1. Computational and Systems Biology
  2. Neuroscience
Download icon

Connectomes: Mapping the mind of a fly

  1. Jason Pipkin  Is a corresponding author
  1. Department of Biology, Brandeis University, United States
Insight
  • Cited 0
  • Views 1,892
  • Annotations
Cite this article as: eLife 2020;9:e62451 doi: 10.7554/eLife.62451

Abstract

Scientists have created the most detailed map of the fruit fly brain to date, identifying over 25,000 neurons and 20 million synapses.

Main text

Every thought, feeling and action emerges from the electrical interplay of billions of neurons in the brain – wired together by an intricate network of cables that connect through hundreds of billions of synapses. Therefore, to fully understand how the brain works we need to consider all parts of the brain and the connections between them.

A connectome is a comprehensive map of the structural and functional neural connections in the brain that enables scientists to explore and compare different pathways, circuits and regions. Creating such a map is a difficult endeavor: neurons are minuscule, and their extensive branches are even smaller (for example, neuronal branches in the brain of a fruit fly are often thinner than 50 nm – about a thousandth of the width of a human hair).

To achieve such a high resolution, ultrathin layers of brain tissue are imaged with an electron microscope, and the neurons and their connections are reconstructed in 3D. This is no small undertaking, and in the case of the roundworm Caenorhabditis elegans (whose brain only consists of 302 neurons), it took the better part of a decade to generate a comprehensive connectome (White et al., 1986). This dissuaded attempts to create connectomes for larger brains, until the technological advances in microscopy and computer vision finally caught up with demand (Denk and Horstmann, 2004; Heymann et al., 2006; Januszewski et al., 2018).

Nowadays, multiple efforts have been underway to conquer the next model organism, the fruit fly Drosophila melanogaster. In a space smaller than a pinhead, the brain of a fruit fly contains over 100,000 neurons with around 100 million synapses (Simpson, 2009). So far, 3D reconstructions of its brain have tended to either be sparse or limited to small sub-regions (Eichler et al., 2017; Takemura et al., 2017; Zheng et al., 2018). Now, in eLife, Louis Scheffer, Stephen Plaza (both Janelia Research Campus, HHMI) and colleagues – including Scheffer, Shan Xu, Michal Januszewski, Zhiyuan Lu, Shin-ya Takemura, Kenneth Hayworth, Gary B. Huang and Kazunori Shinomiya as joint first authors – report that they have created the largest high-resolution connectome in any animal to date (Scheffer et al., 2020).

The researchers (who are based in the US, Canada, Switzerland, Japan, Germany and UK) generated a connectome for one half of the largely symmetrical fly brain, comprising over 25,000 neurons and 20 million synapses (Figure 1A). To achieve this, Scheffer et al. analyzed 3D reconstructions of the neurons and their synapses using machine learning algorithms and more than 50 person-years of work to proofread the computer-generated map. This involved transforming more than 20 terabytes of raw image data into a 26-megabyte network diagram.

The connectome of a fruit fly.

(A) Scheffer et al. have used machine learning to reconstruct half of the fruit fly brain (area shown in teal), providing the most comprehensive connectome of any animal to date. (B) The different regions in the brain were labelled according to known brain maps. The fly brain can be subdivided based on tracts of neuronal branches and large tangles called neuropils. Any given neuron might innervate several neuropils, and neurons are often identified by the neuropils they make synapses with. (C) Neuropils of the odor processing center (the antennal lobe AL, the calyx CA and the lateral horn LH) and the neurons (yellow) that connect them. (D) The connectome can differentiate neurons into separate types that would have been difficult to distinguish using light microscopy due to their similar anatomy. (E) The connectome contains a complete list of all neurons and their synapses, making it possible to search for specific neurons and their connections. For example, the database shows that a neuron from a region called the mushroom body (shown in tan) connects to two other neurons in this neuropil (shown in magenta and teal).

Image credit: Figure 1A–D adapted from Scheffer et al., 2020 (CC BY 4.0), Figure 1E obtained from the online portal Neuprint (CC0).

A good map should be informative, accurate and usable. Towards that end, Scheffer et al. have labelled regions according to existing atlases of the fly brain (Figure 1B). In addition, they subdivided the brain based on the tracts of neuronal branches and large tangles called neuropils, which are thought to be sites of local computation (Figure 1C). With regions delineated in this way, individual neurons could be categorized based on the neuropils they connect to. This is particularly useful for distinguishing neurons that have similar anatomies, which would be difficult to tell apart using light microscopy alone (Figure 1D). Moreover, the connectome is freely available at the online portal ‘Neuprint’, and anyone can search the database for a neuron of interest, observe the images it was traced from, plot it in 3D and see which brain regions it interacts with (Figure 1E).

Scheffer et al. were also able to quantify how highly interconnected the fly brain is. For example, the path distance between neurons (i.e., how many neurons there are between a connected pair) is low – three quarters of neurons were linked by three or fewer interneurons. Moreover, they demonstrated that the different neuropils are indeed segregated electrically, suggesting that the same neuron can perform separate computations in different regions at the same time.

Of course, no map is perfect. Most branches traced belong to yet unidentified neurons (presumably the ones that reside outside the borders of the sequenced brain). Many of the tiniest twigs that are difficult for both human and machine eyes to trace may have been missed or lie ‘disconnected’ from their true parent neuron, which may reside in the other unmapped half of the brain. This could potentially reduce the number of synapses between two cells.

Scheffer et al. have pledged to improve and update their connectome over time – after all, another half of the fly brain remains to be analyzed. Nevertheless, the current map marks a significant increase in scale over the one of C. elegans and will undoubtedly help unravel the neurological basis underlying a fly’s behavior. And it may bring us one step closer to creating connectomes of larger animals, including vertebrates.

References

    1. White JG
    2. Southgate E
    3. Thomson JN
    4. Brenner S
    (1986) The structure of the nervous system of the nematode Caenorhabditis elegans
    Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 314:1–340.
    https://doi.org/10.1098/rstb.1986.0056

Article and author information

Author details

  1. Jason Pipkin

    Jason Pipkin is in the Department of Biology, Brandeis University, Waltham, United States

    For correspondence
    jpipkin@brandeis.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-5525-3951

Publication history

  1. Version of Record published: October 8, 2020 (version 1)

Copyright

© 2020, Pipkin

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,892
    Page views
  • 99
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Computational and Systems Biology
    2. Immunology and Inflammation
    Johannes Pettmann et al.
    Research Article Updated

    T cells use their T cell receptors (TCRs) to discriminate between lower-affinity self and higher-affinity non-self peptides presented on major histocompatibility complex (pMHC) antigens. Although the discriminatory power of the TCR is widely believed to be near-perfect, technical difficulties have hampered efforts to precisely quantify it. Here, we describe a method for measuring very low TCR/pMHC affinities and use it to measure the discriminatory power of the TCR and the factors affecting it. We find that TCR discrimination, although enhanced compared with conventional cell-surface receptors, is imperfect: primary human T cells can respond to pMHC with affinities as low as KD ∼ 1 mM. The kinetic proofreading mechanism fit our data, providing the first estimates of both the time delay (2.8 s) and number of biochemical steps (2.67) that are consistent with the extraordinary sensitivity of antigen recognition. Our findings explain why self pMHC frequently induce autoimmune diseases and anti-tumour responses, and suggest ways to modify TCR discrimination.

    1. Computational and Systems Biology
    2. Neuroscience
    Jack Goffinet et al.
    Research Article Updated

    Increases in the scale and complexity of behavioral data pose an increasing challenge for data analysis. A common strategy involves replacing entire behaviors with small numbers of handpicked, domain-specific features, but this approach suffers from several crucial limitations. For example, handpicked features may miss important dimensions of variability, and correlations among them complicate statistical testing. Here, by contrast, we apply the variational autoencoder (VAE), an unsupervised learning method, to learn features directly from data and quantify the vocal behavior of two model species: the laboratory mouse and the zebra finch. The VAE converges on a parsimonious representation that outperforms handpicked features on a variety of common analysis tasks, enables the measurement of moment-by-moment vocal variability on the timescale of tens of milliseconds in the zebra finch, provides strong evidence that mouse ultrasonic vocalizations do not cluster as is commonly believed, and captures the similarity of tutor and pupil birdsong with qualitatively higher fidelity than previous approaches. In all, we demonstrate the utility of modern unsupervised learning approaches to the quantification of complex and high-dimensional vocal behavior.