Профиль

Thanks Tyler!!

We are extremely grateful to the #maleCNS team, incl. @janelia-flyem.bsky.social, @jefferis.bsky.social and lab, and many, many others. Backstory >> bsky.app/profile/jeff... Website >> male-cns.janelia.org

We, in this case, are stellar postdocs and shared first-authors @sanderliessem.bsky.social and @skasin.bsky.social with a team from @uni-wuerzburg.de, @hhmijanelia.bsky.social and @columbiauniversity.bsky.social in a collaboration between the Ache and Card labs & friends.

Overall, DNs form parallel, loosely-overlapping ensembles that span a continuum from command-like neurons to synergistic population codes. Upstream, distinct combinations of sensory feature detectors differentially recruit DN ensembles to enable flexible, context-dependent behavioral control.

5th, it turns out - very! Based on their output similarity in the VNC, we were able to cluster all 1300 DNs into functional ensembles covering the entire behavioral space of the fly. DNs previously associated with specific behaviors clustered together. So these clusters align with specific behaviors

At this point, we’ve shown that landing and takeoff are controlled by separate, parallel populations of neurons on each sensorimotor level - from feature detection to pre-motor assemblies. Despite being triggered by the same visual stimulus. How widespread is this parallel processing architecture?

Despite relying on a different set of sensory feature detectors, landing DNs are responsive to frontal looming stimuli. But their response differs from that of takeoff DNs, showing that the networks upstream of each DN ensemble encode different parameters of the same stimulus.

4th, we asked whether the variants of the landing movements driven by different DNs are matched to their sensory tuning - it turns out they are! Different Landing DNs receive visual input from different spatial directions, and the sensory tuning matches the movements they drive.

This makes sense, functionally, because landing is only relevant in the context of flight, and the landing motor pattern would disrupt other actions, such as walking or courtship, when evoked in the wrong context.

Our connectome analysis predicts functional connectivity, which we verify by patching/optogenetic activation experiments in behaving flies. Importantly, though, actual connectivity is heavily state-dependent, so that sensory activity gets routed into landing pathways only during flight.

3rd we went upstream to ask how DNs get recruited by sensory systems. We identified all visual inputs to landing and takeoff DN ensembles - and find they are recruited by separate, parallel sets of visual feature detectors. And core input neurons to the landing pathway drive landing phenotypes.

One neat little feat: within each ensemble, individual DNs make unique connections to motor neurons via specific interneurons, which enable variations of the core motor pattern - leading to landing movements with unique leg kinematics, such as reaching overhead, suitable for landing on the ceiling.

2nd we identified the Core Motor Assemblies linking DN ensembles to motor neurons. These assemblies predict movement sequences driven by DNs, such as extensions of all legs for landing through inhibition of leg flexors and excitation of extensors. Again, circuits for landing and takeoff > separated.

Neurons within each ensemble drive similar movements of the entire body, which we confirmed using optogenetics and behavioral tracking.

Видео

1st we analyzed pathways for landing and takeoff - behaviors elicited by the same visual stimulus that involve different body movements. Comparing output connectivity of all descending neurons (DNs) to known DNs for landing and takeoff, we identify separate DN ensembles controlling each behavior.

Thank you @katrinvogt.bsky.social! I had a great time with all the impressive science going on in Konstanz!

We thank @dfg.de and the #NeuroNex program for generously funding our work, and our neighbours and colleagues at @uni-wuerzburg.de, especially the Department of Neurobiology and Genetics (Charlotte Förster, Chris Wegener, and @silkesachse.bsky.social). Stay tuned - more walking papers on the way...

the beatles are walking across a zebra crossing on a city street .ALT: the beatles are walking across a zebra crossing on a city street .media.tenor.com

This study was a team effort, with some surprises and directional changes on the way. I am grateful to the entire crew, especially Stefan & Sander, Fathima, Feffo, Mert, and Aleyna from my lab, Adrián and @postpop.bsky.social from the Clemens Lab, and Axel and Ansgar from the Büschges lab.

Both MDN and DopaMeander are gated out (switched off) during flight. Showing that behavioral state-dependent gating occurs simultaneously across levels of motor control and is bidirectional in nature, allowing the specific boosting of motor control modules in the appropriate behavioral context.

Finally, we leveraged the fact that we identified neurons controlling walking across hierarchical levels to ask how the brain ensures that appropriate motor output is generated in different behavioral states. To do so, we recorded MDN and DopaMeander in flies transitioning into and out of flight.

Видео