Hox Activity Levels Governs the Evolution of Homologous Behaviors

Hox Activity Levels Governs the Evolution of Homologous Behaviors. | Research Square Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Help Center Sign In Submit a Preprint Cite Share Download PDF Article Hox Activity Levels Governs the Evolution of Homologous Behaviors. Jimena Berni, Jesús Rodriguez Curt, Meg Sambrook, Sara Deppieri, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9072454/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Central neuronal network rewiring is a key mechanism for the evolution of distinct behaviors. However, elucidating the genetic processes that have taken place to reconfigure circuits has rarely been performed. Here, taking advantage of the diversification of locomotor behaviors along the body of a single animal, we demonstrate how changes in the levels of the conserved developmental Hox genes govern the assembly of distinct circuits generating specific motor patterns. We show how these genetic changes occur during the late stages of development, refine the neuronal morphologies of a subset of neurons and lead to their incorporation into different circuits ultimately generating new function and behavior. Our work therefore uncovers a mechanism whereby evolutionary selected deviations in the expression levels of specification genes, refines connectivity in an otherwise unchanged network allowing for the evolution of a new behavior while ensuring system stability. Biological sciences/Developmental biology/Differentiation Biological sciences/Evolution Biological sciences/Neuroscience/Neural circuits Biological sciences/Neuroscience/Motor control Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Main text The extraordinary diversity of behaviors we observe today across or within species, has emerged through the evolution of the neural circuits that generate them. A mechanism that is beginning to be explored is the remodeling of central circuits through synaptic connectivity changes (aka rewiring) 1-6 . Working with simple nervous systems, such as the fruit fly or nematodes, has permitted identification of the neurons as well as the characterization of the synaptic changes that support distinct behavioral outputs. For example, the mating disinterest of Drosophila simulans males to females of the sister species Drosophila melanogaster correlates with a change in the excitatory and inhibitory balance onto a courtship promoting neuron, which is present in both species 1 . However, understanding the genetic changes, selected through evolution, that have induced circuit remodeling and shaped the diversification of function and behavior is still to be achieved. The Hox genes are conserved master regulators of regional identity along the body in most animals, including humans. Evidence of their necessity for the diversification of body regions, both across species but also within the same species, have been well documented since the discovery of the development of an extra pair of wings in the Drosophila Ubx mutant 7,8 . In the vertebrate nervous system, we know that distinct patterns of Hox expression are necessary and sufficient for the assembly of the respiratory network 9,10 . A range of insightful studies have shown how the spatial and temporal patterns of Hox gene expression dictate the formation of along the spinal cord 11-14 and the appropriate connections with muscles 15-17 . Furthermore, distinct Hox patterning has been correlated with distinct locomotor outputs 18,19 . Therefore, the Hox genes are prime candidates to be driving central circuit rewiring to support the evolution of new behaviours. In here, we used the locomotor behavior of the fruit fly Drosophila melanogaster larva as a model for behavioral evolution. The larvae explore their environment using two discrete behaviors: crawls and turns. Each one is autonomously generated by central circuits located at distinct levels along the ventral nerve cord 20-22 . Left-right symmetrical waves of neuronal activity propagating anteriorly or posteriorly along the abdominal(a) nervous system (neuromeres) generate forward and backward crawls respectively. Forward and Backward crawls differ in the neuromere they originate from (a1 or a8) and also involve specific neurons that differentially control the wave propagation and the sequence of muscles contractions in each segment 23-27 . Turns, are generated by asymmetrical neuronal activity in the anterior segments, from thorax to a4 (Fig.1C) 21,28,29 . Despite the operational diversity between anterior and posterior abdominal segments, each neuromere arises from lineages of neuronal progenitors that are largely invariant 30-32 . Thus, this system offers the unique opportunity to investigate the role of neuronal circuit rewiring and its genetic control for the evolution of a novel locomotor behavior. The extraordinary diversity of behaviors we observe today across or within species, has emerged through the evolution of the neural circuits that generate them. A mechanism that is beginning to be explored is the remodeling of central circuits through synaptic connectivity changes (aka rewiring) 1-6 . Working with simple nervous systems, such as the fruit fly or nematodes, has permitted identification of the neurons as well as the characterization of the synaptic changes that support distinct behavioral outputs. For example, the mating disinterest of Drosophila simulans males to females of the sister species Drosophila melanogaster correlates with a change in the excitatory and inhibitory balance onto a courtship promoting neuron, which is present in both species 1 . However, understanding the genetic changes, selected through evolution, that have induced circuit remodeling and shaped the diversification of function and behavior is still to be achieved. The Hox genes are conserved master regulators of regional identity along the body in most animals, including humans. Evidence of their necessity for the diversification of body regions, both across species but also within the same species, have been well documented since the discovery of the development of an extra pair of wings in the Drosophila Ubx mutant 7,8 . In the vertebrate nervous system, we know that distinct patterns of Hox expression are necessary and sufficient for the assembly of the respiratory network 9,10 . A range of insightful studies have shown how the spatial and temporal patterns of Hox gene expression dictate the formation of along the spinal cord 11-14 and the appropriate connections with muscles 15-17 . Furthermore, distinct Hox patterning has been correlated with distinct locomotor outputs 18,19 . Therefore, the Hox genes are prime candidates to be driving central circuit rewiring to support the evolution of new behaviours. In here, we used the locomotor behavior of the fruit fly Drosophila melanogaster larva as a model for behavioral evolution. The larvae explore their environment using two discrete behaviors: crawls and turns. Each one is autonomously generated by central circuits located at distinct levels along the ventral nerve cord 20-22 . Left-right symmetrical waves of neuronal activity propagating anteriorly or posteriorly along the abdominal(a) nervous system (neuromeres) generate forward and backward crawls respectively. Forward and Backward crawls differ in the neuromere they originate from (a1 or a8) and also involve specific neurons that differentially control the wave propagation and the sequence of muscles contractions in each segment 23-27 . Turns, are generated by asymmetrical neuronal activity in the anterior segments, from thorax to a4 (Fig.1C) 21,28,29 . Despite the operational diversity between anterior and posterior abdominal segments, each neuromere arises from lineages of neuronal progenitors that are largely invariant 30-32 . Thus, this system offers the unique opportunity to investigate the role of neuronal circuit rewiring and its genetic control for the evolution of a novel locomotor behavior. Differences in neuronal morphology in a subset of interneurons underlie the specialization of function. Given that the number and identity of interneurons is largely invariant across the abdominal segments a2-a7, with fewer neurons in a8 (Fig. 1a, b and Supplementary Fig. 1), we analyzed the regional differences in neuronal morphologies. Using extensive labelling of individual neurons from the Janelia collection 33 in third instar larvae with the multi-colour flip out method 34 we obtained over 20,000 images that we used to build an atlas (mini-atlas) of neural system diversity. We analysed the progeny of six neuronal progenitors, called Neuroblasts (NBs), that produce 43 of 270 abdominal interneuron types and are amenable to genetic targeting 35 (Fig. 1 and mini-atlas).  Segment-specific morphological differences were defined by a series of conservative criteria. The presence/absence or change of direction of a primary dendritic or axonal branch was considered a change and was classified as posterior, anterior or commissural change. The length of a process was considered different only if it crossed to a new segment, whereas neurons whose dendritic shape changed in the same region of the neuromere were considered invariant (Fig. 1). We found that the shape of more than half of the interneurons is invariant across abdominal segments (Figure 1a and c), suggesting that any diversity in the abdominal nervous system is incorporated into a basic segmental neuromere rather than each segment being completely reorganised. Diversity in segmentally variant interneurons mostly arises from differences in the rostro-caudal direction and length of dendrites and axons or both, or in a few cases, in the differential presence of commissural projections (Fig. 2c; Supplementary Mini-atlas). The caudal shift of dendritic processes is the most common phenotype and consists of a gradual transition that occurs from anterior to posterior segments (Fig. 2b). For example, A23e (see method for neuron naming details) dendrites project mostly rostrally in a2. In a3 to a5, the dendritic tree extends rostral and caudally with branches of similar length but more lateral. In a6, the dendrites are mostly extending caudally (Fig. 1b). The dendritic shift could be accompanied by a shift in the axon in the same direction, as observed in A23e, or by no axonal change (i.e., A18b2 Supplementary Mini-atlas). Other neurons showed an abrupt change in morphology as exemplified by the A18c and the A18g5/Canon neurons (when multiple names for an interneuron were used in the literature, we mention both once). Canon neurons, from abdominal segments 1 to 7, have an axonal projection that terminates in a common region in thoracic segment 1 24 suggesting a shared developmental programme. The homologous A18g5 interneuron in a8 has a short axonal projection terminating in a5 (Fig. 1b). The abrupt changes in morphology are often observed in neurons that are present in a1 (e.g., A18c) or a8 (e.g., canon). Morphological analysis of homologous interneurons defines three diversification modes of neurons located in the VNC: i) neurons that are invariant regardless of the segment where they are located; ii) neurons with gradual changes of morphology along the antero-posterior neuraxis and iii) neurons that present abrupt changes in morphology. A common feature of variable neurons is the conservation of the main skeleton (Supplemental Fig. 2) suggesting that their identity has been maintained but they may have responded to segment- and cell- specific developmental programs that have changed their fine wiring. Hox expression deviates in a proportion of differentiating interneurons. The antero-posterior differences in neuronal morphology suggested a role for the Hox genes, that are conserved master regulators of cellular identity along the body axis, in shaping wiring diversity 18,19 . Their effect needed to be not solely regional, but also generate cell specific differences. Hox genes are re-expressed in postmitotic neurons as they begin to differentiate (embryonic stage 16), put out axons and dendrites and form synaptic connections 37,38 suggesting they could drive the necessary network diversification. We analysed the levels of Hox proteins (ANTP, UBX, ABD-A and ABD-B) expressed in the abdominal segments in the NB lineages previously mentioned (Fig. 2a, b). The expression values normalised as a percentage of the highest expressing neuron in the same embryo 39 . The quantification revealed a wide heterogeneity in Hox combinations and levels (Fig. 2a). To understand the main characteristics of the Hox protein distribution, we used a k ernel d ensity e stimation (KDE) analysis dividing the probability mass in five evenly spaced regions of 20% density (Fig. 2a and Supplementary Fig. 3). The KDE showed: first, that more than half (60%) of the neurons are expressing the generic Hox code expected for the segment. Second, that only a small proportion of neurons (20%) are expressing relatively high levels of one or more Hox protein (values above the last ring). Finally, it revealed peaks of maximum density (the most observed intensity, smallest ring) with very low levels of Hox expression (e.g. a2-a7 peak at 10% ABD-A and UBX). We then wondered if the heterogeneity was only due to differences in UBX and ABD-A expression between the anterior and posterior compartment of each segment, as previously described 40-42 or, if it was present within lineages indicating a more complex regulation and possible role. We found the expected compartmental differences, with UBX highly expressed in the NB lineages located in the anterior half of the parasegment and ABD-A higher in NB lineages in the posterior half (Fig. 2B, fig S3).More interesting is our discovery that cells belonging to the same lineage expressed different profiles of Hox proteins (Fig. 2b-d, and Supplementary Fig. 4 and 5).This is evident in NB 2-4, where ABD-A is absent or very low (<20%) in half of the interneurons from the lineage while the other half expresses an intermediate level of ABD-A (Fig. 1d) when located in a4, a canonical ABD-A region. Similarly, in a4, the neurons from NB 1-2 lineage express either high ABD-A or high UBX (Supplementary Fig. 4). Finally, we also found one neuron from NB5-6 that expressed UBX in a8, outside of the general pattern of expression between a2-a7 41 . These cell-specific differences in the combinations of Hox proteins within NB lineages were present in all segments and in all NB lineages that we analysed (Supplementary Fig. 4 and 5). Overall, our data show that for most neurons, the postmitotic Hox protein expression combination matches the generic parasegmental pattern described in the literature, but a proportion of individual neurons deviates considerably (Figure 1E). The fact that these deviations precisely occur during the process of differentiation, suggests that Hox proteins may play a role in the segment specific differentiation of equivalent cells along the antero-posterior axis. Cell-autonomous Postmitotic Control of Neuronal Diversity. To evaluate if the postmitotic expression of Hox genes drives the segment-specific differences in morphology, we focused on interneurons we could reliably identify in embryo and larva (Supplementary. Fig. 6). We worked with six interneurons from NB2-4 that could be recognised based on their morphology, birth order as determined with electron microscopy, and position within the lineage (Supplementary Fig. 6). Furthermore, two of these neurons, A18a and Canon, belong to the circuits generating crawling and they can be manipulated postmitotically 23,24 (Supplementary Fig. 7). Additionally, we included an interneuron involved in proprioception and required for left-right symmetry during crawling (A08e2) 43 . Invariant neurons have very low levels of Hox protein as in A18a (and A18b3, A18g3 A18c) and are insensitive to Hox over-expression as seen when over-expressing abd-A, Ubx and ABD-B in A18a (Fig. 3a and Supplementary Fig. 8). In contrast, differences in morphology correlate with segment specific higher levels of Hox protein suggesting the existence of a threshold which needs to be overcome for rewiring (Fig 3b,c, and Supplementary Fig. 8). A statistical comparison of the Hox levels in invariant versus variant neurons confirmed that significantly higher levels of Hox genes are present in variant neurons (Fig. 3d). To empirically test the role of postmitotic Hox levels, we carried out gain of function experiments. We predicted that increasing Hox levels during differentiation will induce the transformation of morphology. In Canon neurons, ABD-B is relatively high in a8 and this correlates with a shorter axon. Over-expressing Abd-B in a1 and a6 resulted in much shorter axons, resembling the morphology of the homologous neuron in a8 (Fig. 3b’,b”). This result indicates that ABD-B levels refine a specific axon developmental program in the terminal segment a8. In A08e2, UBX levels are only high in a1 coinciding with anteriorly projecting axons and dendrites (Fig. 3c). When Ubx was over-expressed, we observed that the axons, which normally bifurcate into an anterior and posterior branch in segments a2 until a6, were uniquely projecting anteriorly as observed in a1. Ubx over-expression did not produce a clear effect on the dendritic morphology of A08e2, only inducing an anterior lengthening of the lateral dendritic branch in a7 (Fig. 3b). Once more, we see that UBX levels modulate the development of the segmental A08e2 axonal direction of growth (Fig. 3b’,b”). These results indicate that cell-autonomous postmitotic increase in Hox activity in specific differentiating neurons modulates specific developmental programs that determine the diversification of neuronal architecture. Rewiring incorporates neurons into novel functional motor circuits. We then asked to what extent Hox-driven changes in wiring affect their recruitment into motor circuits controlling specialized behavioral patterns. Working with newly hatched first instar larvae and using calcium imaging, we evaluated the patterns of activity of A18a and Canon neurons. Both neurons remained spontaneously active in isolated nervous system and are part of the circuitry for locomotion. As a control for an invariant neuron, we analyzed the pattern of activity of A18a. Consistently with observations in a later larval stage (20) , A18a activity propagated in forward and backward waves, being part of the circuitry for forward and backward crawls (Fig. 4a). Over-expressing Ubx , abd-A or Abd-B in postmitotic A18a interneurons had no effect on their activity patterns (Fig. 4b,c and Supplementary Fig. 9). This observation is consistent with the lack of effect of this manipulation on the A18a morphology and suggests that Hox gene expression is not altering the physiological properties of the neuron. We then performed the equivalent experiment analyzing the spontaneous activity pattern of Canon neurons, which are only active during fictive backward locomotion (Fig. 4d) and regulate the timing of muscle relaxation via excitation onto inhibitory pre-motor interneurons 24 . Over-expression of Abd-B in Canon neurons (called Canon> Abd-B from now on) induced a shortening of the axon, which is likely to affect the intersegmental connectivity between homologous neurons. The driver line used labels a variable number of Canon neurons, but when over-expressing Abd-B the number of neurons expressing calcium signals was decreased, suggesting they were inactive or absent (Fig. 4e and Supplementary Fig. 9). The active neurons showed different levels of intersegmental coordination. Some waves of activity propagated backwards along the nerve cord reaching a8 while in others, the pattern of activity was irregular within and across segments. We also observed the appearance of a greater proportion of forward waves of activity; a phenotype absent in the control line (Fig. 4d,f and 24 ). This phenotype was specific to the over-expression of Abd-B and absent when over-expressing Ubx or abd-A (Fig. 4f and 24 ). To test if the forward waves on neuronal activity observed in Canon> Abd-B neurons are associated with fictive forward waves of locomotion, we recorded canon neurons simultaneously to aCC motorneurons. We found that a proportion of Canon neurons were active both during forward and backward fictive locomotion, while many neurons remained silent. This showed that the active neurons have become part of the forward and backward crawling circuit, while confirming that the development of a proportion of Canon > Abd-B neurons has been affected some to the extent of becoming inactive (Fig. 4g,h). Finally, we wondered how the connectivity of Canon> Abd-B neurons might have changed to enable the integration into forward crawl circuit. On possible mechanism would be that the weight of their pre-synaptic connections to neurons which are active during forward and backward waves has changed 24 . We evaluated the identity of presynaptic partners and the location of synapses onto the axon of a wildtype a6 canon neurons, using reconstructions from an electron microscopy volume of a first instar larval central nervous system 24 , 44 (supplementary Fig. 10). We found that the majority of the expected Canon-Canon synapses, necessary for the propagation of backward activity, occur along two to three segments away from the soma (Fig. 4i). The shortening of the axonal projection is likely to have affected these inputs, decreasing the drive during backward waves (Fig. 4j). Canon neurons also receive inputs from the brain descending backward locomotion command neurons MDNs 24,45 .  MDNs synapses are mostly located in the more rostral axonal region of canon neurons, indicating that a shorter axon would probably diminish the strength of these inputs, affecting the backward propagation of Canon neurons activity in response to sensory stimulation. To become activated in forward waves, Canon> Abd-B neurons need to receive inputs from forward active neurons. In a6, the canon neuron on the right (not the left) receives one synapse from A27h, a neuron necessary for the propagation of forward crawl. This synaptic input is not sufficient to drive wildtype canon’s activity during forward locomotion, but the close proximity between Canon and A27h neurons, that make contact several times at the dendrites close to the soma, suggests that Canon’s rewiring could promote more or stronger inputs from the circuit involved in forward crawling. This combined with the predicted decrease of backward related inputs is likely responsible of the incorporation of canon> Abd-B neurons into forward pattern of activity. Overall, these experiments show that rewiring specific neurons can have a dramatic impact on their connectivity and function allowing them to be incorporated into a novel specialized circuit. Postmitotic Hox driven rewiring drives the evolution of behaviour. It is difficult to imagine how the integration of rewired Canon > Abd-B interneurons into forward locomotion circuits, might drive circuit diversification. Wildtype Canon neurons induce segment relaxation after a backward peristaltic wave of contraction has passed. If the wave is travelling forward, the relaxation should occur one or two segment in front of the wave, in segments yet to be contracted (Fig. 3h). This might have no obvious behavioural effect or might affect the wave propagation. We found no difference in the speed of propagation of forward waves in canon> AbdB (Fig. 5a)suggesting there is no effect. We also quantified the time of backward wave propagation, that should be decreased if our prediction on Canon-Canon neurons connectivity weakening stands true (Fig. 5b and 24 ). We did not find a significant effect but could observe a clear bimodal distribution of the larval behaviour, where half of the animals had their behaviour affected. This dichotomy was probably due to differences in the number of neurons affected in each individual as consequence of stochastic expression in the driver line (supplementary Fig. 7). Finally, we over-activated the neurons using thermogenetics (with UAS- TRPA1 ). In contrast to to the Canon > TRPA1 that were almost completely paralyzed during the manipulation 24 , many Canon > TRPA1, AbdB larvae performed forward waves (Fig. 5c). These escapers showed aberrant phenotypes in that their peristaltic waves of muscle contraction progressed laboriously, indicating a behavioural impact of the rewiring induced by changes in the levels of Abd-B during differentiation. These results encouraged us to further test the role of rewiring in the VNC for the diversification of behaviour. Many neurons show changes in morphology along the abdominal segments (Fig. 1). The additive effect of rewiring in several interneurons, that has taken place during evolution, is likely to have driven the establishment of distinct motor capacities for specific segments, allowing the diversification of homologous behaviours along the larval body 46-48 . To test this hypothesis, we evaluated the larval locomotor behaviour while over-expressing the Ubx, abd-A and Abd-B genes respectively in the thoracic and abdominal segments using the driver line tsh -gal4. A Gal80 ts -mediated conditional expression allowed us to perform this manipulation after NB proliferation has ended, during neuronal differentiation (Fig. 5d). The larvae showed both losses and gains of phenotype as consequence of these manipulations. All larvae were able to execute coordinated crawls and turns (Fig. 5e-h) but they showed certain defects in locomotor performance ( < 2.5%) (Fig. 5i) such as: i) incomplete waves that stop before reaching the end of the body, ii) waves of contraction starting simultaneously in different segments (e.g a8 and a4) and iii) waves of contraction that changed direction before reaching the end of the body. Defective waves can be observed in all genotypes including control animals, but their frequency was slightly higher in genetically modified individuals. tsh > Abd-B presented a singular balance defect, their body shifting from left to right, as they steadily crawled (movie 1 and 2). The frequency and coordination of turns and head casts was unaltered (Fig. 5j). These phenotypes indicate that the establishment of segmental connections, that determine particular functions within the locomotor network, depends on Hox genes level of expression during differentiation. Strikingly, we found two gain of function phenotypes. Over-expression of Ubx induced a significant increase of head lifting both during crawls and head casts, including turns (Fig. 5k-m and movie 3). Body lifting and lowering are necessary for displacement during peristaltic forward crawling as they allow the contracted segment to move forward, without being dragged over the substrate. Ubx mutants have defective lifting and lowering of segments a1 to a3 49 . Therefore, we see that rewiring, occurring during differentiation, has been sufficient to drive the acquisition of a novel behaviour, fundamental for the proper execution of forward crawling, in thoracic segments, suggesting a transformation towards forward waves. The second phenotype was a significant increase in the number of backward waves upon the over-activation of abd-A (Fig. 5f and movie 4). The phenotype consists of the spontaneous initiation of bouts of backward waves, interrupted by head cast, as occurs during forward crawling. The anterior initiation cite suggests a reorganization of connections such that thoracic segments have become more similar to abdominal segments. The changes are likely to include motor and proprioception circuits that drive the initiation and propagation of peristaltic crawling waves. In summary, locomotion was not impaired by the changes in Hox genes levels of expression during differentiation, but the neuronal circuits were remodeled in ways that allowed the emergence of new behaviours in homologous segments. Discussion Together our results support the idea that the evolution of a new homologous behaviour is determined by the rewiring of a subset of interneurons. Differences in morphology change the synaptic connectivity between homologous neurons, allowing cells to meet new or more partners or modifying the synaptic weight between current synaptic partners. These differences have the potential of reconfiguring the patterns of activity between neurons supporting the evolution of new circuits, or circuit motifs, for specialized behaviours (Fig. 5n). In this scenario, region-specific changes of morphology of a subset of interneurons within a conserved network constitute a mechanism by which a new behavioral program can evolve while retaining the main properties of the network 6 . From a genetic point of view, the rewiring mechanism underlying network evolution results in low impact for the general structure of the central nervous system or low pleiotropy 6 . Our experiments show that specialization of morphology does not completely reconfigure the topology of the neurons. Although these neurons show diversity in presence, length, or orientation of their processes, they retained their main skeleton, which is an identity feature. Thus, only the late program of differentiation has been modified in a cell-autonomous manner in a subset of neurons, while retaining the main developmental program intact. We propose that modifications occur only in the final stages of development, which has the advantage of sparing the nervous system from wider complexity circuit reorganization that may generate deleterious consequences. In the motor system, we found that cell-specific deviations in postmitotic Hox protein levels determine the diversification of morphology during differentiation. It is likely that they modify the expression of effector proteins (or also called realizators) 50-52 that control axonal and dendritic trajectories 53 , synaptic organisation, and connectivity 15,38,54,55 . There is no apparent rule for the way Hox genes control the neuronal morphology, but rather a cell-specific effect where each cell slightly tunes their endogenous developmental program differentially in response to distinct levels of Hox genes 51 . Notably in the cortex, several transcription factors expressed postmitotically contribute to the establishment of subtype-specific morphology features 56 . Therefore, we see a conserved mechanism, where specification genes refine the neuronal architecture and connectivity allowing for further functional diversity 56-58 , while retaining the main developmental program. Our work thus uncovers a mechanism by which evolutionary selected fortuitous deviations in specification genes levels rewire a subset of neurons in ways that drive their incorporation into a novel specialized circuit. Such a mechanism ensures the system stability and simplifies circuit diversification—within organisms and potentially also across homologous behaviour in all animals (Fig. 5n). Declarations Acknowledgments: We thank Michael Bate for useful discussions and making constructive comments on earlier versions of the manuscript and Sarah Newbury and Natalie Page for editorial comments. We are grateful to Ernesto Sánchez-Herrero, Stefan Thor, Claudio Alonso and Matthias Landgraf for sharing fly stocks. Aki Nose for sharing EM data and Deeptha Vasudevan and Ellie Heckscher who helped retrieving EM data for birth order analysis. We thank Megan McIntosh, Dan Young and Nadia Amin for technical support, Flybase for providing gene-related information, the Bloomington Stock Center for fly stocks, the Developmental Studies Hybridoma Bank for antibodies, The Wolfson Imaging Center and its technician Yan Gu. We thank the funding sources: Royal Society and Wellcome Trust Henry Dale Fellowship Grant 105568/Z/14/Z (JB) UK Biotechnology and Biological Sciences Research Council grant BB/X009289/1 (JB) National Institutes of Health NS122903 (HL) Leukaemia UK John Goldman Fellowship 2020/JGF/003 (SM) UKRI Future Leaders Fellowship MR/T041889/1 (SM) National Science Foundation under Grant No. NSF PHY-1748958 partial support (JB). Author contributions: Conceptualization: JB Methodology: RN, EM, HL, NV, SAM, JWT, JB Investigation: JRC, MS, SD, KN, RN, EM, HL, NV, SAM, JWT, JB Visualization: JRC, MS, KN, RN, JWT, JB Funding acquisition: HL, SAM, JB Project administration: JB Supervision: SAM, JB Writing – original draft: JB Writing – review & editing: KN, EM, HL, NV, SAM, JWT, JB Competing interests: Authors declare that they have no competing interests. Data and materials availability: Mini-atlas will be uploaded on webpage, it is available on request. KDE was performed in Julia programming language using the MarginalHist function from StatsPlots.jl (https://github.com/JuliaPlots/StatsPlots.jl), which utilises KernelDensity.jl (https://github.com/JuliaStats/KernelDensity.jl). References Seeholzer, L. F., Seppo, M., Stern, D. L. & Ruta, V. Evolution of a central neural circuit underlies Drosophila mate preferences. Nature 559 , 564-569, doi:10.1038/s41586-018-0322-9 (2018). Ding, Y. et al. Neural Evolution of Context-Dependent Fly Song. Curr Biol 29 , 1089-1099 e1087, doi:10.1016/j.cub.2019.02.019 (2019). Auer, T. O. et al. Olfactory receptor and circuit evolution promote host specialization. Nature 579 , 402-408, doi:10.1038/s41586-020-2073-7 (2020). Bumbarger, D. J., Riebesell, M., Rodelsperger, C. & Sommer, R. J. System-wide rewiring underlies behavioral differences in predatory and bacterial-feeding nematodes. Cell 152 , 109-119, doi:10.1016/j.cell.2012.12.013 (2013). Hong, R. L. et al. Evolution of neuronal anatomy and circuitry in two highly divergent nematode species. Elife 8 , doi:10.7554/eLife.47155 (2019). Roberts, R. J. V., Pop, S. & Prieto-Godino, L. L. Evolution of central neural circuits: state of the art and perspectives. Nat Rev Neurosci 23 , 725-743, doi:10.1038/s41583-022-00644-y (2022). Lewis, E. B. A gene complex controlling segmentation in Drosophila. Nature 276 , 565-570, doi:10.1038/276565a0 (1978). Merabet, S. & Carnesecchi, J. Hox dosage and morphological diversification during development and evolution. Semin Cell Dev Biol 152-153 , 70-75, doi:10.1016/j.semcdb.2022.11.009 (2024). Lumsden, A. & Krumlauf, R. Patterning the vertebrate neuraxis. Science 274 , 1109-1115, doi:10.1126/science.274.5290.1109 (1996). del Toro, E. D. et al. Generation of a novel functional neuronal circuit in Hoxa1 mutant mice. J Neurosci 21 , 5637-5642 (2001). Briscoe, J., Pierani, A., Jessell, T. M. & Ericson, J. A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell 101 , 435-445, doi:10.1016/s0092-8674(00)80853-3 (2000). Dasen, J. S., Tice, B. C., Brenner-Morton, S. & Jessell, T. M. A Hox regulatory network establishes motor neuron pool identity and target-muscle connectivity. Cell 123 , 477-491, doi:10.1016/j.cell.2005.09.009 (2005). Dasen, J. S. & Jessell, T. M. Hox networks and the origins of motor neuron diversity. Curr Top Dev Biol 88 , 169-200, doi:10.1016/S0070-2153(09)88006-X (2009). Dasen, J. S., Liu, J. P. & Jessell, T. M. Motor neuron columnar fate imposed by sequential phases of Hox-c activity. Nature 425 , 926-933, doi:10.1038/nature02051 (2003). Catela, C., Shin, M. M., Lee, D. H., Liu, J. P. & Dasen, J. S. Hox Proteins Coordinate Motor Neuron Differentiation and Connectivity Programs through Ret/Gfralpha Genes. Cell Rep 14 , 1901-1915, doi:10.1016/j.celrep.2016.01.067 (2016). Shin, M. M., Catela, C. & Dasen, J. Intrinsic control of neuronal diversity and synaptic specificity in a proprioceptive circuit. Elife 9 , doi:10.7554/eLife.56374 (2020). Philippidou, P. & Dasen, J. S. Hox genes: choreographers in neural development, architects of circuit organization. Neuron 80 , 12-34, doi:10.1016/j.neuron.2013.09.020 (2013). Jung, H. et al. The Ancient Origins of Neural Substrates for Land Walking. Cell 172 , 667-682 e615, doi:10.1016/j.cell.2018.01.013 (2018). Jung, H. et al. Evolving Hox activity profiles govern diversity in locomotor systems. Dev Cell 29 , 171-187, doi:10.1016/j.devcel.2014.03.008 (2014). Berni, J., Pulver, S. R., Griffith, L. C. & Bate, M. Autonomous circuitry for substrate exploration in freely moving Drosophila larvae. Curr Biol 22 , 1861-1870, doi:10.1016/j.cub.2012.07.048 (2012). Berni, J. Genetic dissection of a regionally differentiated network for exploratory behavior in Drosophila larvae. Curr Biol 25 , 1319-1326, doi:10.1016/j.cub.2015.03.023 (2015). Sims, D. W., Humphries, N. E., Hu, N., Medan, V. & Berni, J. Optimal searching behaviour generated intrinsically by the central pattern generator for locomotion. Elife 8 , doi:10.7554/eLife.50316 (2019). Hasegawa, E., Truman, J. W. & Nose, A. Identification of excitatory premotor interneurons which regulate local muscle contraction during Drosophila larval locomotion. Sci Rep 6 , 30806, doi:10.1038/srep30806 (2016). Hiramoto, A. et al. Regulation of coordinated muscular relaxation in Drosophila larvae by a pattern-regulating intersegmental circuit. Nat Commun 12 , 2943, doi:10.1038/s41467-021-23273-y (2021). Kohsaka, H. et al. Regulation of forward and backward locomotion through intersegmental feedback circuits in Drosophila larvae. Nat Commun 10 , 2654, doi:10.1038/s41467-019-10695-y (2019). Zarin, A. A., Mark, B., Cardona, A., Litwin-Kumar, A. & Doe, C. Q. A multilayer circuit architecture for the generation of distinct locomotor behaviors in Drosophila. Elife 8 , doi:10.7554/eLife.51781 (2019). Heckscher, E. S., Lockery, S. R. & Doe, C. Q. Characterization of Drosophila larval crawling at the level of organism, segment, and somatic body wall musculature. J Neurosci 32 , 12460-12471, doi:10.1523/JNEUROSCI.0222-12.2012 (2012). Lahiri, S. et al. Two alternating motor programs drive navigation in Drosophila larva. PLoS One 6 , e23180, doi:10.1371/journal.pone.0023180 (2011). Pulver, S. R. et al. Imaging fictive locomotor patterns in larval Drosophila. J Neurophysiol 114 , 2564-2577, doi:10.1152/jn.00731.2015 (2015). Thomas, J. B., Bastiani, M. J., Bate, M. & Goodman, C. S. From grasshopper to Drosophila: a common plan for neuronal development. Nature 310 , 203-207, doi:10.1038/310203a0 (1984). Doe, C. Q. Molecular markers for identified neuroblasts and ganglion mother cells in the Drosophila central nervous system. Development 116 , 855-863 (1992). Stollewerk, A. & Simpson, P. Evolution of early development of the nervous system: a comparison between arthropods. Bioessays 27 , 874-883, doi:10.1002/bies.20276 (2005). Li, H. H. et al. A GAL4 driver resource for developmental and behavioral studies on the larval CNS of Drosophila. Cell Rep 8 , 897-908, doi:10.1016/j.celrep.2014.06.065 (2014). Nern, A., Pfeiffer, B. D. & Rubin, G. M. Optimized tools for multicolor stochastic labeling reveal diverse stereotyped cell arrangements in the fly visual system. Proc Natl Acad Sci U S A 112 , E2967-2976, doi:10.1073/pnas.1506763112 (2015). Rickert, C., Kunz, T., Harris, K. L., Whitington, P. M. & Technau, G. M. Morphological characterization of the entire interneuron population reveals principles of neuromere organization in the ventral nerve cord of Drosophila. J Neurosci 31 , 15870-15883, doi:10.1523/JNEUROSCI.4009-11.2011 (2011). Monedero Cobeta, I., Salmani, B. Y. & Thor, S. Anterior-Posterior Gradient in Neural Stem and Daughter Cell Proliferation Governed by Spatial and Temporal Hox Control. Curr Biol 27 , 1161-1172, doi:10.1016/j.cub.2017.03.023 (2017). Miguel-Aliaga, I. & Thor, S. Segment-specific prevention of pioneer neuron apoptosis by cell-autonomous, postmitotic Hox gene activity. Development 131 , 6093-6105, doi:10.1242/dev.01521 (2004). Velten, J. et al. Single-cell RNA sequencing of motoneurons identifies regulators of synaptic wiring in Drosophila embryos. Mol Syst Biol 18 , e10255, doi:10.15252/msb.202110255 (2022). Hirth, F., Hartmann, B. & Reichert, H. Homeotic gene action in embryonic brain development of Drosophila. Development 125 , 1579-1589 (1998). White, R. A. & Wilcox, M. Distribution of Ultrabithorax proteins in Drosophila. EMBO J 4 , 2035-2043, doi:10.1002/j.1460-2075.1985.tb03889.x (1985). White, R. A. & Wilcox, M. Protein products of the bithorax complex in Drosophila. Cell 39 , 163-171, doi:10.1016/0092-8674(84)90202-2 (1984). Gebelein, B. & Mann, R. S. Compartmental modulation of abdominal Hox expression by engrailed and sloppy-paired patterns the fly ectoderm. Dev Biol 308 , 593-605, doi:10.1016/j.ydbio.2007.05.017 (2007). Heckscher, E. S. et al. Even-Skipped(+) Interneurons Are Core Components of a Sensorimotor Circuit that Maintains Left-Right Symmetric Muscle Contraction Amplitude. Neuron 88 , 314-329, doi:10.1016/j.neuron.2015.09.009 (2015). Schneider-Mizell, C. M. et al. Quantitative neuroanatomy for connectomics in Drosophila. Elife 5 , doi:10.7554/eLife.12059 (2016). Carreira-Rosario, A. et al. MDN brain descending neurons coordinately activate backward and inhibit forward locomotion. Elife 7 , doi:10.7554/eLife.38554 (2018). Averof, M. & Patel, N. H. Crustacean appendage evolution associated with changes in Hox gene expression. Nature 388 , 682-686, doi:10.1038/41786 (1997). Jarvis, E., Bruce, H. S. & Patel, N. H. Evolving specialization of the arthropod nervous system. Proc Natl Acad Sci U S A 109 Suppl 1 , 10634-10639, doi:10.1073/pnas.1201876109 (2012). Martin, A. et al. CRISPR/Cas9 Mutagenesis Reveals Versatile Roles of Hox Genes in Crustacean Limb Specification and Evolution. Curr Biol 26 , 14-26, doi:10.1016/j.cub.2015.11.021 (2016). Dixit, R., Vijayraghavan, K. & Bate, M. Hox genes and the regulation of movement in Drosophila. Dev Neurobiol 68 , 309-316, doi:10.1002/dneu.20589 (2008). Garcia-Bellido, A. Genetic control of wing disc development in Drosophila. Ciba Found Symp 0 , 161-182, doi:10.1002/9780470720110.ch8 (1975). Akam, M. Hox genes: from master genes to micromanagers. Curr Biol 8 , R676-678, doi:10.1016/s0960-9822(98)70433-6 (1998). Hobert, O. Homeobox genes and the specification of neuronal identity. Nat Rev Neurosci 22 , 627-636, doi:10.1038/s41583-021-00497-x (2021). Baek, M. & Mann, R. S. Lineage and birth date specify motor neuron targeting and dendritic architecture in adult Drosophila. J Neurosci 29 , 6904-6916, doi:10.1523/JNEUROSCI.1585-09.2009 (2009). Dasen, J. S., De Camilli, A., Wang, B., Tucker, P. W. & Jessell, T. M. Hox repertoires for motor neuron diversity and connectivity gated by a single accessory factor, FoxP1. Cell 134 , 304-316, doi:10.1016/j.cell.2008.06.019 (2008). Catela, C., Chen, Y., Weng, Y., Wen, K. & Kratsios, P. Control of spinal motor neuron terminal differentiation through sustained Hoxc8 gene activity. Elife 11 , doi:10.7554/eLife.70766 (2022). Lodato, S. & Arlotta, P. Generating neuronal diversity in the mammalian cerebral cortex. Annu Rev Cell Dev Biol 31 , 699-720, doi:10.1146/annurev-cellbio-100814-125353 (2015). Lai, T. et al. SOX5 controls the sequential generation of distinct corticofugal neuron subtypes. Neuron 57 , 232-247, doi:10.1016/j.neuron.2007.12.023 (2008). Liu, J. et al. FOXG1 sequentially orchestrates subtype specification of postmitotic cortical projection neurons. Sci Adv 8 , eabh3568, doi:10.1126/sciadv.abh3568 (2022). Additional Declarations There is NO Competing Interest. Supplementary Files RodriguezCurtSupplMaterial.pdf Supplementary materials and Figures NNA95163movie1controltsh.mov Supplementary Movie 1 NNA95163movie2DiffAbdBtshabdB.mov Supplementary Movie 2 NNA95163movie3DiffUbxtshUbx.mov Supplementary Movie 3 NNA95163movie4DiffabdAtshabdA.mov Supplementary Movie 4 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information

Related documents

• Dynamic Polyethism and Competition for Tasks in Threshold Reinforcement Models of Social Insects#622717
• PAVXploreRL: Physical-Action-Visual World Model Reinforcement Learning with Action Exploration#999828
• The elusivity of nature and the mind-matter problem#743385
• Integrating morphological and molecular perspectives on foraging tactic constancy in a nectar-robbing bumble bee#796442
• Long-term camera trap monitoring of mammalian diversity and activity rhythms of threatened species in Xianxialing Nature Reserve, Zhejiang Province, China#964951