Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 May 22;23(8):2365-2378.
doi: 10.1016/j.celrep.2018.04.079.

Pre-synaptic Muscarinic Excitation Enhances the Discrimination of Looming Stimuli in a Collision-Detection Neuron

Ying Zhu  1 , Richard B Dewell  2 , Hongxia Wang  2 , Fabrizio Gabbiani  3
Affiliations

Pre-synaptic Muscarinic Excitation Enhances the Discrimination of Looming Stimuli in a Collision-Detection Neuron

Ying Zhu et al. Cell Rep. .

Abstract

Visual neurons that track objects on a collision course are often finely tuned to their target stimuli because this is critical for survival. The presynaptic neural networks converging on these neurons and their role in tuning them remain poorly understood. We took advantage of well-known characteristics of one such neuron in the grasshopper visual system to investigate the properties of its presynaptic input network. We find the structure more complex than hitherto realized. In addition to dynamic lateral inhibition used to filter out background motion, presynaptic circuits include normalizing inhibition and excitatory interactions mediated by muscarinic acetylcholine receptors. These interactions preferentially boost responses to coherently expanding visual stimuli generated by colliding objects, as opposed to spatially incoherent controls, helping to discriminate between them. Hence, in addition to active dendritic conductances within collision-detecting neurons, multiple layers of inhibitory and excitatory presynaptic connections are needed to finely tune neural circuits for collision detection.

Keywords: DCMD; LGMD; collision avoidance; divisive normalization; lateral excitation; looming; muscarine; scopolamine.

PubMed Disclaimer

Conflict of interest statement

DECLARATION OF INTERESTS

The authors declare no competing interests.

Figures

Figure 1
Figure 1. Spatio-temporal Activation of LGMD Excitatory Inputs to Looming Stimuli
(A) a: to map dendritic receptive fields, brief square flashes were presented at 5 locations on the display. 3 colors represent flashes with different distances from the display center. b: a looming stimulus was then presented. It expanded from the center square and eventually covered locations where the other flashes were presented. c: LGMD excitatory dendritic field. Dendritic branches activated by the 5 flashes are shown in red, green, and blue. (B) The black dashed line indicates the start of the stimulus. a: angular size time course (l/|v| = 120 ms). Stimulus parameters were as follows: radius l, speed v, and angular size 2θ (schematics). b: average relative fluorescence change at the center branch (red) and two sets of branches symmetrically surrounding it (green and blue) in one animal. Time is relative to maximal stimulus size (80 ms before projected collision). c: mean (solid lines) and SD (shaded areas, n = 5; n, number of animals) of average dF/F (normalized to maximum dF/F in response to center square flash for each animal) at the center branch (red) and at two surrounding branch sets (green and blue). b and c: red, green, and blue rectangles (arrowheads) represent the time during which the stimulus edges moved across locations activated by red, green, and blue flashes. Over the white intermediate regions, the stimulus edge would be within the receptive fields of branches on either side. (C) The same as (B) but enlarged. The black dashed line indicates the time when the stimulus reached its maximum size; the dashed blue line represents the projected collision time. The red double arrow indicates the time interval during which the stimulus edge activates both the red and green branches (A, c) because of overlapped activation (Zhu and Gabbiani, 2016). (D) a: LGMD excitatory dendritic field with 10 selected dendritic branches indicated in colors. b: time course of average dF/F at each selected branch of a. The horizontal black dashed line is threshold for branch activation. (E) a: the y axis represents the time when the rising phase of each trace in (D), b, crosses the threshold. The branch with the earliest rise time at threshold is the looming center. The x axis represents the distance in the imaging plane from each dendritic branch to the looming center. The colors match those in (D). b: the same as a across 5 animals. All dendritic branches are plotted in the same color for each animal. (F) a: time-integrated dF/F for each dendritic branch in (D) as a function of distance from the looming center. b: the same as a across 5 animals. Data from each animal are plotted in a different color. See also Figure S1.
Figure 2
Figure 2. Comparison of Calcium Responses and Subthreshold Vm
(A) LGMD excitatory dendritic field with selected dendritic branches in different colors. White dashed lines indicate intracellular electrodes. (B) a: time course of stimulus angular size (l/|v| = 120 ms). b: LGMD dendritic membrane potential (Vm; black line). Gray shading, SD (4 trials). Thin colored lines represent time courses of average dF/F within the selected dendritic branches (A). (C) Enlarged view of (B). dF/F traces were normalized to their peaks. (D) Cross-correlation (xcorr) of the Vm rising phase and dF/F. The colors match (B). The black dashed line indicates zero-time lag. (E) Enlarged view of (D). (F) The peak of each cross-correlation in (D) as a function of peak dF/F for each dendritic branch in (B). (G) The same as (F) across animals. (H) The peak of each cross-correlation trace in (D) as a function of its branch distance to the looming center. (I) The same as (H) across animals. (J) Time lag at the peak of each cross-correlation trace plotted across animals as a function of its branch distance to the looming center. (K) Peak cross-correlation as a function of peak time lag across animals. In (G) and (I)–(K), n = 7, Vm was recorded at 11 locations. See also Figure S2.
Figure 3
Figure 3. Effect of Muscarine and Scopolamine on Looming Responses
(A) a: LGMD activation to looming, shown as peak dF/F at each activated pixel (color scale, right; average over 3 trials in 1 animal). Blue and yellow dashed arrows indicate the first and second principal axes. b: the same after puffing muscarine (average over 3 trials). (B) LGMD excitatory dendritic field with selected dendritic branches indicated by different colors. (C) a: stimulus angular size time course (l/|v| = 120 ms). b: average dF/F in response to this stimulus at each selected dendritic branch (B) in one animal. (D) The same as (C) after puffing muscarine. (E) Peak amplitude of dF/F on each dendritic branch after versus before muscarine (p = 4.84 × 10−10, paired t test). (F) a: projection of peak dF/F onto the first principal axis as a function of distance along the axis after normalizing to peak value without (blue) and with muscarine (red). Red and blue shadings, SD. FWHM before muscarine, 148.7 ± 13.3 μm (mean ± SD); after: 252.2 ± 13.2 μm (p = 0.0002, paired t test, before versus after). b: the same analysis but along the second principal axis. FWHM before muscarine: 61.2 ± 15.7 μm; after: 90.6 ± 18.2 μm (p = 0.0016). (G) Peak amplitude of dF/F on each dendritic branch as a function of branch distance to the looming center before (control) and after puffing muscarine (n = 5). Corresponding lines are linear regressions. Line slopes are significantly different (38 data points, p = 0.023, ANCOVA). (H) Rise time at threshold (as in Figure 1D) after versus before puffing muscarine (p = 0.31, paired t test). (I) a: LGMD activation to looming, shown as peak dF/F at each activated pixel (color scale, right; average over 3 trials in 1 animal). b: the same after puffing scopolamine (average over 3 trials). (J) LGMD excitatory dendritic field with selected dendritic branches indicated by different colors. (K) a: stimulus angular size time course (l/|v| = 120 ms). b: average dF/F at selected dendritic branches, with colors matched to (J). (L) The same as (K) after puffing scopolamine. In (K) and (L), the vertical dashed lines indicate the time of maximal stimulus size. (M) Peak dF/F amplitude on selected dendritic branches after versus before puffing scopolamine (p = 2.8 × 10−15, paired t test). (N) The same as (F) except that scopolamine was puffed. a: FWHM before scopolamine application: 168.3 ± 39.8 μm (mean ± SD); after: 141.2 ± 62.9 μm (p = 0.21, paired t test, before versus after). b, FWHM before scopolamine: 72.9 ± 13.1 μm; after: 60.8 ± 31.2 μm (p = 0.36). (O) The same as (G) except that scopolamine was puffed (53 data points, p = 0.22, ANCOVA). (P) Rise time at threshold (defined as in Figure 1D) after versus before puffing scopolamine (p = 1.92 × 10−8, paired t test). In (C) and (D), the vertical dashed line indicates the time of maximal stimulus size. In (E)–(H), n = 5. In (M)–(P), n = 7. See also Figure S3.
Figure 4
Figure 4. Test of the Influence of Scopolamine on Lateral Inhibition
(A) a: stimulus schematics (control). b: stimulation time course. Time 0–ta, bright display; ta, stimulus on at right display edge; ta–tb, stimulus stays still; tb–tc, stimulus translates from right to left until it moves out of display. (B) a: stimulus schematics (lateral inhibition); flanking gratings temporal frequency, 6 luminance transitions per second, 90°/s. Grating width, 25°. sep., separation. b: stimulation time course. Time 0–ta, bright display; ta, stimulus on at right display edge, gratings on; ta–tb, stimulus stays still; tb–tc, gratings drift; tb+0.2 s–tc, stimulus translates from right to left until it moves out of display; tc, gratings stop drifting. (C) a: dendritic activation to translating stimulus. b: dendritic activation to the same stimulus after puffing scopolamine. (D) LGMD excitatory dendritic field with selected dendritic branches indicated by different colors. (E) a: average dF/F time course (over 3 trials) at each dendritic branch in (D) to the stimulus as in (A). Vertical dashed lines: black, time when stimulus appears; red, time when stimulus starts translating; blue, time when stimulus moves out of display. The asterisk and plus markers on each curve indicate calcium baseline and peak dF/F to the translating stimulus, respectively. b: dF/F amplitude computed by subtracting baseline (asterisks) from peak dF/F (plus symbols) for each curve in a (matched color) and plotted as a function of peak dF/F time. (F) a: average dF/F time course (over 3 trials) at each dendritic branch in (D) to the stimulus as in (B), with a 50° grating separation. Vertical dashed lines: black, time when stimulus and lateral gratings appear; red, time when lateral gratings start drifting (0.2 s before the small stimulus starts translating); blue, time when small visual stimulus moves outside of display and lateral gratings stop drifting. Asterisks and plus symbols as in (E). b: the same as (E), b. (G) The same as (E) after puffing scopolamine. (H) The same as (F) after puffing scopolamine. (I) dF/F amplitude on every selected dendritic branch after versus before puffing scopolamine (41 dendritic branches total, average of 3 trials per condition). Purple dots: translating stimuli, no drifting gratings. Green, light green, red, and blue dots: translating stimuli and lateral drifting gratings separated by 70°, 50°, 30°, and 7.5°, respectively. Purple, green, light green, red, and blue diamonds represent the mean (center of mass) of matched color dots, respectively. (J) Peak dF/F time on every selected dendritic branch after versus before puffing scopolamine with the same plotting conventions as in (I). (K) Boxplots of mean dF/F amplitudes for all branches in each animal to translating stimuli without and with drifting gratings. Blue and magenta, after and before puffing scopolamine, respectively (*p < 0.05, Wilcoxon signed-rank test, uncorrected for multiple comparisons; p = 0.025 for control and p = 0.0001 for scopolamine, one-way ANOVA; p = 0 for control versus scopolamine, two-way ANOVA). In this and subsequent boxplots, the central bar indicates the median, the top and bottom box edges indicate the 25th and 75th percentiles, the whiskers represent the extent of the data within 1.5 times the box height, and data points outside of this range (outliers) are plotted individually as crosses. (L) Boxplots of the difference of mean amplitudes for all branches in each animal after versus before puffing scopolamine to translating stimuli without and with drifting gratings. No significant statistical difference of means was found between any two groups (p = 0.66 one-way ANOVA). In (I)–(L), n = 5. See also Figure S4.
Figure 5
Figure 5. Calcium Responses to Looming Stimuli Restricted to Horizontal and Vertical Bands
(A) a: from left to right, schematics of looming stimulus, HBR, and VBR looming stimuli, respectively. b: from left to right, LGMD dendritic field activation to the same stimuli. Red and blue arrows on left indicate the first and second principal axis, respectively. (B) a: 8 selected regions of interest (ROIs) represented by different colors in an example LGMD. The white arrow points to the looming center. The red arrow indicates the first principal axis of the looming response. b: peak dF/F in the 8 ROIs to SL (black crosses), HBR (red dots), and VBR (blue circles) stimuli. Solid lines indicate linear regressions (n = 5; **p < 0.01; n.s., not significant [i.e., p > 0.05], paired t test, ANCOVA test of slopes). (C) Projection of peak dF/F onto the first (a) and second (b) principal axes as a function of distance along the axis after normalizing to peak value in response to SL (black), HBR (red), and VBR (blue) stimuli, respectively. Red, black, and blue shadings, SD (n = 10). (D) Boxplots of the full width at half-maximum (FWHM) of the curves in (C) along the first (a) and second principal axis (b) to the SL or the HBR/VBR stimuli (n = 10, *p < 0.05, **p < 0.01, paired t test). Box plotting conventions are described in the legend to Figure 4K. See also Figures S5 and S6.
Figure 6
Figure 6. Effect of Scopolamine and Muscarine on Calcium Responses to Flicker-Off Stimuli
(A) a: schematics of flicker-off stimuli with diameter D. b: time course of visual stimulus. Time 0–2 s, bright display; after 2 s, black stimulus flashed on at center of display. (B) LGMD activation to flicker-off stimuli of increasing size. (C) Boxplots of peak dF/F for the stimulus center branch to flicker-off stimuli of increasing size. Magenta and blue represent without and with scopolamine, respectively (n = 5; p = 0.0007 for control versus scopolamine, two-way ANOVA; p = 0.028 for control, p = 0.069 for scopolamine, one-way ANOVA). (D) Boxplots of mean peak dF/F over all branches to flicker-off stimuli with increasing size. Magenta and blue are as in (C) (n = 5; p = 0.0009 for control versus scopolamine, two-way ANOVA; p = 0.07 for control, p = 0.35 for scopolamine, one-way ANOVA). (E) Boxplots of peak dF/F for the stimulus center branch to flicker-off stimuli with increasing size. Magenta and blue represent without and with muscarine (n = 6; p = 0 for control versus muscarine, two-way ANOVA; p = 0.098 for control, p = 0.073 for muscarine, one-way ANOVA). Plotting conventions are as in (D). (F) Boxplots of mean peak dF/F over all branches to flicker-off stimuli with increasing size. Magenta and blue are as in (E) (n = 6; p = 0 for control versus muscarine, two-way ANOVA; p = 0.053 for control, p = 0.001 for muscarine, one-way ANOVA). In (C)–(F), *p < 0.05, **p < 0.01, n.s. not significant, paired t test. Box plotting conventions are described in the legend to Figure 4K.
Figure 7
Figure 7. Scopolamine Reduced Coherence Preference and Schematic Models of Candidate Neural Circuits Presynaptic to Dendritic Field A of the LGMD
(A) Illustration of coherent and incoherent stimuli. For coarse looms, grayscale levels are set so that the luminance in each coarse pixel is equal to that of standard looms in every frame. For reduced coherence looms, the spatial locations of coarse pixels were altered (Experimental Procedures). (B) LGMD’s response to looming stimuli (l/|v| = 50 ms) decreased after addition of scopolamine. The lines and shaded region indicate mean and ± 1 SD. (C) LGMD’s response to a spatially randomized stimulus also decreased after addition of scopolamine. Plotting conventions are as in (B). (D) LGMD responses increase with stimulus coherence (Pearson r = 0.90, p = 0.006), but this coherence preference decreased after scopolamine (p = 0.008, ANCOVA test of slopes). Thin lines indicate data from individual animals. Thick lines and points represent population average. Error bars indicate ± 1 mad. (E) Scopolamine caused a decrease in coherence preference (calculated from the slopes of the lines in C; p = 0.022, paired t test). There is an average decrease from 0.28 spikes to 0.07 spikes per percent coherence. Gray lines connect individual data points, and dots and whiskers indicate mean ± 1 SD. (F) Overview of the pathways impinging on the LGMD. Feedforward inhibition impinging on two other dendritic fields was not studied here. For feedforward excitation onto the excitatory dendritic field, colored dots illustrate retinotopy. Ph, photoreceptors; LMC, lamina large monopolar cells; TmA, transmedullary afferents. (G) Earlier model with lateral inhibition presynaptic to the LGMD. The LGMD dendrites are shown at the bottom, TmAs in the center, and axonal terminals of lamina neurons at the top. The synapses of axonal terminals of adjacent TmAs are connected through inhibitory synapses (red arrows). (H) Current model with lateral excitation presynaptic to the LGMD, normalizing inhibition, and upstream short-range lateral inhibition, with the same conventions as in (G). The exact locations of short-range lateral inhibition and global inhibition were not determined here, except that they are necessarily located upstream of the local lateral excitation. In (B)–(E), n = 5.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Behnia R, Clark DA, Carter AG, Clandinin TR, Desplan C. Processing properties of ON and OFF pathways for Drosophila motion detection. Nature. 2014;512:427–430. - PMC - PubMed
    1. Brown DA. Muscarinic acetylcholine receptors (mAChRs) in the nervous system: some functions and mechanisms. J Mol Neurosci. 2010;41:340–346. - PubMed
    1. Carandini M, Heeger DJ. Normalization as a canonical neural computation. Nat Rev Neurosci. 2011;13:51–62. - PMC - PubMed
    1. Caulfield MP. Muscarinic receptors–characterization, coupling and function. Pharmacol Ther. 1993;58:319–379. - PubMed
    1. Collin C, Hauser F, Gonzalez de Valdivia E, Li S, Reisenberger J, Carlsen EM, Khan Z, Hansen NO, Puhm F, Søndergaard L, et al. Two types of muscarinic acetylcholine receptors in Drosophila and other arthropods. Cell Mol Life Sci. 2013;70:3231–3242. - PMC - PubMed

Publication types

LinkOut - more resources