The brain extracts information from signals delivered from the eyes and other sensory organs in order to direct behavior. Understanding how the interactions and wiring of a multitude of individual nerve cells process and transmit this critical information to the brain is a fundamental goal in the field of neuroscience.

One question many neuroscientists have tried to understand is how nerve cells in an animal’s brain detect direction when an animal sees movement of some kind – so-called motion vision. The raw signal from the light receptors in the eye does not discriminate whether the light moves in one direction or the other. So, the nerve cells in the brain must somehow compute the direction of movement based on the information relayed by the eye.

For more than half a century, major debates have revolved around two rival models that could explain how motion vision works. Both models could in principle lead to neurons that prefer images moving in one direction over images moving in the opposite direction – so-called direction selectivity. In both models, the information about the changing light levels hitting two light-sensitive cells at two points on the eye are compared across time. In one model, signals from images moving in a cell’s preferred direction become amplified. In the other model, signals moving in the unfavored direction become canceled out. However, neither model perfectly explains motion vision.

Now, Haag, Arenz et al. show that both models are partially correct and that the two mechanisms work together to detect motion across the field of vision more accurately. In the experiments, both models were tested in tiny fruit flies by measuring the activity of the first nerve cells that respond to the direction of visual motion. While each mechanism alone only produces a fairly weak and error-prone signal of direction, together the two mechanisms produce a stronger and more precise directional signal. Further research is now needed to determine which individual neurons amplify or cancel the signals to achieve such a high degree of direction selectivity.

Flies see the world through a hexagonal array of facets each equipped with its own small lens focusing the light onto 8 photoreceptors. Photoreceptors send their axons into the optic lobe, which consists of four consecutive layers of neuropil called lamina, medulla, lobula and lobula plate (Figure 1a). Each neuropil layer is made up of retinotopically organized columns together containing roughly 100 different neurons per column (Fischbach and Dittrich, 1989). Within the optic lobe, visual motion information is extracted in parallel pathways encoding light increments (ON) and decrements (OFF) (Joesch et al., 2010; Eichner et al., 2011; Joesch et al., 2013). Both pathways bifurcate in the lamina and lead, via a set of specific medulla interneurons, onto the dendrites of T4 and T5 cells, respectively (Takemura et al., 2013; Shinomiya et al., 2014). Both T4 and T5 cells exist in 4 subgroups tuned to one of the four cardinal directions of motion and project into four layers of the lobula plate (Maisak et al., 2013) (Figure 1b). There, they form monosynaptic cholinergic excitatory connections with the dendrites of the large-field lobula plate tangential cells (Mauss et al., 2014; Schnell et al., 2012) as well as with lobula plate intrinsic neurons. These intrinsic neurons, in turn, inhibit tangential cells in the adjacent layer (Mauss et al., 2015). Since none of the neurons upstream from T4 and T5 respond to visual motion in a direction-selective way (Behnia et al., 2014; Strother et al., 2014; Meier et al., 2014; Ammer et al., 2015; Serbe et al., 2016), T4 and T5 cells are the first neurons in the processing chain where directional information is represented explicitly (Maisak et al., 2013; Fisher et al., 2015).

Figure 1

Download asset Open asset

To investigate the mechanism leading to direction selectivity in T4 cells, we applied apparent motion stimuli where, instead of continuously moving (Figure 1c, top), a bar is abruptly stepped from one location to a neighboring one (Figure 1c, bottom). These stimuli lend themselves well to discriminate between preferred direction (‘PD’) enhancement (Hassenstein and Reichardt, 1956) (Figure 1d, left) and null direction (‘ND’) suppression (Barlow and Levick, 1965) (Figure 1d, right) as the response to the sequence can be compared with the sum of the responses to the luminance pulses given in isolation (‘linear expectation’): in case of preferred direction enhancement, the response to the sequence along the preferred direction is larger than the linear sum of responses to the isolated pulses and identical if the sequence is along the null direction; In case of null direction suppression, the response to the sequence along the preferred direction is identical to the linear sum of responses to the isolated pulses and smaller if the sequence is along the null direction.

The columnar organization of the optic lobes stretches from the laminar cartridges through the medulla and into the lobula and lobula plate. Individual T4 and T5 neurons extend their dendrites across multiple columns (Figure 1e) and receive synaptic inputs from different medulla cell types located in different columns relative to the home columns of the T4 or T5 neuron and to each other (Takemura et al., 2013; Shinomiya et al.; 2014). To understand the particular contribution of those inputs it seemed necessary to precisely place the stimuli onto the columnar raster of the fly’s optic lobe.

The structure of the optical system and of the neuronal wiring in flies obeys the neural superposition principle (Kirschfeld, 1967; Braitenberg, 1967). Those photoreceptors R1-6 from 6 neighboring ommatidia that share the same optical axis converge on the same lamina cartridge, thus forming an optical column that represents the unit of spatial resolution (‘neuro-ommatidium’). To visually stimulate these neuro-ommatidia precisely and in isolation, stimulation must consist of parallel rays at angles along the optical axes of those photoreceptors and aligned to the columnar raster. For this, we adopted a telescopic stimulation device (Franceschini, 1975; Schuling et al., 1989) (Figure 1f). As the fly rhabdomeric photoreceptors work as light guides, the raster of optical columns can be visualized by shining light from within the head capsule (antidromic illumination) to align the raster of neuro-ommatidia to the stimulation locations on a micro-display with the help of a CMOS camera.

As a proof of principle, we first expressed the genetically encoded calcium indicator GCaMP6m (Chen et al., 2013) in lamina cells L2, recorded from their terminals in the medulla by 2-photon microscopy (Denk et al., 1990) and stimulated them with light spots of 1176 ms duration covering a single optical column at various positions. L2 cells responded maximally to stimuli at a certain column, with less than 30% response amplitude to stimuli positioned on surrounding columns, and negligible responses (<12%) to stimuli positioned on columns in the next outer ring (Figure 2a). The slight activation of neighboring columns results directly from the optics of the fly’s eye, where the visual fields of single ommatidia have Gaussian sensitivity profiles with an acceptance angle roughly matching the inter-ommatidial separation (Götz, 1965). These experiments illustrate the specificity of the telescopic stimulation to single lamina cartridges and thus optic lobe column with minimal cross-stimulation of neighboring columns at the physical limit of the fly optical system.

Figure 2

Download asset Open asset

We next used a driver line specific for those T4 and T5 cells sending their axons into layer 3 of the lobula plate that are hence sensitive to upward motion. We recorded from single T4 cells by selecting individual processes in layer three of the lobula plate and confirmed their preference for luminance increments. Repeating the above experiment with T4 cells, we again found maximal responses to the stimulus placed in a single column. Compared to L2, however, the receptive field was found to be larger, with about 50% amplitude to the stimuli onto the six surrounding columns, and roughly 25% to the next outer ring (Figure 2b). This indicates that T4 neurons pool excitatory synaptic input from more than one column, consistent with their morphology (Figure 1e) and expected from a motion detector that is required to integrate information from spatially offset input.

In order to discriminate between preferred direction enhancement and null direction suppression, we tested T4 cells with three light pulses of 472 ms duration positioned along the dorso-ventral axis of the eye (Figure 3a, left). T4 cells responded to the individual pulses with different amplitudes, depending on the position of the stimulus (Figure 3a, ‘Flicker’; see also Figure 2b). When stimulated sequentially from ventral to dorsal (Figure 3a, ‘Sequence’, top middle), the cell responded more strongly (thick blue line) than expected from the sum of the responses to the individual stimuli (thin blue line). The opposite was observed when we stimulated the cell sequentially from dorsal to ventral (Figure 3a, ‘Sequence’, bottom middle): now the cell responded more weakly (thick red line) than expected from the sum of the responses to the individual stimuli (thin red line). We then calculated the nonlinear response component by subtracting the linear expectation from the actual response and found that both preferred direction enhancement (Figure 3a, ‘Sequence’, top right) and null direction suppression (Figure 3a, ‘Sequence’, bottom right) contributes to the directionally selective responses of T4 cells.

Figure 3

Download asset Open asset

We then asked whether these two mechanisms occupied separate or overlapping receptive fields. For that, we presented pulses in four neighboring columns along the dorso-ventral axis, individually (Figure 3b, left) as well as sequences of pairs in adjacent columns (Figure 3b, middle). Stepping the stimulus up and down between the uppermost columns, we observed only null direction suppression but no preferred direction enhancement (Figure 3b, right top). The opposite was true for stimuli at the two lower most locations: here, only preferred direction enhancement was observed, but no null direction suppression (Figure 3b, right bottom). For sequences between the two inner locations, both phenomena were visible (Figure 3b, right middle). We conclude that the different mechanisms are offset in their receptive fields: preferred direction enhancement is shifted towards the ‘null-side’, and null direction suppression towards the ‘preferred side’ with respect to each other.

To assess the time-course of both mechanisms, we presented two pulses of 472 ms duration and delayed one with respect to the other by a variable time-lag (Figure 3c). When we placed the pulses onto the ventral part of the receptive field, delaying the upper with respect to the lower one (PD), responses peaked at a delay of about 500 ms (Figure 3c, top). When the pulses were placed onto the dorsal part of the receptive field, delaying the lower with respect to the upper one (ND), responses were suppressed and returned to base line at a delay larger than 500 ms (Figure 3c, bottom).

We also noticed that this null direction suppression builds up over longer sequences of steps (Figure 3d). When the stimulation of the central column was preceded by the stimulation of one directly adjacent column on the ‘preferred side’, the response to the central pulse was strongly suppressed as before (red trace). This suppression led to a response to the sequence that on average not only falls below the sum of both individual stimulations, but even below the flicker response to the central column (black trace), indicating inhibition. When this null-direction sequence was extended from two to three columns, the resulting response was even smaller (green trace).

The experiments shown in Figure 3c also revealed that for zero onset time differences, i.e. simultaneous stimulation of two columns, the response was suppressed as compared to stimulation of just one column. We tested the receptive field of this suppression of the central column systematically by simultaneously stimulating the central column together with one of the columns in the 2 rings surrounding it (Figure 3e). We normalized the response to the two stimulated columns with respect to the response to the isolated stimulation of the central column. The resulting response in the T4 neuron was found to be suppressed in comparison to the isolated central column response for simultaneous stimulation of the central column together with another one in the dorsal part of the receptive field.

This suggests that employing apparent motion with flicker stimuli of larger bars should only lead to small flicker responses, possibly occluding null-direction suppression. We tested this prediction by presenting bright horizontal bars of different spatial extent at different locations within the receptive field of individual T4 cells on an LED arena (Figure 4a,c). Using a bar size of 180° × 4.5°, indeed, flicker responses of T4 cells were almost undetectable (Figure 4a). Apparent motion stimuli along the preferred direction led to pronounced responses, larger than the linear prediction, while null direction sequences did not suppress the negligible sum of flicker responses by a significant amount (Figure 4b). This was much different when repeating the same stimuli using bars of only 4.5° × 3° instead: now, as in the experiments employing telescopic stimulation of individual columns, strong flicker responses appeared (Figure 4c). Furthermore, in addition to the preferred direction enhancement, pronounced null direction suppression was observed with peak sensitivity within the dorsal part of the receptive field (Figure 4d).

Figure 4

Download asset Open asset

What is the virtue of having null direction suppression in addition to the preferred direction enhancement? To address this question, we constructed a simple algorithmic model of a motion detector (Figure 5a). In this model, the motion detector receives 3 inputs. A central direct input (D) is flanked by two low-pass filtered inputs: one enhancing input (E) implementing a multiplicative and one suppressing input (S) implementing a divisive non-linearity. This way, the detector is designed to combine a preferred direction enhancement on one and a null direction suppression on the other side within a single motion-sensing unit. Testing the model with single pulses, responses were only detected for positions −1 and 0 (Figure 5b, left). Stimulating the model with 2-pulse-sequences at positions −1 and 0 led to pronounced preferred direction enhancement but no null direction suppression (Figure 5b, top 2 rows). When the 2-pulse-sequences were delivered at positions 0 and 1, no pronounced preferred direction enhancement occurred, but null direction suppression was substantial (Figure 5b, bottom 2 rows). Thus, these simulation results qualitatively match the respective experimental data (compare with Figure 3a and b). We next simulated an array of such units and calculated its responses to grating motion at different speeds and directions. We found that the model exhibits a high degree of directional selectivity over a broad range of velocities, with little responses to null direction motion (Figure 5c, middle left). In contrast, the model without null direction suppression revealed substantial null direction responses almost as large as its preferred direction responses (Figure 5c, middle center). The same was true for a model with no preferred direction enhancement (Figure 5c, middle right). Interestingly, no such qualitative differences were observed at the level of a model tangential cell, i.e. after subtraction of opponent units (Figure 5c, bottom). When we simulated the responses of the model to grating motion into different directions, we found the full model T4 cell to be more sharply tuned to its preferred direction than the one without preferred direction enhancement or the one without null direction suppression (Figure 5d). We conclude that combining preferred direction enhancement with null direction suppression leads to a strong direction selectivity of the primary motion-sensing unit.

Figure 5

Download asset Open asset

We next investigated the direction selectivity of T4 and T5 cells by recording their responses to moving gratings presented on an LED arena moving at various speeds and directions. Over a wide range of velocities spanning more than two orders of magnitude, T4/T5 cells responded almost exclusively to upward motion, i.e. along their preferred direction, with little or no responses at all to motion along their null direction (Figure 6a). When stimulating the cells with 1 Hz grating motion at various directions, we found a rather sharp directional tuning with about 90° half-width around its preferred direction (Figure 6b; see also Figure 3g in Maisak et al., 2013). To rule out that the highest degree of directional selectivity in T4/T5 cells is achieved by a hitherto unknown reciprocal inhibition between T4/T5 cells with opposite preferred direction, e.g. via inhibitory lobula plate interneurons (Mauss et al., 2015) we blocked the synaptic output from all T4 and T5 cells by expression of tetanus toxin light-chain (Sweeney et al., 1995). Having confirmed the effectiveness of the block (Supplementary file 1), we repeated the above experiments. T4/T5 cells revealed the same high degree of directional selectivity as before (Figure 6c,d). From this, we conclude that the mechanism underlying the high degree of directional selectivity in T4 cells does not include the output of oppositely tuned T4 cells but rather originates in the dendrite of the cells.

Figure 6

Download asset Open asset

In the Hassenstein-Reichardt model (Hassenstein and Reichardt, 1956) as well as in the Barlow-Levick model (Barlow and Levick, 1965) direction selectivity emerges by a nonlinear interaction of asymmetrically filtered signals from adjacent image points: an enhancement along the preferred direction in the first, and a suppression along the null direction in the second model. Both models lead to weakly direction-selective signals in the first place, which, in the Hassenstein-Reichardt model, are improved downstream by subtraction of oppositely tuned components (for review, see Borst and Helmstaedter, 2015. Surprisingly, however, a high degree of directional selectivity is found already at the first stage where directional responses are observed, i.e. in T4 and T5 cells (Maisak et al., 2013) (Figure 6). This can now be explained by the fact that both preferred direction enhancement and null direction suppression are implemented in T4 cells. While the output of the full Hassenstein-Reichardt-correlator after the subtraction stage shows a high degree of directional selectivity in the absence of null-direction suppression, the relatively small differences between large, but poorly tuned signals (see Figure 5c, middle panel) would be highly prone to noise. Improving the direction-tuning already at the level of the half-detectors by the additional null-direction suppression increases robustness to noise and might in addition be energetically less costly.

A recent study addressing the mechanisms underlying the elementary motion detectors in Drosophila concluded preferred direction enhancement as the sole mechanism (Fisher et al. 2015). However, we find that their stimulation paradigm consisting of flashes at 2 positions simultaneously does not allow to discriminate between different mechanisms including lateral inhibition, adaptation and null direction suppression.

Since the operations of preferred direction enhancement and null direction suppression have offset receptive fields, it requires the cells to collect their input from more than two adjacent image points, in line with anatomical observations (Figure 1e; Takemura et al., 2013; Shinomiya et al., 2014) . In contrast to the cellular implementation of the motion detector proposed by Behnia et al, 2014 our findings also require that T4 cells receive input from more than two cell types. This again is in agreement with recent connectomic data (for T4 cells: Louis Scheffer, Janelia Research Campus, personal communication; for T5 cells: Shinomiya et al., 2014). Finally, functional studies where the synaptic input elements onto T4 and T5 neurons were blocked also suggested the involvement of more than two input elements in the ON- (Ammer et al., 2015) as well as in the OFF-pathway (Serbe et al., 2016).

Considering the columnar nature of the input elements and the structure of the T4 dendrites covering multiple columns, we predict, that the inputs representing the enhancing, direct and suppressing input segregate on different sub-regions of the T4 dendrite, with enhancing inputs on the ‘null-side’ and the suppressing inputs on the ‘preferred side’ of the dendrite. Future experiments will have to map the different cell types of the medulla onto the proposed model (Figure 5a) as well as to identify the transmitter receptors, responsible for enhancing, exciting and inhibiting the dendrites of T4 and T5 cells.

1. 1. Ammer G
   2. Leonhardt A
   3. Bahl A
   4. Dickson BJ
   5. Borst A

   (2015) Functional specialization of neural input elements to the Drosophila ON motion detector

   Current Biology 25:2247–2253.

   https://doi.org/10.1016/j.cub.2015.07.014

   • PubMed
   • Google Scholar
2. 1. Barlow HB
   2. Levick WR

   (1965) The mechanism of directionally selective units in rabbit's retina

   Journal of Physiology 178:477–504.

   https://doi.org/10.1113/jphysiol.1965.sp007638

   • Google Scholar
3. 1. Behnia R
   2. Clark DA
   3. Carter AG
   4. Clandinin TR
   5. Desplan C

   (2014) Processing properties of ON and OFF pathways for Drosophila motion detection

   Nature 512:427–430.

   https://doi.org/10.1038/nature13427

   • PubMed
   • Google Scholar
4. 1. Borst A
   2. Helmstaedter M

   (2015) Common circuit design in fly and mammalian motion vision

   Nature Neuroscience 18:1067–1076.

   https://doi.org/10.1038/nn.4050

   • Google Scholar
5. 1. Brainard DH

   (1997) The Psychophysics Toolbox

   Spatial Vision 10:433–436.

   https://doi.org/10.1163/156856897X00357

   • Google Scholar
6. 1. Braitenberg V

   (1967) Patterns of projection in the visual system of the fly. I. Retina-lamina projections

   Experimental Brain Research 3:271–298.

   https://doi.org/10.1007/BF00235589

   • PubMed
   • Google Scholar
7. 1. Chen T-W
   2. Wardill TJ
   3. Sun Y
   4. Pulver SR
   5. Renninger SL
   6. Baohan A
   7. Schreiter ER
   8. Kerr RA
   9. Orger MB
   10. Jayaraman V
   11. Looger LL
   12. Svoboda K
   13. Kim DS

   (2013) Ultrasensitive fluorescent proteins for imaging neuronal activity

   Nature 499:295–300.

   https://doi.org/10.1038/nature12354

   • Google Scholar
8. 1. Denk W
   2. Strickler JH
   3. Webb WW

   (1990) Two-photon laser scanning fluorescence microscopy

   Science 248:73–76.

   https://doi.org/10.1126/science.2321027

   • PubMed
   • Google Scholar
9. 1. Eichner H
   2. Joesch M
   3. Schnell B
   4. Reiff DF
   5. Borst A

   (2011) Internal structure of the fly elementary motion detector

   Neuron 70:1155–1164.

   https://doi.org/10.1016/j.neuron.2011.03.028

   • Google Scholar
10. 1. Fischbach K-F
    2. Dittrich APM

    (1989) The optic lobe of Drosophila melanogaster. I. A Golgi analysis of wild-type structure

    Cell and Tissue Research 258:441–475.

    https://doi.org/10.1007/BF00218858

    • Google Scholar
11. 1. Fisher YE
    2. Silies M
    3. Clandinin TR

    (2015) Orientation selectivity sharpens motion detection in Drosophila

    Neuron 88:390–402.

    https://doi.org/10.1016/j.neuron.2015.09.033

    • Google Scholar
12. 1. Franceschini N

    (1975)

    Photoreceptor Optics

    98–125, Sampling of the visual environment by the compound eye of the fly: fundamentals and applications, Photoreceptor Optics, 10.1007/978-3-642-80934-7_6.

    • Google Scholar
13. 1. Götz KG

    (1965) Die optischen Uebertragungseigenschaften der Komplexaugen von Drosophila

    Kybernetik 2:215–221.

    https://doi.org/10.1007/BF00306417

    • Google Scholar
14. 1. Hassenstein B
    2. Reichardt W

    (1956) Systemtheoretische Analyse der Zeit-, Reihenfolgen- und Vorzeichenauswertung bei der Bewegungsperzeption des Rüsselkäfers Chlorophanus

    Zeitschrift Für Naturforschung 11:513–524.

    https://doi.org/10.1515/znb-1956-9-1004

    • Google Scholar
15. 1. Joesch M
    2. Schnell B
    3. Raghu SV
    4. Reiff DF
    5. Borst A

    (2010) ON and OFF pathways in Drosophila motion vision

    Nature 468:300–304.

    https://doi.org/10.1038/nature09545

    • Google Scholar
16. 1. Joesch M
    2. Weber F
    3. Eichner H
    4. Borst A

    (2013) Functional specialization of parallel motion detection circuits in the fly

    Journal of Neuroscience 33:902–905.

    https://doi.org/10.1523/JNEUROSCI.3374-12.2013

    • PubMed
    • Google Scholar
17. 1. Kirschfeld K

    (1967) Die Projektion der optischen Umwelt auf das Raster der Rhabdomere im Komplexauge von Musca

    Experimental Brain Research 3:248–270.

    https://doi.org/10.1007/BF00235588

    • Google Scholar
18. 1. Maisak MS
    2. Haag J
    3. Ammer G
    4. Serbe E
    5. Meier M
    6. Leonhardt A
    7. Schilling T
    8. Bahl A
    9. Rubin GM
    10. Nern A
    11. Dickson BJ
    12. Reiff DF
    13. Hopp E
    14. Borst A

    (2013) A directional tuning map of Drosophila elementary motion detectors

    Nature 500:212–216.

    https://doi.org/10.1038/nature12320

    • PubMed
    • Google Scholar
19. 1. Mauss AS
    2. Meier M
    3. Serbe E
    4. Borst A

    (2014) Optogenetic and pharmacologic dissection of feedforward inhibition in Drosophila motion vision

    Journal of Neuroscience 34:2254–2263.

    https://doi.org/10.1523/JNEUROSCI.3938-13.2014

    • Google Scholar
20. 1. Mauss AS
    2. Pankova K
    3. Arenz A
    4. Nern A
    5. Rubin GM
    6. Borst A

    (2015) Neural circuit to integrate opposing motions in the visual field

    Cell 162:351–362.

    https://doi.org/10.1016/j.cell.2015.06.035

    • PubMed
    • Google Scholar
21. 1. Meier M
    2. Serbe E
    3. Maisak MS
    4. Haag J
    5. Dickson BJ
    6. Borst A

    (2014) Neural circuit components of the Drosophila OFF motion vision pathway

    Current Biology 24:385–392.

    https://doi.org/10.1016/j.cub.2014.01.006

    • PubMed
    • Google Scholar
22. 1. Pologruto TA
    2. Sabatini BL
    3. Svoboda K

    (2003) ScanImage: Flexible software for operating laser scanning microscopes

    BioMedical Engineering Online 2:13.

    https://doi.org/10.1186/1475-925X-2-13

    • Google Scholar
23. 1. Reiser MB
    2. Dickinson MH

    (2008) A modular display system for insect behavioral neuroscience

    Journal of Neuroscience Methods 167:127–139.

    https://doi.org/10.1016/j.jneumeth.2007.07.019

    • PubMed
    • Google Scholar
24. 1. Schnell B
    2. Raghu SV
    3. Nern A
    4. Borst A

    (2012) Columnar cells necessary for motion responses of wide-field visual interneurons in Drosophila

    Journal of Comparative Physiology A 198:389–395.

    https://doi.org/10.1007/s00359-012-0716-3

    • Google Scholar
25. 1. Schuling FH
    2. Mastebroek HAK
    3. Bult R
    4. Lenting BPM

    (1989) Properties of elementary movement detectors in the fly Calliphora erythrocephala

    Journal of Comparative Physiology A 165:179–192.

    https://doi.org/10.1007/BF00619192

    • Google Scholar
26. 1. Serbe E
    2. Meier M
    3. Leonhardt A
    4. Borst A

    (2016) Comprehensive characterization of the major presynaptic elements to the Drosophila OFF motion detector

    Neuron 89:829–841.

    https://doi.org/10.1016/j.neuron.2016.01.006

    • PubMed
    • Google Scholar
27. 1. Shinomiya K
    2. Karuppudurai T
    3. Lin TY
    4. Lu Z
    5. Lee CH
    6. Meinertzhagen IA

    (2014) Candidate neural substrates for off-edge motion detection in Drosophila

    Current Biology 24:1062–1070.

    https://doi.org/10.1016/j.cub.2014.03.051

    • PubMed
    • Google Scholar
28. 1. Strother JA
    2. Nern A
    3. Reiser MB

    (2014) Direct observation of ON and OFF pathways in the Drosophila visual system

    Current Biology 24:976–983.

    https://doi.org/10.1016/j.cub.2014.03.017

    • PubMed
    • Google Scholar
29. 1. Sweeney ST
    2. Broadie K
    3. Keane J
    4. Niemann H
    5. O'Kane CJ

    (1995) Targeted expression of tetanus toxin light chain in Drosophila specifically eliminates synaptic transmission and causes behavioral defects

    Neuron 14:341–351.

    https://doi.org/10.1016/0896-6273(95)90290-2

    • PubMed
    • Google Scholar
30. 1. Takemura S-Y
    2. Bharioke A
    3. Lu Z
    4. Nern A
    5. Vitaladevuni S
    6. Rivlin PK
    7. Katz WT
    8. Olbris DJ
    9. Plaza SM
    10. Winston P
    11. Zhao T
    12. Horne JA
    13. Fetter RD
    14. Takemura S
    15. Blazek K
    16. Chang L-A
    17. Ogundeyi O
    18. Saunders MA
    19. Shapiro V
    20. Sigmund C
    21. Rubin GM
    22. Scheffer LK
    23. Meinertzhagen IA
    24. Chklovskii DB

    (2013) A visual motion detection circuit suggested by Drosophila connectomics

    Nature 500:175–181.

    https://doi.org/10.1038/nature12450

    • Google Scholar

Author details

1. Juergen Haag

   Max Planck Institute of Neurobiology, Martinsried, Germany

   Contribution

   JH, Designed the experiments and evaluations, Performed and analyzed the experiments using the telescopic stimulation, Conception and design

   Contributed equally with

   Alexander Arenz

   Competing interests

   No competing interests declared.

2. Alexander Arenz

   Max Planck Institute of Neurobiology, Martinsried, Germany

   Contribution

   AA, Performed and analyzed the experiments using the LED arena, Designed the experiments and evaluations, Conception and design

   Contributed equally with

   Juergen Haag

   Competing interests

   No competing interests declared.

3. Etienne Serbe

   Max Planck Institute of Neurobiology, Martinsried, Germany

   Contribution

   ES, Recorded from tangential cells in T4/T5 block flies

   Competing interests

   No competing interests declared.

4. Fabrizio Gabbiani

   Baylor College of Medicine, Houston, United States

   Contribution

   FG, Supported by a fellowship of the Alexander-von-Humboldt Foundation and participated in the early phase by constructing the telescopic stimulus device

   Competing interests

   No competing interests declared.

5. Alexander Borst

   Max Planck Institute of Neurobiology, Martinsried, Germany

   Contribution

   AB, Designed the experiments and evaluations, Wrote the manuscript with the help of all authors, Conception and design, Drafting or revising the article

   For correspondence

   borst@neuro.mpg.de

   Competing interests

   AB: Reviewing editor, elife

• 3,064

  views

• 554

  downloads

• 84

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.