=0.14±0.02, respectively), at the same intensities that failed to evoke responses when presented dorsally. Thus, the spectral sensitivity of ventral POL behavior is different, extending to longer wavelengths.

Dorsal polarotactic behavior is mediated by the ‘Dorsal Rim Area’

To determine the necessity of different photoreceptor classes for dorsal POL vision we disrupted synaptic transmission through expression of shibirets, a temperature-sensitive, dominant negative mutant of dynamin ([27], Figure S2). While expression in some photoreceptor subtypes non-specifically reduced behavioral responses by less than 50% (Figure 2A), only inactivation using rh3-GAL4 (expressed in R7p and DRA inner photoreceptors) completely abolished polarotactic responses (A=0.02±0.01). Furthermore, dorsal POL behavior was completely lost upon photoreceptor inactivation using hth-GAL4 (expressed in R7DRA and R8DRA; A = 0.03±0.01), as well as rh6+DRA-GAL4 (A=0.01±0.005). In this driver, a point-mutation introduced into the rh6 promoter sequence leads to expression of GAL4 in R8y as well as R7DRA and R8DRA (Figure S2). Thus, DRA ommatidia provide the retinal substrate of dorsal POL vision. However, we could not rule out contributions of other photoreceptor classes.

Figure 2. Dorsal polarotactic behavior is mediated by the ‘Dorsal Rim Area’.

Figure 2

Open in a new tab

A. Testing the necessity of photoreceptor subtypes mediating behavioral responses to UV-POL stimuli presented dorsally. Polarotactic responses were measured in flies expressing UAS-shibirets under the control of GAL4 drivers expressed in various subtypes of photoreceptors. Unlabeled bars: not significantly different from the control. B. Sufficiency of photoreceptor subtypes mediating behavioral responses to POL stimuli presented dorsally. Opsin drivers (both wild type and mutated) and hth-GAL4 were used to rescue photoreceptor function by expressing eye-specific Phospholipase C (NorpA) from newly generated UAS-norpA transgenes (shown schematically, see methods), in norpA−/norpA− mutant flies. Open bars denote experimental genotypes, gray bars denote negative controls (a norpA/norpA mutant, bearing UAS-norpA, without a GAL4 driver).

To test for sufficient photoreceptor classes, we functionally rescued the phototransduction mutant norpA [28], using GAL4 drivers. As expected, norpA mutants were blind (A=0.00±0.001, Figure 2B). This defect was specifically rescued by expressing UAS-norpA using rh3-GAL4 (A=0.10±0.01), or rh6+DRA-GAL4 (A=0.11±0.02), but by none of the other opsin drivers. Rescue of different photoreceptor subclasses in addition to rh3-expressing cells (rh1+rh3, rh3+rh4, rh3+rh5, rh3+rh6) never led to A values significantly higher than rh3>norpA alone. Although hth-GAL4 did not rescue, this driver is only weakly expressed in the adult retina. Therefore, specifically restoring function to both R7DRA and R8DRA is sufficient to restore dorsal POL behavior.

The rhabdomeric photoreceptors of insects are inherently polarization-sensitive since the rhodopsin molecules are aligned within the microvillar membrane so that linearly polarized light is maximally absorbed when the e-vector orientation is parallel to the microvilli [11, 29, 30]. Hence, polarization sensitivity is maximal when the microvilli are well aligned along the rhabomere [31, 32]. However, rhabdomeres are generally twisted in flies [33, 34, 35, 36]. We therefore assessed rhabdomere twist of R7DRA and R8DRA (Figure 3) by measuring microvilli orientation in serial electron microscopic cross-sections. R7DRA and R8DRA were easily identifiable by their enlarged rhabdomeres (compare Figure 3A with Figure 3D, [15, 17]), and displayed strongly reduced twist when compared to non-DRA ommatidia (Figure 3A,B,C,E). Based on the twist functions we estimated their polarization sensitivity as PSR7.DRA = 8.1±0.6 (n=8) and PSR8.DRA = 7.9±1.1 (n=7) (see Supplemental Experimental Procedures). The e-vector orientations of maximal sensitivity (ϕmax) of R7 and R8 were approximately orthogonal to each other (82°±8°). Thus, R7DRA and R8DRA have polarization sensitivity characteristics appropriate to an orthogonal analyzer system as previously described [2]. In contrast, R7 and R8 immediately adjacent to the DRA displayed considerable twisting (Figure 3B, C, F), resulting in lower estimated PS values (PSR7 = 3.6±1.2, n = 7; PSR8 = 2.2±0.5, n=8). R7 and R8 rhabdomeres at the ventral eye rim (VR) also exhibited significant twist and comparatively low estimated PS values (PSR7.VR = 2.6±0.89, PSR8.VR = 2.2±0.5, n=5), consistent with the absence of a specialized ventral rim area.

Figure 3. Rhabdomeres of inner photoreceptors in the DRA are untwisted.

Figure 3

Open in a new tab

Ommadidia of the Dorsal Rim Area (DRA) and of the adjacent dorsal area (DA) were studied A, D Transmission electron micrographs showing rhabdom structure (a) and microvilli orientation of R7 in individual ommatidia at distal (b) and proximal (c) levels of R7. Numbers indicate receptor types. Straight lines in rhabdomere cross sections give microvilli orientations. Calibration bars 1 μm. B, C Total range of microvilli directions expressed by fans in different groups of ommatidia (same as in line graphs E, F). Red fans represent R7, blue fans R8. Fine black line marks the boundary between the DRA and the DA. Fat black line shows the eye rim. Interrupted arrowed line is the v-axis of the ommatidial pattern pointing dorsal (compare Figure S3D). dco: dorso-caudal origin of ommatidial rows. Calibration bars 10 μm. Note that one ommatidium has a R7DRA (large, non-twisting rhabdomere) but a R8DA (small, twisting rhabdomere). E–F Graphic representation of microvilli orientation at different retinal levels (twist functions) in R7 (left family of curves) and R8 (right family of curves). The ordinate indicates microvilli orientation relative to a straight line through the centers of R1 and R3 rhabdomeres (0°; stipled line). The abscissa gives retinal level relative to the surface of the eye (0 μm indicates level of first section containing rhabdoms). Colors mark data from different, identified ommatidia.

Low twist R7 photoreceptors in the ventral eye can mediate polarotactic responses

We next assessed which photoreceptor classes are necessary for the ventral UV-POL response, using shibirets (Figure 4A). Of the single opsin drivers, only rh3-GAL4 caused a significant response decrease (A=0.11±0.03). Inactivation of R7DRA and R8DRA using hth-GAL4, and rh6+DRA-GAL4 had no effect (A=0.30±0.01 and A=0.29±0.02, respectively). Thus, the DRA is not required for the ventral POL response. Rather, this behavior depends on UV-sensitive rh3-expressing ventral R7p cells.

Figure 4. Ventral R7 photoreceptors can mediate polarotactic responses.

Figure 4

Open in a new tab

A. Behavioral responses to a ventral POL stimulus after inactivation of photoreceptor subtypes with two copies of UAS-shibirets. Open bars denote experimental genotypes, gray bars denote control genotypes. B. No single photoreceptor subtype is required for an orientation response of upside-down walking flies to linearly polarized green light, but behavior gets strongly abrogated upon simultaneous inactivation of rh1- and [rh5+rh6] subtypes, and completely disappears using 3 copies of UAS-shibirets (compare dark and light green bars).

Ventral POL responses to green light cannot be mediated by the exclusively UV-sensitive R7. In fact, inactivation of the three main photoreceptor classes by themselves (R1–R6, R7, or R8 cells) did not significantly affect the Green-POL response (Figure 4B), which was only abrogated by a combination of rh1+[rh5+rh6]-GAL4 drivers (A=0.02±0.02). We infer that R1–R6 and R8, but not R7, are redundantly required for the response to green light (510±10 nm). Hence, changing the stimulus wavelength shifted the retinal inputs to the POL vision circuitry.

To examine how the ventral retina mediates polarotactic responses, we estimated PS of ventral photoreceptors, by characterizing their rhabdomere twist (Figure 5). Since our behavioral data suggest a prominent role of R7p, we first compared rhabdomeric twist of R7 and R8 subtypes, after specifically labeling p ommatidia, using rh3-GAL4 and UAS-CD2:HRP (see Supplemental Experimental Procedures; Figure S3A,B). Specific differences in rhabdomeric twist between p and y ommatidia have been described for Calliphora R8 cells [33]. However, analysis of three patches of ventral retina revealed that R7p and R7y as well as R8p and R8y were equally twisted (Figure S3C). We now broadly searched the ventral retina, by analyzing rhabdomeric twist of R7 in seven, partly overlapping groups of 6 to 25 ventral ommatidia (Figure S3D). In four groups, R7 cells were significantly twisted, with an average estimated PSR7 ranging from 2.2 to 2.9 (Figure 5G; 2.2±0.8, n=11; 2.5±0.6, n=6; 2.9±0.8, n=11, 2.6±0.9, n=25). In two groups, twisting was restricted to the proximal half of the R7 rhabdomeres (Figure 5I) resulting in estimated PS of >3 in one group (PS R7 =3.1±0.4, n=10), and reaching 5 in the other group (PS R7 = 5.0±1.0, n=7). The last group (Figure 5D,E,H) contained several ommatidia with low twist (13° to 26°), resulting in high estimated PS values of 6.4 to 8.0 (average PSR7 of the group was 5.5±1.6, n=13). Comparison of twist functions between overlapping ommatidia from three different individuals (VA2, VA5, VA6; Figure S3D) showed considerable differences in estimated PS, arguing against a precisely positioned ventral POL area. However, in these ommatidia, the ϕmax orientations of R7 were strongly aligned, showing variations of only ±5.6° to ±13.6° (circular s.d.). Thus, even in the more twisted rhabdomeres, the microvilli orientations still had a strong directional bias, resulting in significant estimated PS values. Hence, by pooling the responses of neighboring ommatidia even moderately polarization-sensitive R7 cells can provide reliable e-vector information.

Figure 5. Moderate- and low-twist R7 cells exist in the ventral eye.

Figure 5

Open in a new tab

Three different groups of ommatidia in the ventral eye (VA1, VA2, VA3). Rhabdomeres generally twist but the amount of twist and the shape of the twist functions differed between groups. A, D as in Figure 2 A, D. Calibration bars 1 μm. B, C, E, F as in Figure 2 B, C (same ommatidia as in line graphs G,H). White asterisks on some fans in C indicate that data for the most proximal rhabdomere are missing. vfo ventro-frontal, vco ventro-caudal origin of ommatidial rows. Calibration bars 10 μm. G–I as in Figure 2 E, F.

The R8 cells in all ommatidial groups had strong rhabdomeric twist (Figure 5C,F,G,H,I) and low estimated PS values with averages ranging from 1.5 to 2.2. Only in VA2, the group exhibiting strong estimated R7 PS (Figure 5F,H), a few R8 rhabdomeres were extremely short (10–20 μm) resulting in small net twist and correspondingly high estimated PS reaching 4 to 7 in 5 out of the 13 R8 cells.

Outer photoreceptors R1–R6 contribute to ventral polarotactic responses

Using a ventral UV-POL stimulus, we assessed sufficiency of photoreceptor classes using UAS-norpA (Figure 6A). Only weak rescue was obtained when norpA was expressed in any of rh1-, rh3-, or rh5-positive cells (A=0.09±0.02, A=0.07±0.01, and A=0.08±0.02, respectively). However, rh1+rh3 together rescued ventral POL responses to wild type levels (A=0.29±0.02). Other combinations did not show this effect, and hth-GAL4 and rh6+DRA-GAL4 failed to rescue ventral POL behavior. Thus, specific synergy between R1-R6 and rh3-expressing R7p cells is required for a robust ventral UV POL response.

Figure 6. Outer photoreceptors R1–R6 contribute to ventral polarotaxis.

Figure 6

Open in a new tab

A. NorpA rescue experiments (open bars, and shaded bar control) were used to define photoreceptor sub-type sufficiency for behavioral responses to polarized UV light. B. Orienting responses to polarized green light in norpA rescue experiments. Alignment responses upon outer photoreceptors rescue (rh1-norpA) were eliminated by the QWP, and restored by rotating it 45°. C. Twist functions of receptors R1, R2, R4 and R5. Typical twist functions of each cell type in three ommatidia are shown. The twist functions of R4 and R5 are generally flatter than those of R1 and R2. D. Polarization sensitivity (PS) of receptor types R1–8. PS of samples of 4, 5 and 8 ommatidia in three different eyes are shown. Black circles indicate average PS. Note that R4, R5 and R6 have higher PS than R1–3. E. Model summarizing photoreceptor contributions to linearly polarized stimuli presented to the dorsal, or the ventral retina, respectively. Left: insects encounter linearly polarized light originating from atmospheric scattering, or from reflections off of shiny surfaces such as water. Middle: schematic representation of the dorsal half (top) or ventral half (bottom) of the fly retina (necessary ommatidia are labeled red), followed by a schematic representation of photoreceptor classes in these ommatidia (photoreceptors that provide input to UV polarotaxis, green polarotaxis, or both behaviors are shown in violet, green, and blue, respectively). Right: photoreceptor types providing behavioral contributions. Behavioral output is symbolized by a sinusoid function, synergistic interactions between photoreceptor subtypes are symbolized by a ‘+’ sign.

In contrast, under a green POL stimulus (Figure 6B), robust behavior was observed upon rescue of R1–R6 function (A=0.24±0.03). The ‘Quarter Wave Plate’ (QWP) abrogated the behavioral response, both for wild type, as well as rh1-norpA rescued animals, and rotation of the QWP by 45 degrees again restored behavior. As expected, rescue of R7 cells, which cannot detect green light, was never sufficient (A=0.00±0.03). To our surprise, rescue of either R8 photoreceptor subtypes was sufficient to mediate ventral polarotaxis (AR8p = A=0.19±0.01; AR8y = 0.10±0.01). Thus R1–R6, and R8 cells, are sufficient to mediate responses to polarized green light presented ventrally.

To estimate PS of ventral R1–R6, we measured their twist functions (Figure 6C) in 17 ommatidia from three ventral groups. While average estimated PS in R1–3 was <2 (PSR1 = 1.7±0.4, PSR2 = 1.7±0.4, PSR3 = 1.8±0.6, n=17), PS in R4–6 was enhanced (PSR4 = 3.0±0.9, PSR5 = 2.8±0.7, PSR6 = 2.4±0.4, n=17). Thus ventral R4 to R6 cells with their reduced twist can serve as an additional retinal substrate for ventral polarotaxis.

DISCUSSION

We define the retinal substrates for both dorsal and ventral polarization vision in Drosophila. The DRA is necessary and sufficient for dorsal polarotactic responses, a result that strengthens studies in other insects concluding that this region mediates responses to celestial polarized light [37, 38, 39]. In addition, our work defines the retinal substrate for responses to ventral polarotactic stimuli, as would occur naturally by reflections from shiny surfaces like water or leaves. Our work resolves the differences between previous behavioral studies of polarotactic behavior in flies [23, 25] by demonstrating that flies possess separate detectors to respond to distinct wavelengths, and sources, of polarized light.

A ventral POL region has previously been described in the backswimmer Notonecta, which uses polarized reflections to locate water bodies [6]. In this insect, inner photoreceptors in a small ventral region form orthogonal analyzer pairs with untwisted rhabdomeres much like a DRA [6]. Drosophila uses a different strategy by exploiting the fact that photoreceptors with moderate or weak twist still provide enough PS to serve as polarization analyzers. In this way other visual senses, such as the detection of motion and spectral cues, should be affected only minimally by the polarization of light. Hence, unlike Notonecta with its specialized ventral retina, the generalist Drosophila incorporates ventral POL detectors whilst preserving other critical visual capacities.

An interesting feature of this design is that different classes of photoreceptors form ventral POL analyzers depending on stimulus wavelength. In the UV range, R7p cells are necessary for normal polarotactic responses; correspondingly, we describe ventral R7 cells with moderate to high estimated PS. However, our sufficiency experiments also revealed the involvement of outer photoreceptors in polarization vision. While R1–R3 appear to be weakly polarization-sensitive, R4, R5 and possibly R6, show pronounced estimated PS due to reduced rhabdomeric twist. These results are consistent with intracellular recordings in Calliphora describing two classes of R1–R6, one of which retains some PS, even in the UV [40].

While R4 to R6 with their pronounced PS provide a basis for ventral polarotaxis via the outer photoreceptors, the contribution of R8 is less clear. We found that R8 rhabdomeres twist strongly and, thus, R8 cells are expected to have low PS. In contrast, our behavioral tests demonstrate that R8 can rescue polarotaxis in norpA mutants (Figure 6B). Consistent with this, we found rare cases of very short R8 rhabdomeres exhibiting small twist ranges and correspondingly high expected PS. However, we cannot exclude the possibility that the apparent behavioral contributions of R8 could reflect low-level, expression of our driver lines in R1–R6.

In larger flies, R7y and R8y as well as R1–R6 contain a UV-sensitizing pigment [13]. Since this molecule is not covalently linked to the opsin protein, its function is independent of microvillar orientation, thereby diminishing PS in the UV range [40]. In addition, these cells contain a C40 carotinoid, which both gives them their yellow appearance, and induces anomalous dichroism which further reduces PS [13]. This may explain why R7p but not R7y can mediate ventral UV polarotaxis in Drosophila (Figure 4). Our data describe an unexpected new role for ‘pale’ ommatidia outside the DRA. Moreover, the behavioral data confirm that in R1–R6 the UV sensitizing pigment does not completely eliminate PS in the UV, as was previously reported [40]. The contributions of the outer photoreceptors therefore become more pronounced when polarized green light is presented. That cellular contributions to ventral POL vision differ as a function of wavelength is, particularly interesting since reflections from leaves contain much less UV (and more green light) than reflections from water [6]. Hence, activation of distinct combinations of photoreceptors might convey specific meanings to the fly.

The combination of polarization-sensitive outer and inner photoreceptors represents a new analyzer design, differing from those described in the DRA and the ventral retina of Notonecta [6]. In particular, our morphological data does not reveal an orthogonal organization of ventral analyzers. However, comparison between these channels might still increase quality and robustness of the signal. Nothing is known about the subtype-specific connectivity of R7p/R8p and their post-synaptic partners, and no electrophysiological data on polarization-opponent interneurons [41], or ‘compass neurons’ [42], exist in flies. By establishing Drosophila as a model of polarization vision, our studies will enable genetic screens using quantitative behavioral assays to allow a complete dissection of the neural circuits involved in responding to this fundamental quality of light.

EXPERIMENTAL PROCEDURES

Fly stocks

Oregon R (FlyBase), UAS-shibirets on II and III (Bruce Baker, Stanford), rh1-Gal4 on X (Jessica Treisman, New York), rh2-Gal4 on II (Andrea Brand, Cambridge), rh3-Gal4 on II (FlyBase), rh4-Gal4 on II (FlyBase), rh5-GAL4 on II (FlyBase), rh6-GAL4 on II (FlyBase), rh3+4-Gal4 on II, rh5+6-Gal4 on II, (C. Desplan), hth-GAL4 (C. Desplan), rh1-NorpA on III (R. Shortridge, Buffalo), UAS-CD8:GFP on X and II (Chris Potter), norpA[36] (FlyBase), UAS-CD2:HRP (FlyBase), rh6+DRA.R7-GAL4 on II (Tiffany Cook, Cincinnati).

Generation of UAS-norpA transgenes

A ~900 bp 5′ fragment was PCR-amplifed from VDRC full-length cDNA clone GH28834, [primers: 5′-TGACGAATTCGGTACCGTGCAGGGCAACGGAAACGGAAGCGTC-3′, and 5′-CAACGTTTCTCCTCGTAGAGAGGGTA-3]′. This product was cut with EcoRI/SacII, and ligated into pUAST (EcoRI/XhoI) together with a ~2.5 kb SacII/XhoI fragment excised from GH28834.

Immunohistochemistry

Brains were fixed for 45 minutes in 2% paraformaldehyde and blocked in 10% normal goat serum, then incubated with 1:10 mouse anti-24B10 (Developmental Studies Hybridoma Bank), 1:2000 chicken anti-GFP (Abcam) and 10% normal goat serum, and detected with goat-anti chicken Alexa 488 (Invitrogen), and goat anti-mouse Alexa 594 (Invitrogen), at a 1:200 dilution.

Behavior

All stocks were maintained on molasses, under 12:12 light/dark cycles, with circadian temperature changes between 18°C and 25°C, under 45%–60% humidity. 66 mated female flies were collected 1–3 days after eclosion and sorted onto fresh food. After 2 days, flies were tested within three hours of the onset of light, or four hours before the offset of light. All experiments were performed at 34°C.

Experimental Setup

An unpolarized light source (see below) illuminated a filter set consisting of a polarizer and a diffuser (‘polarizer/diffuser pair), which was rotated by a computer-controlled motor [software: ‘NMC Simple Sequencer’ (Jeffrey Kerr)]. Within a large temperature-controled (Peltier device) chamber 66 flies were contained in a small arena formed by a heavily sanded, plexiglass ring (Ø = 7.5 cm, height = 2.5 cm) between two plates of UV-transparent plexiglass. The distance from polarizer/diffuser pair was 3.5 cm for flies walking on the ‘ceiling’ of the arena and 6 cm for those on the floor. The arena was surrounded by infrared LEDs (880 nm), and flies were filmed from below. Tracking software extracted the position and orientation of each individual fly in real time, sampling at 30Hz [22].

Stimulus

The light of an EXFO X-cite exacte DC light source passed one of three bandpass filter combinations. UV: Schott UV1 (365+/−10nm) + Thorlabs FGB37S, BLUE: Newport 20BPF10-460 (460+/−10nm) + FGL435S), or GREEN: Newport 20BPF10-510 (510+/−10nm) + FGL435S). All stimuli were calibrated with an Ocean Optics USB 2000 spectrophotometer. Polarizer (HN42HE, Polaroid) and diffuser (2 sheets of tracing paper: ‘Transparentpapier’, Max Bringmann KG, Germany) were illuminated through a 35 mm Zeiss collimating adapter. The light stimulus was either linearly polarized or unpolarized depending on which side of the polarizer/diffuser pair faced the flies. The stimulus aperture was limited to 5 cm using a black plastic sheet with circular opening. Only flies walking directly under this aperture were tracked.

Metrics

A value: The ‘alignment metric’, A, for quantification of the behavioral response is extracted in several steps. 1) All fly angular headings during the stopped epochs for a given experiment are binned in 2° increments from 0 to 2π and transformed into a probability distribution. 2) This probability distribution was fitted to A*cos(2*θ+ϕ)+b (where θ is the fly heading angle, ϕ the phase shift of the cosine function, and b is the offset). 3) A percent modulation (PM) = Amplitude/mean(probability) was then calculated. 4) ‘A value’ = PM * cos(ϕ). Thus, if the phase shift ϕ is zero, then the A value equals the PM. However, if ϕ is shifted, then the A value decreases. Inspection of the polar histograms revealed that the amplitude of the modulation (the strength of the behavioral response) was invariably coupled to the phase of the cosine function. That is, we never observed flies to align precisely at any position other than parallel to the e-vector.

Morphology

For electron microcopy (EM) of the dorsal-most retina, the eyes of wild type Drosophila (OregonR) were split in the horizontal plane. The dorsal eye halves were fixed with 2% glutaraldehyde in 0.05 M Na-cacodylate (pH 7.2) for 2 h at 4° C, and postfixed with 2% OsO4 in 0.05 M Na-cacodylate (pH 7.2) for 2 h at 4° C, followed by dehydration with 2,2 dimethoxypropane, and embedded in Epon 812. Silver sections were stained with uranyl acetate and lead citrate.

Tangential sections of groups of identified ommatidia were taken at 5 μm intervals and photographed in the electron microscope. Microvilli orientations were measured relative to a straight line through the rhabdomere centers of R1 and R3 as a reference. Twist functions were obtained by graphing microvilli orientation vs. retinal level.

Supplementary Material

01

Highlights.

  • Drosophila orients to polarized light presented either dorsally, or ventrally.

  • Dorsal polarotactic behavior is mediated by the ‘Dorsal Rim Area’

  • Low twist R7 photoreceptors in the ventral eye can mediate polarotactic responses.

  • Outer photoreceptors R1–R6 contribute to ventral polarotactic responses.

Acknowledgments

The authors thank Bob Schneeveis, David Profitt, and Primoz Pirih for technical assistance. Claude Desplan, Bruce Baker, Russell Shortridge, Tiffany Cook, Jessica Treisman, and Martin Heisenberg provided fly stocks. This work was supported by the Helen Hay Whitney Foundation (MFW), the Jane Coffin Childs Foundation (DAC), a Ruth L. Kirschstein Graduate Fellowship Award (MMV), and by an NIH Director’s Pioneer Award (DP1 OD003530) to TRC. This work was also supported by a Burroughs-Wellcome Career Development Award (TRC), a Mcknight Scholar Award (TRC), and Klingenstein Fellowship (TRC), and a Searle Scholar Award (TRC).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Wehner R. Polarization vision – a uniform sensory capacity? J Exp Biol. 2001;204:2589–2596. doi: 10.1242/jeb.204.14.2589. [DOI] [PubMed] [Google Scholar]
  • 2.Wehner R, Labhart T. Polarization vision. In: Warrant EJ, Nilsson D-E, editors. Invertebrate vision. Cambridge: 2006. [Google Scholar]
  • 3.Nilsson DE, Warrant EJ. Visual discrimination: Seeing the third quality of light. Curr Biol. 1999;9:R535–537. doi: 10.1016/s0960-9822(99)80330-3. [DOI] [PubMed] [Google Scholar]
  • 4.Rossel S. Navigation by bees using polarized skylight. Comp Biochem Physiol. 1993;104A:695–708. [Google Scholar]
  • 5.Wehner R. Desert ant navigation: how minibrains solve complex tasks. J Comp Physiol A. 2003;189:579–588. doi: 10.1007/s00359-003-0431-1. [DOI] [PubMed] [Google Scholar]
  • 6.Schwind R. Zonation of the optical environment and zonation of the rhabdom structure within the eye of the backswimmer, Notonecta glauca. Cell Tissue Res. 1983;232:53–63. doi: 10.1007/BF00222373. [DOI] [PubMed] [Google Scholar]
  • 7.Wildermuth H. Dragonflies recognize the water of rendezvous and oviposition sites by horizontally polarized light: a behavioral field test. Naturwissenschaften. 1998;85:297–302. [Google Scholar]
  • 8.Shashar N, Sabbah S, Aharoni N. Migrating locusts can detect polarized reflections to avoid flying over the sea. Biol Letters. 2005;4:472–475. doi: 10.1098/rsbl.2005.0334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Horvath G, Majer J, Horvath L, Szivak I, Kriska G. Ventral polarization vision in tabanids: horseflies and deerflies (Diptera: Tabanidae) are attracted to horizontally polarized light. Naturwissenschaften. 2008;95:1093–1100. doi: 10.1007/s00114-008-0425-5. [DOI] [PubMed] [Google Scholar]
  • 10.Labhart T, Meyer EP. Neural mechanisms in insect navigation: polarization compass and odometer. Curr Opin Neurobiol. 2002;12:707–714. doi: 10.1016/s0959-4388(02)00384-7. [DOI] [PubMed] [Google Scholar]
  • 11.Hardie RC. Properties of photoreceptors R7 and R8 in dorsal marginal ommatidia in the compound eyes of Musca and Calliphora. J Comp Physiol A. 1984;154:157–165. [Google Scholar]
  • 12.Labhart T, Meyer EP. Detectors for polarized skylight in insects: a survey of ommatidial specializations in the dorsal rim area of the compound eye. Microsc Res Tech. 1999;47:368–379. doi: 10.1002/(SICI)1097-0029(19991215)47:6<368::AID-JEMT2>3.0.CO;2-Q. [DOI] [PubMed] [Google Scholar]
  • 13.Hardie RC. Functional organization of the fly retina. In: Autrum H, Ottoson D, Perl ER, Schmidt RF, Shimazu H, Willis WD, editors. Progress in Sensory Physiology. Berlin: Springer; 1985. pp. 1–79. [Google Scholar]
  • 14.Wernet MF, Desplan C. Building a retinal mosaic: cell fate decisions in the fly eye. Trends Cell Biol. 2004;14:576–584. doi: 10.1016/j.tcb.2004.09.007. [DOI] [PubMed] [Google Scholar]
  • 15.Wada S. Spezielle randzonale Ommatidien der Fliegen (Diptera: Brachycera): Architektur und Verteilung in den Komplexaugen. Z Morph Tiere. 1974;77:87–125. [Google Scholar]
  • 16.Fortini M, Rubin GM. The optic lobe projection pattern of polarization-sensitive photoreceptor cells in Drosophila melanogaster. Cell Tissue Res. 1991;265:185–191. doi: 10.1007/BF00318153. [DOI] [PubMed] [Google Scholar]
  • 17.Wernet MF, Labhart T, Baumann F, Mazzoni EO, Pichaud F, Desplan C. Homothorax switches function of Drosophila photoreceptors from color to polarized light sensors. Cell. 2003;115:267–79. doi: 10.1016/s0092-8674(03)00848-1. [DOI] [PubMed] [Google Scholar]
  • 18.Heisenberg M, Buchner E. The role of retinula cell types in visual behavior of Drosophila melanogaster. J Comp Physiol. 1977;117:127–162. [Google Scholar]
  • 19.Gao S, Takemura SY, Ting CY, Huang S, Lu Z, Luan H, Rister J, Thum AS, Yang M, Hong ST, Wang JW, Odenwald WF, White BH, Meinertzhagen IA, Lee CH. The neural substrate of spectral preference in Drosophila. Neuron. 2008;60:328–342. doi: 10.1016/j.neuron.2008.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Yamaguchi S, Desplan C, Heisenberg M. Contribution of photoreceptor subtypes to spectral wavelength preference in Drosophila. Proc Natl Acad Sci. 2010;107:5634–9. doi: 10.1073/pnas.0809398107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Rister J, Pauls D, Schnell B, Ting CY, Lee CH, Sinakevitch I, Morante J, Strausfeld NJ, Ito K, Heisenberg M. Dissection of the peripheral motion channel in the visual system of Drosophila melanogaster. Neuron. 2007;56(1):155–170. doi: 10.1016/j.neuron.2007.09.014. [DOI] [PubMed] [Google Scholar]
  • 22.Katsov AY, Clandinin TR. Motion processing streams are behaviorally specialized. Neuron. 2008;59:322–335. doi: 10.1016/j.neuron.2008.05.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.von Philipsborn A, Labhart T. A behavioural study of polarization vision in the fly, Musca domestica. J Comp Physiol A. 1990;167:737–743. [Google Scholar]
  • 24.Wunderer H, Smola U. Fine structure of ommatidia at the dorsal eye margin of Calliphora Erythrocephala Meigen (Diptera: Calliphoridae): an eye region specialized for the detection of polarized light. Int J Insect Morphol & Embryol. 1982a;11:25–38. [Google Scholar]
  • 25.Wolf R, Gebhardt B, Gademann R, Heisenberg M. Polarization sensitivity of course control in Drosophila melanogaster. J Comp Physiol. 1980;139:177–191. [Google Scholar]
  • 26.Labhart T. How polarization-sensitive interneurones of crickets perform at low degrees of polarization. J Exp Biol. 1996;199:1467–1475. doi: 10.1242/jeb.199.7.1467. [DOI] [PubMed] [Google Scholar]
  • 27.Kitamoto T. Conditional modification of behavior in Drosophila by targeted expression of a temperature-sensitive shibire allele in defined neurons. J Neurobiol. 2001;47:81–92. doi: 10.1002/neu.1018. [DOI] [PubMed] [Google Scholar]
  • 28.Bloomquist BT, Shortridge RD, Schneuwly S, Perdew M, Montell C, Steller H, Rubin GM, Pak WL. Isolation of a putative phospholipase C gene of Drosophila, norpA, and its role in phototransduction. Cell. 1988;54:723–733. doi: 10.1016/s0092-8674(88)80017-5. [DOI] [PubMed] [Google Scholar]
  • 29.Israelachvili JN, Wilson M. Absorption characteristics of oriented photopigments in microvilli. Biol Cybernetics. 1976;21:9–15. doi: 10.1007/BF00326667. [DOI] [PubMed] [Google Scholar]
  • 30.Goldsmith TH, Wehner R. Restrictions on rotational and translational diffusion of pigment in the membranes of a rhabdomeric photoreceptor. J Gen Physiol. 1977;70:453–490. doi: 10.1085/jgp.70.4.453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Wehner R, Bernard GD, Geiger E. Twisted and non-twisted rhabdoms and their significance for polarization detection in the bee. J Comp Physiol. 1975;104:225–245. [Google Scholar]
  • 32.Nilsson D, Labhart T, Meyer EP. Photoreceptor design and optical properties affecting polarization sensitivity in ants and crickets. J Comp Physiol A. 1987;161:645–658. [Google Scholar]
  • 33.Wunderer H, Smola U. Morphological differentiation of the central visual cells R7/8 in various regions of the blowfly eye. Tissue and Cell. 1982b;14:341–358. doi: 10.1016/0040-8166(82)90032-5. [DOI] [PubMed] [Google Scholar]
  • 34.Smola U, Tscharntke H. Twisted rhabdomeres in the Dipteran eye. J Comp Physiol. 1979;133:291–297. [Google Scholar]
  • 35.Smola U, Wunderer H. Fly rhabdomeres twist in vivo. J Comp Physiol. 1981a;142:43–49. [Google Scholar]
  • 36.Smola U, Wunderer H. Twisting of blowfly (Calliohora erythrocephala Meigen)(Diptera, Calliphoridae) rhabdomeres: an in vivo feature unaffected by prepreation or fixation. Int J Insect Morphol Embryol. 1981b;10:331–344. [Google Scholar]
  • 37.Stalleicken J, Labhart T, Mouritsen H. Physiological characterization of the compound eye in monarch butterflies with focus on the dorsal rim area. J Comp Physiol A. 2006;192:321–331. doi: 10.1007/s00359-005-0073-6. [DOI] [PubMed] [Google Scholar]
  • 38.Brunner D, Labhart T. Behavioural evidence for polarization vision in crickets. Physiological Entomology. 1987;12:1. [Google Scholar]
  • 39.Mappes M, Homberg U. Behavioral analysis of polarization vision in tethered flying locusts. J Comp Physiol A. 2004;190(1):61. doi: 10.1007/s00359-003-0473-4. [DOI] [PubMed] [Google Scholar]
  • 40.Guo A. Elektrophysiologische Untersuchungen zur Spektral- und Polarisationsempfindlichkeit der Sehzellen von Calliphora erythrocephala III. Sci Sin. 1981;24:272–286. [Google Scholar]
  • 41.Labhart T. Polarization-opponent interneurons in the insect visual system. Nature. 1988;331:435–437. [Google Scholar]
  • 42.Heinze S, Homberg U. Maplike representation of celestial e-vector orientations in the brain of an insect. Science. 2007;327(2):385–398. doi: 10.1126/science.1135531. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS339789-supplement-01.pdf (2.1MB, pdf)

RESOURCES