A projectome of the bumblebee central complex

  1. Marcel Ethan Sayre
  2. Rachel Templin
  3. Johanna Chavez
  4. Julian Kempenaers
  5. Stanley Heinze

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This paper addresses researchers interested in architecture and function of the insect central complex as it represents the first comprehensive projectome dataset of any central complex outside Drosophila. The authors use the bumblebee as representative for hymenopterans with their navigation skills. Further, they mine their data for conserved and diverged aspects compared to fly (Drosophila) knowledge and hypothesize how the differences may relate to diverged neural circuit function. Hence, they provide an excellent and comprehensive descriptive resource providing a point of reference for others and a starting point for comparative studies of neural circuits. In particular, this study is the first comprehensive description of columnar neurons in the bumblebee central complex, described through the lens of the recently published fruit fly connectome of the same, homologous neuropil. The comparative approach used here holds promise for describing neural circuits in bees and flies in shared frame of reference. The authors use an approach that reflects a compromise between quick collection of electron microscopy (EM) data and being able to fully reconstruct all neurons in the bumble bee's central complex. The authors are transparent about the method's limitations and draw appropriate conclusions.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Insects have evolved diverse and remarkable strategies for navigating in various ecologies all over the world. Regardless of species, insects share the presence of a group of morphologically conserved neuropils known collectively as the central complex (CX). The CX is a navigational center, involved in sensory integration and coordinated motor activity. Despite the fact that our understanding of navigational behavior comes predominantly from ants and bees, most of what we know about the underlying neural circuitry of such behavior comes from work in fruit flies. Here, we aim to close this gap, by providing the first comprehensive map of all major columnar neurons and their projection patterns in the CX of a bee. We find numerous components of the circuit that appear to be highly conserved between the fly and the bee, but also highlight several key differences which are likely to have important functional ramifications.

Article activity feed

  1. Version published to 10.7554/elife.68911 on eLife
  2. eLife

    Evaluation Summary:

    This paper addresses researchers interested in architecture and function of the insect central complex as it represents the first comprehensive projectome dataset of any central complex outside Drosophila. The authors use the bumblebee as representative for hymenopterans with their navigation skills. Further, they mine their data for conserved and diverged aspects compared to fly (Drosophila) knowledge and hypothesize how the differences may relate to diverged neural circuit function. Hence, they provide an excellent and comprehensive descriptive resource providing a point of reference for others and a starting point for comparative studies of neural circuits. In particular, this study is the first comprehensive description of columnar neurons in the bumblebee central complex, described through the lens of the recently published fruit fly connectome of the same, homologous neuropil. The comparative approach used here holds promise for describing neural circuits in bees and flies in shared frame of reference. The authors use an approach that reflects a compromise between quick collection of electron microscopy (EM) data and being able to fully reconstruct all neurons in the bumble bee's central complex. The authors are transparent about the method's limitations and draw appropriate conclusions.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    Sayre et al. report on an impressive data set of over 1,300 neuronal arbours reconstructed, without synapses, from mesoresolution volume electron microscopy, therefore providing a first data set from a densely labelled image volume of a complete central complex of an insect other than a Drosophila. The authors readily acknowledge the limitations of their study, namely the choice of a coarse resolution of 126 nm/voxel which precluded the reconstruction of neuronal arbours with axons thinner than 600 nm, affecting primarily one class of neurons (the tangential cells) which interlink different compartments within the same neuropil region (be it the ellipsoid body (EB), fan-shaped body (FB) or the protocerebral bridge (PB)) and therefore do not preclude the analysis of inter-neuropil regions, which is dominated by columnar neurons of thicker caliber and therefore reconstructed sufficiently here. The low-resolution volume EM was then complemented with both less coarse volume EM (at 100 nm/voxel) and light-microscopy labellings.

    The courage it takes to work with an animal as large as a Bombus terrestris worker using volume electron microscopy is commendable. Obtaining even a projectome, namely, the low-order branches of neurons without the synapses, is already an extraordinary achievement.

    The authors draw direct comparisons with the locust (and briefly the butterfly) and particularly with the fly Drosophila, where the synaptic connectivity circuits and functional data is most abundantly known and published. The comparisons are apt, and the focus on the differences interesting: in resolving the lack of thoroidal shape of the ellipsoid body-equivalent in the bumblebee, and in relating the differences in the relative number of neurons to additional features known to be implemented by the central complex such as path integration. All the above makes for a lengthy discussion, but the specialists will appreciate the detailed one-to-one comparisons between fly, bee, butterfly and locust.

    To remark as well the willingness of the authors to use the Drosophila-centered nomenclature in the naming of compartments and neurons of the bumblebee, which helped perform these revisions and will most certaintly assist a number of fly-trained readers to distill and take home the key findings of this work.

    As dry as anatomy can be, the authors manage to bring in putative functional roles, based on computational models and functional data acquired on the central complex of other species, which make the findings lighter to read and immediately relatable. The authors have carefully acknowledged the limitations of their study data and comparisons appropriately and transparently.

  4. eLife

    Reviewer #2 (Public Review):

    This study by Sayre et al aims to better characterize neural circuitry of the central complex (CX) in the bee, a neuropil that is important for the control of many navigational behaviors. Much of the current knowledge of the CX circuits and function currently comes from flies, while many complex navigational behaviors have been described in other insects. Thus, this study is significant because (a) it begins to close the gap between the investigations of neural circuits in the fly and behavioral work in other insects, and (b) because it allows for cross-species comparisons of CX circuits and identification of evolutionary adaptations to perform certain behaviors.

    Besides the biological insights, which I will address below, this work demonstrates an approach to electron microscopy that balances imaging throughput with dataset completeness. Rather than image the full bumble bee CX at resolution sufficient for a complete reconstruction of all neurons in the CX, the authors collect lower resolution data and focus on reconstructing main neurites of a selected number of cell types. The authors present a "projectome", a map of neural projections between neuropils of the central complex (and in several cases to sub-regions of said neuropils). However, the method has limitations that prevent the authors from (a) fully tracing neuron processes below a certain size and (b) measuring connections between neurons. Furthermore, the morphology of some neurons cannot be fully reconstructed because the respective neurons leave the imaged volume. This limits the extent and level of detail at which different neurons can be grouped into types. The authors are very transparent about these limitations. They compare three data sets with different resolution to identify which information is lost at lower resolution and focus their analysis on cell types that can be characterized to a high degree with the data at hand: columnar neurons.

    Columnar neurons have characteristic projection patterns that are tightly linked to the function of the CX circuitry. The authors provide an atlas of the different types of columnar neurons and their numbers and compare their findings to the fly and where possible the locust. This reveals a beautiful functional homology of the head direction circuitry across the three species despite different anatomical implementations. The authors also identify several differences that could point at circuit adaptations that allow bees to excel at path integration-based navigation.

    For one region, the noduli (NO), the authors collected EM data with higher resolution, which enabled characterization intra-neuropil organization. Many studies suggest that in the central complex structure is often tightly linked with function, which makes this NO dataset a valuable contribution to further understand the organization of one of the main input structures to the central complex. In the larger neuropils, the ellipsoid body (EB) and fan-shaped body (FB), the resolution of the projectome unfortunately does not allow for detailed characterization of the intra-neuropil structure. Here, instead, the authors provide a rough characterization based on layers defined by immunolabelling of TH and 5TH as well as entry sites of primary neurites.

    This work presents an important first step to better relate findings about the central complex circuitry from fruit flies to the function of the homologous circuits in (bumble)bees, which might help us understand the adaptations that give rise to the astonishing behavioral repertoire of bees such as their ability to navigate accurate on a relatively large spatial scale.

  5. eLife

    Reviewer #3 (Public Review):

    The insect central complex (CX) is a brain part, which processes multimodal sensory input to guide orientation and directed locomotion. It is built by a large number of different neuron types with intriguing and unique connectivity patterns, which together form several interconnected and midline spanning neuropils. So far, the many neuron types have mainly been identified by sparse marking of cells in a number of insect species, which allowed for precise determination of projection patterns of selected neurons. This approach has revealed an overall conserved CX architecture but it suffers from the possibility that neuron types may be missed and that the relative position of the projections among neurons cannot be determined exactly. Recently, a comprehensive connectomics map was generated for the fruit fly Drosophila melanogaster providing projection and connectivity information in unprecedented detail. While the most thorough understanding of CX function is based on the elaborate toolkit available in the fly, the most complex navigation behaviors are known from ants and bees, calling for more comprehensive work in those species.

    Sayre et al. now provide a comprehensive projectome of the bumblebee CX by using serial block face EM and subsequent 3D reconstruction of 1,300 neurons. The projectome of the entire CX is complemented by analyses of sub-parts (with focus on the noduli) with higher resolution and with some immunohistochemical data. With this work, the authors provide a very extensive and valuable resource allowing for comparisons of neuron types and projections between species. This approach has the power to reveal on one hand the conserved core of CX projections and the specific differences, which might underlie the different navigational abilities of these insects.

    On the technical level, they find that significant additional information can be gained by adding high resolution SBFEM to the less resolved overall reconstruction. Further, the authors confirm that single cell reconstructions based on confocal microscopy can be well mapped onto the projectome, allowing for adding more detail to the reconstruction in the future.

    On the scientific side, they first establish a comprehensive resource for the bumblebee CX - the first such dataset outside Drosophila. They describe the setup of the entire bumblebee CX, defining the neuron types and their projection patterns with a focus on columnar neurons. Second, they mine this resource for conserved and diverged aspects compared to Drosophila with respect to numbers of certain neural cell types, conservation of key projection circuits and - importantly - identifying specific differences between these insects. Based on this careful analysis, several hypotheses are formulated as to what the differences might mean for the functions of the bumblebee CX. For instance, the authors confirm an overall conserved CX architecture and suggest a very conserved head direction circuit. But they also find differences in fan shaped body layering, noduli organization and other aspects like the lack of obvious deltaV and FX cells. Interestingly, they describe changes in projection patterns that could underlie the different morphologies of the ellipsoid body (bar like versus donut shaped) and the potential functional meaning of a number of bee-specific aspects of noduli projections.

    In summary, the strength of the paper lies in the establishment of a quite comprehensive cell atlas of the bumblebee CX displayed in extensive and clear figures. Further, the analysis for conservation and divergence with respect to the fly CX has been done in a very careful and comprehensive way. Intriguing divergences lead to hypotheses on the functional implications. All claims are well founded in the data.

    Almost unavoidably for data gained outside the Drosophila cosmos, the resource does not have single cell resolution when it comes to the fine terminal projections and, hence, does not provide information on specific cell-cell connectivity. Further, cells with small diameter neurites are likely missing outside of the noduli dataset. The authors are aware of these issues calling their resource projectome rather than connectome.

    The paper describes very interesting differences and respective hypotheses are presented but these are not tested.
    The reconstructions will be available as interactive datasets, which is essential for future use.

  6. Version published to 10.1101/2021.03.26.437137v1 on bioRxiv