Postural control is the neural capacity to detect and correct inappropriate body orientation. The larva of Drosophila melanogaster executes a stereotyped self-righting (SR) manoeuvre when turned upside-down. This study combines whole-circuit connectomic reconstruction with targeted functional assays to trace the sensory, interneuronal and motor architecture that implements SR along the antero-posterior axis.
The authors anchored their analysis at a previously implicated motor output—the pair of segmentally repeated lateral transverse motor neurons, LT1/2-MNs—and traced upstream partners through synaptic connectivity. Complementary behavioural perturbations used neuron-specific thermogenetic silencing to evaluate the contribution of identified nodes to SR performance.
Starting from the LT1/2-MNs, the connectomic reconstruction revealed a set of pre-motor interneurons and their principal upstream partners that form the immediate motor pathway for SR. A distinct population of class IV multidendritic sensory neurons emerged as key sensory inputs to this circuit. The identified elements collectively constitute the core components that link sensory detection of body orientation to motor output executed by LT1/2-MNs.
The manuscript reports on the identity and connectivity of these circuit elements along the antero-posterior (AP) axis, documenting how upstream sensory neurons converge onto specific interneurons that, in turn, synapse onto LT1/2-MNs and related pre-motor cells. The reconstruction therefore yields a detailed, segmentally resolved map of sensory-to-motor connections implicated in SR.
To test the functional importance of the anatomically defined circuit, the authors performed neuron-specific thermogenetic silencing of many of the pre-motor, interneuronal, and sensory elements identified in the connectome. Inhibiting the great majority of these elements significantly impaired SR performance, indicating that the traced pathway is necessary for normal self-righting behaviour.
These perturbations provide a causal link between the structural connectome and the behavioural phenotype, supporting the role of the mapped sensory, interneuronal and motor elements in executing SR.
The SR circuit shows substantial connectivity overlap with circuits previously described for nociceptive rolling and touch-induced crawling. Several interneurons act as shared convergence points across these behaviours; the authors highlight DnB, A02o (Wave-1) and TePn05 as shared interneurons receiving diverse sensory inputs and participating in multiple motor programmes. This overlap suggests a conserved or multi-use interneuronal scaffold that routes different sensory modalities into appropriate motor outputs.
Network analyses of the reconstructed connectivity reveal a nested, hub-like organisation within the SR circuit. A relatively small number of high-centrality neurons act as hubs, coordinating input across sensory channels and distributing signals to motor pathways. In addition, a small set of integrator neurons bridge sensory and motor subnetworks, positioning them as potential decision or relay nodes that shape the SR response.
This topology—hub nodes plus integrator bridges—provides an organisational principle that may facilitate rapid and flexible routing of sensory information to motor effectors in a segmentally organised nervous system.
Axial connectivity analyses identify a consistent decline in both synapse number and synapse density toward posterior body segments. This gradient indicates that circuit strength and connectivity complexity change along the AP axis, with anterior segments showing denser synaptic connectivity to the SR network than posterior segments. Such antero-posterior connectivity gradients may influence how different body segments contribute to the coordinated motor output required for self-righting.
Taken together, the connectomic mapping and functional silencing experiments yield what the authors report as the first axial wiring diagram for a postural control circuit in any animal. The findings outline structural principles for segmentally organised motor control: a hub-like network architecture, substantial sharing of interneurons across sensory-motor behaviours, and AP connectivity gradients. The authors propose these principles could extend to other segmented nervous systems beyond Drosophila larvae.
This work is presented as a preprint and has not been peer reviewed. Funding sources are disclosed. The manuscript declares no competing interests. Specific methodological details, numerical synapse counts, or statistical measures beyond those summarised in the abstract were not reported in the source text provided here; readers should consult the full preprint for complete datasets and experimental protocols.