Networks of descending neurons transform command-like signals into population-based behavioral control

Jonas Braun, Femke Hurtak, Sibo Wang, Pavan P RamdyabioRxiv 2023 · DOI

TL;DR

This study reveals that "command-like" neurons in fruit flies don't act alone but instead recruit larger networks of other descending neurons (DNs) to generate complex behaviors. This recruitment is crucial for flexible actions, while simpler, stereotyped movements can be driven by individual neurons. This work bridges the gap between simple command signals and complex behavioral output by proposing a hierarchical recruitment model.

What Problem Does This Paper Address?

The brain needs to translate intentions into actions, a process that involves descending neurons (DNs) sending signals from the brain to motor circuits. Some DNs, termed "command-like," appear capable of driving entire behaviors with just a few neurons. However, this observation seemed to conflict with findings that many DNs become active during complex behaviors. The authors sought to reconcile these two views: how can a few neurons be sufficient for a behavior, yet many are active during it? They propose that command-like DNs might act as orchestrators, recruiting other DNs to form larger networks that ultimately control behavior.

Methods

The researchers employed a multi-faceted approach using Drosophila melanogaster:

  • Optogenetics: They used light to activate specific "command-like" DNs (DNp09 for forward walking, aDN2 for grooming, and MDN for backward walking) while simultaneously recording neural activity.
  • Calcium Imaging: They used genetically encoded calcium indicators (GCaMP6s) to visualize the activity of populations of DNs in the gnathal ganglia (GNG), a key brain region for DNs.
  • Connectome Analysis: They leveraged a detailed electron microscopy-based connectome of the Drosophila brain to map the direct synaptic connections between DNs.
  • Behavioral Analysis: They meticulously tracked fly movements using cameras and pose estimation algorithms to quantify behaviors like walking, grooming, and turning.
  • Headless Fly Experiments: To determine the necessity of downstream DN networks, they decapitated flies and then stimulated command-like DNs, observing if the behavior could still be executed without brain-based DN recruitment.

Datasets

The study primarily utilized the female adult fly brain (FAFB) connectome dataset (version hosted on Codex as of August 3, 2023, FlyWire materialization snapshot 630). This dataset provided detailed information on neuronal morphology, connectivity, and classification.

Key Insights and Ideas

  • Command-like DNs are Orchestrators, Not Soloists: The most striking finding is that activating "command-like" DNs doesn't just send a direct signal to the motor system. Instead, it recruits a cascade of other DNs, forming larger networks that are essential for flexible behaviors.
  • A Continuum of Control: DNs exist on a spectrum from "broadcasters" (like DNp09) that recruit many downstream partners for flexible behaviors, to "stand-alone" neurons (like MDN) that can drive stereotyped movements independently.
  • Network Size Predicts Flexibility: The number of downstream DNs a command-like neuron connects to is a strong predictor of whether that neuron drives a flexible behavior (requiring network recruitment) or a stereotyped one (where network recruitment is dispensable).
  • Behavioral Clusters and Inhibition: DN networks are organized into behavior-specific clusters that tend to inhibit each other. This suggests a mechanism for action selection, preventing conflicting behaviors from occurring simultaneously.

Key Results

  • Optogenetic activation of command-like DNs consistently recruited larger populations of GNG-DNs, with the degree of recruitment varying between DN types.
  • Connectome analysis revealed direct excitatory connections between command-like DNs and their downstream partners, explaining the observed functional recruitment.
  • Headless fly experiments demonstrated that while stereotyped behaviors (like backward walking driven by MDN) were retained, flexible behaviors (like forward walking driven by DNp09) were lost, highlighting the necessity of downstream DN networks for flexibility.
  • DNs with more downstream partners were associated with more flexible behaviors, while those with fewer partners drove more stereotyped actions.
  • DN networks were found to form distinct clusters, with excitatory connections within clusters and inhibitory connections between them, suggesting a functional organization for behavior control.

Limitations and Future Work

  • Temporal Resolution: Calcium imaging has limited temporal resolution, making it difficult to distinguish between direct monosynaptic and polysynaptic recruitment of DNs. Future work could employ faster imaging techniques or electrophysiology.
  • Ascending Neuron Contribution: The headless fly experiments cannot fully disentangle the role of descending neurons from ascending neurons, which also play a role in integrating motor information. Future studies could explore the interplay between DNs and ascending neurons.
  • Connectome Completeness: While powerful, the connectome is a static map. Future work could integrate dynamic activity data with the connectome to build more comprehensive models of neural circuits.
  • Generalizability: The study focused on Drosophila. Investigating similar mechanisms in other species, including mammals, would be valuable to understand the broader principles of descending control.
  • New Behaviors: The authors suggest that this hierarchical recruitment model provides a substrate for the evolution of new behaviors through the novel coupling or uncoupling of DNs, a direction for future research.

Summary updated

Abstract

Abstract To transform intentions into actions, movement instructions must pass from the brain to downstream motor circuits through descending neurons (DNs). These include small sets of command-like neurons that are sufficient to drive behaviors—the circuit mechanisms for which remain unclear. Here, we show that command-like DNs in Drosophila directly recruit networks of additional DNs to orchestrate flexible behaviors. Specifically, we found that optogenetic activation of command-like DNs previously thought to drive behaviors alone in fact co-activate larger populations of DNs. Connectome analysis revealed that this functional recruitment can be explained by direct excitatory connections between command-like DNs and networks of interconnected DNs in the brain. The size of downstream DN networks is predictive of whether descending population recruitment is necessary to generate a complete behavior: DNs with many downstream descending partners require network recruitment to drive flexible behaviors, while neurons with fewer partners can alone drive stereotyped behaviors and simple movements. Finally, DN networks reside within behavior-specific clusters that inhibit one another. These results support a mechanism for command-like descending control whereby a continuum of stereotyped to flexible behaviors are generated through the recruitment of increasingly large DN networks which likely construct a complete behavior by combining multiple motor subroutines.
BibTeX
@misc{braun2023networks,
  author = {Jonas Braun and Femke Hurtak and Sibo Wang and Pavan P Ramdya},
  title = {{Networks of descending neurons transform command-like signals into population-based behavioral control}},
  year = {2023},
  doi = {10.1101/2023.09.11.557103},
  howpublished = {bioRxiv},
}

Loading related content...

Figure