The head direction circuit of two insect species

  1. Ioannis Pisokas  Is a corresponding author
  2. Stanley Heinze
  3. Barbara Webb
  1. School of Informatics, University of Edinburgh, United Kingdom
  2. Lund Vision Group and NanoLund, Lund University, Sweden
16 figures, 3 tables and 1 additional file

Figures

Figure 1 with 1 supplement
Anatomical differences between two species.

There are three apparent differences between the CX of the fruit fly (Drosophila melanogaster) and the desert locust (Schistocerca gregaria). (A, B) The ellipsoid body in the fruit fly has a …

Figure 1—figure supplement 1
Connectivity matrices of the two species.

Connectivity matrices of the two species. The connectivity matrices derived by the exact neuronal projections of (A) the fruit fly (Drosophila melanogaster) and (B) the desert locust (Schistocerca …

Effective connectivity of the inhibitory (Delta7) neurons.

On the top row is the fruit fly circuit, on the bottom row is the locust circuit. In A and D, four examples of how the eight types of Delta7 neurons innervate the PB are illustrated. In both …

Figure 3 with 3 supplements
Projection patterns of the excitatory portion of the fruit fly circuit.

(Ai–Fi) Examples of E-PG (combined E-PG and E-PGT, see Table 3), P-EN and P-EG neurons with their synaptic domains and projection patterns. (Aii–Cii) Step by step derivation of the effective circuit …

Figure 3—figure supplement 1
Neuronal activity across PB glomeruli.

The neuronal activity of P-EN, P-EG and E-PG neurons innervating the glomeruli of the PB for the simulated model of the fruit fly. The activity ‘bump’ is centred around identically numbered …

Figure 3—figure supplement 2
Neuronal projections in the fruit fly.

(A, B) Examples of the projection patterns of E-PG neurons (combined E-PG and E-PGT, see Table 3). (C, D) Examples of P-EN neurons with their synaptic domains and projection patterns (see main text …

Figure 3—video 1
Animation illustrating the operation of the excitatory portion of the fruit fly circuit.
Figure 4 with 3 supplements
Projection patterns of the excitatory portion of the locust circuit.

(Ai–Ei) Examples of E-PG, P-EN and P-EG neurons with their synaptic domains and projection patterns. (Aii–Eii) Step by step derivation of the effective circuit (see main text for a complete …

Figure 4—figure supplement 1
Neuronal activity across PB glomeruli.

The neuronal activity of P-EN, P-EG and E-PG neurons innervating the glomeruli of the PB for the simulated model of the locust. The activity ‘bump’ is centred around identically numbered glomeruli …

Figure 4—figure supplement 2
Neuronal projections in the locust.

Examples of the projection patterns of E-PG (A, B) and P-EN (C, D) neurons in the locust. The anatomy and projection patterns differ from those in the fruit fly (see main text for detailed …

Figure 4—video 1
Animation illustrating the operation of the excitatory portion of the locust circuit.
Combined excitatory and inhibitory portion of the ring attractors.

Explanatory drawings of the connectivity of the inhibitory portion with the excitatory portion of the circuit for the fruit fly (A–C) and the locust (D–F). Each coloured disc represents one or more …

Relative synaptic strengths.

Graphical depiction of the synaptic strengths between classes of neurons. (A,C) For the fruit fly ring attractor circuit. (B,D) For the desert locust ring attractor circuit. Synaptic strengths are …

Figure 7 with 1 supplement
Response to abrupt stimulus changes and tuning curves of neurons.

(A and B) The raster plots of the stimuli used to drive the ring attractor during the simulation are shown on top and the spiking rate activity of each neuron at the bottom. In the beginning of the …

Figure 7—figure supplement 1
Response of spiking and rate-based models to step change of heading.

The mean activity ‘bump’ heading and corresponding standard deviation for the fruit fly and the locust models across time when stimulated with a step change of heading by 180° (80 trials each). (A, B

Transition time and rate of the heading signal.

(A) Time required from the onset of the stimulus until the heading signal settles to its new state. The abscissa (horizontal axis) displays the azimuthal difference between initial and target …

Transition regime as function of inhibitory uniformity.

Heading signal transition regime for (A) the fruit fly ring attractor circuit and (B) the desert locust ring attractor circuit. Blue denotes gradual transition of the heading signal, orange denotes …

Distribution of activity ‘bump’ locations.

The distribution of azimuthal location of the heading signal 3 s after stimulus removal is plotted. On the abscissa (horizontal axis), the azimuth where the stimulus is applied is shown. On the …

Effect of synaptic efficacy heterogeneity on ring attractor stability.

(A) Stability of the ring attractor heading signal for the fruit fly, locust and hybrid-species (fruit fly with localised inhibition) model as a function of heterogeneity in the synaptic efficacies …

Figure 12 with 1 supplement
Effect of membrane parameter heterogeneity on ring attractor stability.

(A) Stability of the ring attractor heading signal for the fruit fly and the locust model when the membrane properties are heterogeneous across the neuronal population. (B, C) Stability of the ring …

Figure 12—figure supplement 1
Effect of cell membrane parameter heterogeneity to transition regime.

The difference in the heading signal transition is present at different amounts of heterogeneity in neuron membrane parameters. As neuronal parameters deviate from their nominal values, from top to …

Response to uni-hemispheric stimulation.

Upper plots show the P-EN stimulation protocol and corresponding induced P-EN activity; lower plots show the response of the ring attractor for (A) the fruit fly circuit and (B) the locust circuit. …

Response to uni-hemispheric stimulation.

Response rate of change of the heading signal with uni-hemispheric stimulation of P-EN neurons. The angular rate of change increases exponentially with stimulation strength and does so most rapidly …

Illustration of Gaussian distribution of synaptic strengths.

The Gaussian distribution of synaptic strengths along synapses located in the PB glomeruli. The synaptic strengths along the PB are illustrated for one Delta7 neuron. The example illustrates the …

Illustration of von Mises distributed stimulus.

The curve demonstrates the relative intensity of the stimulus supplied to neurons innervating each EB tile. In this illustration the stimulus is centred at tile 5.

Tables

Table 1
Characteristics of the activity ‘bump’.

The Full Width at Half Maximum (FWHM), the peak impulse rate of the activity ‘bump’ formed across each family of neurons and the amplitude of the activity ‘bump’ measured as the range of firing …

Neuron classDrosophilaLocust
FWHMPeakAmplitudeFWHMPeakAmplitude
(°)
(imp./s)
(imp./s)
(°)
(imp./s)
(imp./s)
E-PG88.3 ± 0.3161.0 ± 0.2160.1 ± 0.368.3 ± 0.1192.6 ± 0.1192.0 ± 0.2
P-EN80.4 ± 0.4190.1 ± 0.2190.1 ± 0.263.1 ± 0.3153.5 ± 0.1153.5 ± 0.1
P-EG71.0 ± 0.2190.1 ± 0.2190.1 ± 0.263.1 ± 0.3153.5 ± 0.1153.5 ± 0.1
Delta7N/A274.7 ± 0.127.1 ± 0.2101.1 ± 0.2266.6 ± 0.2266.6 ± 0.2
Table 2
Characteristics of the neuron tuning curves.

The Full Width at Half Maximum (FWHM), the peak impulse rate of each family of neurons and the activity amplitude measured as the range of firing rates are shown. Numbers are given as median and …

Neuron classDrosophilaLocust
FWHMPeakAmplitudeFWHMPeakAmplitude
(°)
(imp./s)
(imp./s)
(°)
(imp./s)
(imp./s)
E-PG94.7 ± 4.0208.4 ± 2.3208.2 ± 2.273.4 ± 2.6220.8 ± 1.4220.8 ± 1.4
P-EN74.6 ± 3.8230.3 ± 2.3230.3 ± 2.358.9 ± 3.1163.6 ± 0.9163.6 ± 0.9
P-EG74.6 ± 3.8230.3 ± 2.3230.3 ± 2.358.9 ± 3.1163.6 ± 0.9163.6 ± 0.9
Delta7N/A289.9 ± 1.858.1 ± 4.296.0 ± 3.2265.4 ± 2.9265.4 ± 2.9
Table 3
Neuronal nomenclature.

The names used for the homologous neurons differ between Drosophila and other species. The first column shows the name used in this paper to refer to each group of neurons. The other three columns …

ModelDrosophilaLocust
Neuron nameConsensus nameSystematic name (Wolff and Rubin, 2018)Name
E-PGE-PG and E-PGTPBG1-8.b-EBw.s-D/V GA.b and PBG9.b-EB.P.s-GA-t.bCL1a
P-ENP-ENPBG2-9.s-EBt.b-NO1.bCL2
P-EGP-EGPBG1-9.s-EBt.b-D/V GA.bCL1b
Delta7Delta7 or Δ7PB18.s-GxΔ7Gy.b and PB18.s-9i1i8c.bTB1

Additional files

Download links