Encyclopedia of Computational Neuroscience

Living Edition
| Editors: Dieter Jaeger, Ranu Jung

Drosophila Connectome

  • Arjun BhariokeEmail author
  • Louis K. Scheffer
  • Dmitri B. Chklovskii
  • Ian A. Meinertzhagen
Living reference work entry

Later version available View entry history

DOI: https://doi.org/10.1007/978-1-4614-7320-6_275-1


Synaptic Contact Antennal Lobe Optic Lobe Reporter Line Drosophila Brain 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


The Drosophila connectome is a comprehensive description of all the connections between subunits comprising the central nervous system (CNS) of the fruit fly, Drosophila melanogaster.

The CNS of Drosophila consists of the brain and ventral ganglion. It has a volume of ~0.07 mm3 (see Rein et al. 2002, Table 1). To compare, this is similar to the volume of a single cortical column in mouse barrel cortex (~0.09 mm3) (Lefort et al. 2009; Hooks et al. 2011). The Drosophila brain can be subdivided into smaller neuropil compartments (Fig. 1), identified through anatomical boundaries (Ito et al. 2014), as well as the overlapping arborizations of the neurons constituting each neuropil compartment (Chiang et al. 2011; Jenett et al. 2012). In total, across these compartments, the Drosophila brain is thought to contain ~90,000 cells, of which ~90 % are neurons. These neurons can be classified into discrete cell types, with stereotypical shapes and synaptic connections, using both human observation and genetic reporter lines that label individual types.
Fig. 1

Drosophila brain with anatomically identified neuropil compartments individually colored (www.flybrain.org (AB00122)). Scale bar: 100 μm

A Drosophila connectome, within the brain, can be constructed at different levels of resolution. At low resolution, a connectome can be said to comprise neuropil compartments, with connections between them defined by neural pathways, as identified by light microscopy. In contrast, at high resolution, a connectome comprises individual neurons and all the chemical synapses between them. The components of such a cellular connectome are identified through electron microscopy (EM).

As of 2013, neuropil-level connectomes have been constructed for the entire Drosophila brain (Chiang et al. 2011), but only partial cellular connectomes have been reported. In particular, cellular connectomes have been reported for a cartridge of the lamina (Meinertzhagen and O’Neil 1991; Rivera-Alba et al. 2011) and a column of the medulla (Takemura et al. 2013), the first and second neuropil compartments, respectively, within the fly optic lobe (transmitting visual information). Despite their incomplete nature, these partial connectomes have already provided valuable insights into the computational functions of neuronal circuits.

Existing Neuropil-Level Connectomes

Large-scale, compartmental maps of the Drosophila brain have been obtained through the use of genetic reporters imaged by light microscopy (Chiang et al. 2011). Each reporter line fluorescently labels a sparse subset of neurons (Jenett et al. 2012; Ito et al. 2013). Aligning images from multiple lines to a standard Drosophila brain, the overlapping arbors of locally arborizing neurons are used to define individual compartments: local processing units (LPUs) (Chiang et al. 2011). The connections between these LPUs are found by identifying neurons projecting between them. In total, the resulting connectome contains 58 tracts between the 41 identified LPUs, a sparse subset (<10 %) of the 820 numerically possible LPU-LPU connections.

Cellular Connectomes and EM

A neuropil-level connectome, generated from light microscopy, has insufficient resolution to identify synapses, and hence cannot detail the true connections between neurons. The large-scale, dense identification of all synapses within a region of neuropil requires higher resolution EM-based methods, typically serial-section transmission EM (ssTEM) (Fig. 2). However, these techniques are labor intensive and time consuming. Therefore, only a few partial cellular connectomes have been reconstructed, specifically for two Drosophila optic lobe neuropils.
Fig. 2

Reconstructing a cellular connectome using ssTEM. (a) TEM image of part of a 40 nm thick section, one of 2,769 sections imaged to reconstruct the medulla connectome. (b) Applying computational segmentation algorithms, followed by human proofreading on the TEM images, is necessary to isolate neuronal cross sections (neurite profiles in single colors). (c) Profiles in consecutive sections (left) are linked to construct 3D neurites (right). (d) A subset of reconstructed neurons, embedded within the imaged neuropil. (e) Rendering of neurites belonging to 379 reconstructed neurons. Scale bar: 500 nm (a, b)

Optic Lobe Cellular Connectomes

The optic lobe of flies has been the subject of experimental interest for over 50 years, driven by the advanced visual behaviors of flies (Heisenberg and wolf 1984). Despite the large size of the optic lobe, encompassing almost 50 % of the total CNS volume (Rein et al. 2002), the modular structure of its underlying neuropil compartments, corresponding to the ommatidia array of the compound eye, is highly amenable to connectomics.

Within the optic lobes, synapses are divergent polyads (Fig. 3), i.e., each presynaptic specialization (or “T-bar”) is typically associated with four to six postsynaptic dendrites. This divergence is expected to generalize to non-optic-lobe neuropils, though the number of postsynaptic targets may vary across neuropils.
Fig. 3

Divergent polyadic synapse. The presynaptic process, identified with a T-bar (red arrow), is associated with four postsynaptic dendrites (blue arrowheads). These postsynaptic processes can be distinguished from adjacent, unconnected processes (e.g., green circle) by their postsynaptic densities. Scale bar: 250 nm


The lamina receives most of its input directly from photoreceptors in the retina. It is thought to be mostly responsible for encoding contrast and for light adaptation. A cellular connectome of the repeating unit of the lamina, the lamina cartridge, has been reconstructed twice, in two different wild-type strains (Meinertzhagen and O’Neil 1991; Rivera-Alba et al. 2011). Both reconstructions are similar, with most differences apparently introduced by improvements in reconstruction quality and by differences in the interpretation of some postsynaptic elements. The consensus lamina connectome contains the terminals of six photoreceptors, along with 12 types of neurons (Rivera-Alba et al. 2011; Tuthill et al. 2013). In total, each cartridge contains ~480 presynaptic sites, contacting ~1,250 postsynaptic sites.


Directly downstream of the lamina, the medulla is the single largest neuropil of the fly’s brain and receives visual input both through the lamina and directly from a subset of photoreceptors. It is thought to be responsible for (1) the computation of local visual motion, by the comparison of inputs from different points in space, and (2) color vision, utilizing the direct inputs from spectrally tuned photoreceptors.

A connectome of the repeating unit within the medulla, the column, was reconstructed recently (Takemura et al. 2013) (Fig. 4). It reports the connections between 56 identified cell types, of which 27 are modular (contained in every column). As in the lamina, each presynaptic site contacts multiple postsynaptic elements (~3.8 on average). Within a column, the modular cell types are connected by ~2,500 synaptic contacts. However, this is only a subset of the ~5,800 synaptic contacts that these same modular cell types make in total, showing clearly that there are far more cross-column connections in the medulla than in the lamina (Riviera-Alba et al. 2011).
Fig. 4

3D graph of a connectome module: the connections between neurons found in every medulla column. The dominant direction of signal flow is oriented into the page. Cell types with stronger connections are positioned closer to each other. The colors identify three spatially segregated groups of neurons and, hence, three pathways of information transmission within the medulla

Characteristics of the Drosophila Connectome

Presynaptic terminals within the existing connectome reconstructions are typically of uniform size, and the number of synaptic contacts between different neuron pairs ranges over more than two orders of magnitude (1–150). Hence the synaptic weight for any pair is thought to be well approximated by the number of synapses in parallel. However, it is not known if it will be possible to generalize this quantitative assumption to the rest of the Drosophila connectome.

Synapse numbers for identified neurons are consistent across neurons of the same type in the lamina and medulla. However, as has been seen in both the antennal lobe (Chou et al. 2010) and the mushroom body (Murthy et al. 2008; Caron et al. 2013), this may not generalize to other parts of the brain. In general, learning associated or other forms of plasticity may decrease the stereotypy in the numbers of synaptic contacts, thereby increasing the difficulty of obtaining a single, defined connectome.

The Utility of Connectome Information: Functional Studies

One rationale for compiling a cellular connectome is to gain insight into the neuronal implementations of behavior. The existing optic lobe connectomes have already demonstrated such utility. They have been used to direct the attention of genetic experiments towards a specific neuron(s) (Gao et al. 2008) and to provide the neuronal circuit substrates for computational analyses (Takemura et al. 2013). For example, the comprehensive nature of the cellular connectome reconstruction in the medulla allowed the identification of candidate neurons comprising a circuit computing local motion, a computation that had been intensely studied for more than 50 years (Borst et al. 2010), yet with limited prior insight.

In Progress

As of 2013, ongoing Drosophila reconstruction projects include the larval CNS connectome (Cardona et al.), and the connectome for seven medulla columns, utilizing a new imaging technique: focused ion beam milling scanning EM (FIB-SEM). FIB allows improved z-axis resolution, enabling more complete tracing of neurites thinner than the section thickness. This results in a more complete connectome, given that only 50 % of the postsynaptic sites could be traced to an identified neuron in the ssTEM dataset. Further, by including multiple columns, the resulting connectome should also allow an estimate of the variation between columns, as well as the exploration of color vision circuits (Franceschini et al. 1981).

The Future

We are obviously at the early stages of assembling the components of the Drosophila cellular connectome. However, from the current rapid pace of improvements in sample preparation, imaging, and software tools for reconstruction, it seems possible that the cellular connectome will become available for the entire fly’s CNS. This will require improvements in staining and sample preparation, increased reliability and speed of imaging, and improved automated image analysis algorithms. Additionally, information on the polarity of synaptic transmission, and the presence of electrical synapses, neither visible from EM, must be incorporated from other techniques (Meinertzhagen and Lee, 2011). However, even prior to the completion of the entire connectome, the demonstrated utility of the existing pieces of the Drosophila connectome suggests that each additional reconstruction should provide valuable biological insight.


  1. Borst A, Haag J, Reiff DF (2010) Fly motion vision. Annu Rev Neurosci 33:49–70PubMedCrossRefGoogle Scholar
  2. Caron SJ, Ruta V, Abbott L, Axel R (2013) Random convergence of olfactory inputs in the Drosophila mushroom body. Nature 497:113–117Google Scholar
  3. Chiang AS et al (2011) Three‐dimensional reconstruction of brain‐wide wiring networks in Drosophila at single‐cell resolution. Curr Biol 21:1–11PubMedCrossRefGoogle Scholar
  4. Chou Y-H et al (2010) Diversity and wiring variability of olfactory local interneurons in the Drosophila antennal lobe. Nat Neurosci 13:439–449PubMedCentralPubMedCrossRefGoogle Scholar
  5. Franceschini N, Kirschfeld K, Minke B (1981) Fluorescence of photoreceptor cells observed in vivo. Science 213:1264–1267PubMedCrossRefGoogle Scholar
  6. Gao S et al (2008) The neural substrate of spectral preference in Drosophila. Neuron 60:328–342PubMedCentralPubMedCrossRefGoogle Scholar
  7. Heisenberg M, Wolf R (1984) Vision in Drosophila. Genetics of microbehaviour. SpringerCrossRefGoogle Scholar
  8. Hooks B et al (2011) Laminar analysis of excitatory local circuits in vibrissal motor and sensory cortical areas. PLoS Biol 9:e1000572PubMedCentralPubMedCrossRefGoogle Scholar
  9. Ito K, Shinomiya K, Ito M et al (2014) A systematic nomenclature for the insect brain. Neuron (in press)Google Scholar
  10. Ito M, Masuda N, Shinomiya K, Endo K, Ito K (2013) Systematic analysis of neural projections reveals clonal composition of the Drosophila brain. Curr Biol 23:644–655Google Scholar
  11. Jenett A et al (2012) A GAL4‐driver line resource for Drosophila neurobiology. Cell Rep 2:991–1001Google Scholar
  12. Large VRML model with brain strutures labeled. Flybrain on‐line: http://www.flybrain.org. Accession number: AB00122
  13. Lefort S, Tomm C, Floyd Sarria J-C, Petersen CC (2009) The excitatory neuronal network of the C2 barrel column in mouse primary somatosensory cortex. Neuron 61:301–316Google Scholar
  14. Meinertzhagen IA, Lee C-H (2011) The genetic analysis of functional connectomics in Drosophila. Adv Genet 80:99–151CrossRefGoogle Scholar
  15. Meinertzhagen I, O’neil S (1991) Synaptic organization of columnar elements in the lamina of the wild type in Drosophila melanogaster. J Comp Neurol 305:232–263PubMedCrossRefGoogle Scholar
  16. Murthy M, Fiete I, Laurent G (2008) Testing odor response stereotypy in the Drosophila mushroom body. Neuron 59:1009–1023PubMedCentralPubMedCrossRefGoogle Scholar
  17. Rein K, Zöckler M, Mader MT, Grübel C, Heisenberg M (2002) The Drosophila standard brain. Curr Biol 12:227–231PubMedCrossRefGoogle Scholar
  18. Rivera‐Alba M et al (2011) Wiring economy and volume exclusion determine neuronal placement in the Drosophila brain. Curr Biol 21:2000–2005PubMedCentralPubMedCrossRefGoogle Scholar
  19. Takemura SY et al (2013) A visual motion detection circuit suggested by Drosophila connectomics. Nature 500:175–181PubMedCentralPubMedCrossRefGoogle Scholar
  20. Tuthill JC, Nern A, Holtz SL, Rubin GM, Reiser MB (2013) Contributions of the 12 neuron classes in the fly lamina to motion vision. Neuron 79:128–140PubMedCentralPubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  • Arjun Bharioke
    • 1
    Email author
  • Louis K. Scheffer
    • 1
  • Dmitri B. Chklovskii
    • 1
  • Ian A. Meinertzhagen
    • 2
  1. 1.Janelia Farm Research CampusHoward Hughes Medical InstituteAshburnUSA
  2. 2.Department of Psychology and NeuroscienceDalhousie UniversityHalifaxCanada