The locus coeruleus broadcasts prediction errors across the cortex to promote sensorimotor plasticity
- Friedrich Miescher Institute for Biomedical Research, Switzerland
- Faculty of Sciences, University of Basel, Switzerland
Figures
Figure 1 with 1 supplement
LC axonal calcium activity reflects unsigned visuomotor prediction errors.
(A) Left: Schematic of the two-photon microscope and virtual reality system. Right: Schematics of the two virtual reality conditions. Top: closed loop condition in which visual flow speed (green) is …
Figure 1—figure supplement 1
Additional information and analyses on two-photon imaging of LC axons.
(A) Confocal micrographs showing immunolabelling for tyrosine hydroxylase (magenta) and GCaMP6s (green) in the two LC. GCaMP6s was expressed via stereotactic viral vector injection into only the …
Figure 2 with 1 supplement
LC axonal responses are similar in sensory (posterior) and motor (anterior) cortical areas.
(A) Schematic of the two different locations of the two-photon imaging windows used. Posterior windows were centered on V1 and included surrounding structures, while anterior windows included the …
Figure 2—figure supplement 1
Locomotion onset responses of LC axons.
(A) Responses to the onset of locomotion occurring during the closed loop condition (left), during visual flow in the open loop condition (middle), and during no visual flow in the open loop …
Figure 3 with 2 supplements
LC axon stimulation has only a small effect on stimulus responses in layer 2/3 of V1.
(A) AAV vector injections were used to express ChrimsonR-tdTomato in LC NET-positive neurons, and jGCaMP8m in layer 2/3 neurons of V1. Simultaneous two-photon calcium imaging and optogenetic …
Figure 3—figure supplement 1
Dependence of laser stimulation effects on ChrimsonR-labelled LC axon density.
(A) Laser-evoked change in pupil diameter plotted against ChrimsonR-labelled axon density in V1 (see Methods) for the dataset in Figure 3. Pink shows high ChrimsonR-expressing mice, blue shows low …
Figure 3—figure supplement 2
Classification of functional neuronal types in V1 layer 2/3.
(A) Percentage of neurons as a function of the difference in response to locomotion onsets in closed loop conditions and open loop conditions. Neurons with a responses difference below the 33rd …
Figure 4 with 2 supplements
Phasic LC output enhances sensorimotor plasticity in layer 2/3 of V1.
(A) Diagram of the experiment used to determine whether LC axon stimulation during different visuomotor coupling conditions can modulate plasticity. Visual responses are compared before and after …
Figure 4—figure supplement 1
Additional information and analyses on optogenetic stimulation of LC axons to drive plasticity.
(A) Two example confocal microscopy images from V1 showing jGCaMP8m expression (green) and tdTomato expression (magenta), indicating labelled LC axons. The right image is from an animal classified …
Figure 4—figure supplement 2
Differences in locomotion during stimulation cannot explain differences in the change of locomotion modulation index between groups.
Related to Figure 4. (A) Average locomotion speed (excluding stationary periods) for: High ChrimsonR-expressing mice during closed loop stimulation, Control mice during closed loop stimulation, low …
Figure 5 with 1 supplement
LC axon stimulation over minutes recapitulates the visuomotor plasticity seen over days.
(A) Diagram of the three groups of mice reared with different visuomotor experience. Normally reared mice (NR, gray) were raised with a normal light/dark cycle in their cages, with the full …
Figure 5—figure supplement 1
Comparison of locomotion modulation index across different functional neuronal types in layer 2/3 V1.
Related to Figure 5. (A) Average population responses of layer 2/3 neurons in V1 in coupled trained mice to the onset of visual flow during stationary periods (black), and during locomotion (purple) …
Author response image 1
Correlations between axon density and average effect of laser stimulation on stimulus responses and pupil dilation (data from manuscript Figure 3).
Grey points show control mice, blue points show low ChrimsonR-expressing mice, and purple points show high ChrimsonR- expressing mice.
Author response image 2
Difference in responses to visual stimuli caused by optogenetic stimulation, calculated by subtracting the average response when no laser was presented from the average response when the laser was presented concurrent with the visual stimulus.
Pink traces show the response difference for ChrimsonR-expressing mice, and grey shows the same for control mice. Black blocks below indicate consecutive timepoints after stimulation showing a …
Tables
Author response table 1
| Figure panel | Comparison | Value compared | Mean 1 (delta shown for paired tests) | DS1 | Mean 2 (n/a for paired tests) | SD 2 | P value | Test type | N (FoVs) | N (Mice) |
|---|---|---|---|---|---|---|---|---|---|---|
| Figure 1D | Mismatch vs Playback halt (stationary) | Mean response [% ΔF/F] | 2.25 | 2.28 | -1.08 | 1.71 | <10-4 | Bootstrap | 40, 25 | 13, 10 |
| Figure 1D | Mean response vs zero | Mismatch response [% ΔF/F] | 2.25 | 2.28 | n/a | n/a | <10-4 | Bootstrap | 40 | 13 |
| Figure 1D | Playback halt locomotion vs stationary | Mean response [% ΔF/F] | -1.08 | 1.71 | 0.58 | 3.54 | 0.0348 | Bootstrap | 25, 24 | 10, 10 |
| Figure 1D | Mean response vs zero | Playback halt response (stationary) [% ΔF/F] | -1.08 | 1.71 | n/a | n/a | 0.0072 | Bootstrap | 25 | 10 |
| Figure 1D | Mismatch vs Playback halt response (locomotion) | Mean response [% ΔF/F] | 0.58 | 3.54 | 2.25 | 2.28 | 0.0286 | Bootstrap | 24, 40 | 10, 13 |
| Figure 1D | Mean response vs zero | Playback halt response (locomotion) [% ΔF/F] | 0.58 | 3.54 | n/a | n/a | 0.2314 | Bootstrap | 24 | 10 |
| Figure 1G | Mean response vs zero | Visual flow response (locomotion) [% ΔF/F] | -0.58 | 4.19 | n/a | n/a | 0.2821 | Bootstrap | 21 | 11 |
| Figure 1G | Mean response vs zero | Visual flow response (stationary) [% ΔF/F] | 3.16 | 5.27 | n/a | n/a | 0.0055 | Bootstrap | 28 | 11 |
| Figure 1G | Locomotion vs stationary | Mean visual response [% ΔF/F] | 3.16 | 5.27 | -0.58 | 4.19 | 0.0082 | Bootstrap | 28, 21 | 11, 11 |
| Figure 2B | Posterior vs Anterior | Mismatch response [% ΔF/F] | 2.55 | 2.02 | 1.21 | 2.92 | 0.0922 | Bootstrap | 31, 9 | 9, 4 |
| Figure 2B | Posterior vs Anterior | Playback halt response (stationary) [% ΔF/F] | -1.11 | 1.56 | -1.01 | 2.05 | 0.499 | Bootstrap | 16, 9 | 6,4 |
| Figure 2B | Posterior vs Anterior | Visual flow response (stationary) [% ΔF/F] | 2.98 | 5.46 | 3.53 | 5.14 | 0.4194 | Bootstrap | 19, 9 | 7, 4 |
| Figure 2C | Posterior vs Anterior | Air puff response [% ΔF/F] | 24.73 | 10.03 | 24.48 | 11.24 | 0.4841 | Bootstrap | 13, 8 | 6, 4 |
| Figure 2C | Posterior vs Anterior | Locomotion response (closed loop) [% ΔF/F] | 14.99 | 7.07 | 13.33 | 7.86 | 0.3666 | Bootstrap | 28, 9 | 9, 4 |
| Figure 2C | Posterior vs Anterior | Locomotion response (no visual flow) [% ΔF/F] | 14.42 | 8.05 | 14.96 | 11.93 | 0.4204 | Bootstrap | 8, 6 | 6, 3 |
| Figure 2C | Posterior vs Anterior | Locomotion response (visual flow) [% ΔF/F] | 10.85 | 5.89 | 8.12 | 8.46 | 0.3324 | Bootstrap | 14, 6 | 6, 3 |
| Figure S2B | With vs without visual flow | Locomotion onset response (open loop) [% ΔF/F] | 3.03 | 7.82 | 7.17 | 6.45 | 0.071 | Bootstrap | 20, 14 | 9, 9 |
Additional files
-
Supplementary file 1
Table detailing all statistical tests and sample sizes.
- https://cdn.elifesciences.org/articles/85111/elife-85111-supp1-v1.docx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/85111/elife-85111-mdarchecklist1-v1.docx
