Brain bonanza: Researchers can search within the connectome data for specific neurons, brain areas, neurotransmitters and more.
Courtesy of CATMAID

Full connectome of adult fruit fly completed, with help from citizen scientists

The map, by far the largest one of an entire brain to date, contains 130,000 neurons and 53 million synapses.

A team of researchers and citizen scientists has mapped all of the circuits in the entire brain of an adult fruit fly.

The connectome, described in a preprint posted to bioRxiv in June, traces 53 million synapses among nearly 130,000 neurons in Drosophila melanogaster, making it the largest brain map to date. It comes nearly 40 years after researchers finished the first-ever connectome, which mapped all 302 neurons in the nematode Caenorhabditis elegans’ nervous system.

“It’s a huge step up from the connectome of C. elegans,” says Donald O’Malley, associate professor of biology at Northeastern University in Boston, Massachusetts, who was not involved in the new work. “This is really a big breakthrough.”

Complete connectomes already exist for the larval sea squirt Ciona intestinalis, the marine worm Platynereis dumerilii and the larval fruit fly; in 2020, researchers unveiled a partial adult fruit fly map, often referred to as the “hemibrain,” consisting of 25,000 neurons in the central brain.

To create the hemibrain, a team led by researchers at the Janelia Research Campus in Ashburn, Virginia, imaged a portion of the fly brain with electron microscopy, reconstructed the brain via machine learning, and manually reviewed and “proofread” the connections mapped by the computer.

The new connectome is the result of a similar approach carried out by the FlyWire Consortium, which was created in 2019 by Mala Murthy and Sebastian Seung at Princeton University. The consortium involves hundreds of neurobiologists, computer scientists and other contributors around the world.

The consortium’s online platform allows anyone — including citizen scientists — to become trained in the proofreading process and edit the connectome in real time, similar to Wikipedia. Each neuron requires about 19 minutes of proofreading time, according to a 2021 paper describing the platform.

The connectome’s data are publicly available on the Connectome Data Explorer (Codex). Because FlyWire participants also annotated the map, users can search Codex for a range of cell attributes, including brain region, cell type and lineage, and associated neurotransmitters.

Its potential applications are broad, says study investigator Sven Dorkenwald, a graduate student in Seung’s lab. For example, researchers who study particular neurons can pinpoint those neurons’ connections in FlyWire and devise hypotheses about their function that could then be tested in living flies.

“I’m looking forward to seeing what people do and find,” Dorkenwald says. “That’s definitely part of the joy of working on this kind of data.”

T

he FlyWire team built the connectome from electron microscopy images of slices of an adult female fly brain that were published in 2018. Of the 127,978 neurons in the map, 114,423, or 89 percent, have connections only within the brain, a type of cell called an intrinsic neuron. In comparison, only about one-third of neurons in the fruit fly larva brain are intrinsic, according to a complete connectome of a 6-hour-old female fly, published in March.

Nearly 74,000, or 64 percent, of the adult intrinsic neurons are contained within the fly’s vision areas, or optic lobes, compared with about 32,000, or 28 percent, located in the central brain.

“I found that absolutely shocking,” says Michael Winding, group leader at the Francis Crick Institute in London, England, and an investigator on the larval fly connectome, who was not involved in the new work. “It really highlights how visual that animal is.”

The fruit fly larva has a simple eye and about 40 visual neurons, relying primarily on its sense of smell instead. Studying how the brain builds the visual system as the insect transitions from a larva to an adult could help elucidate how new circuits integrate into existing networks, Winding says, and how more complex brain structures evolved on top of older ones.

Making connections: Neurons and synapses in the antennal lobe are visualized in CATMAID, a tool for exploring the connectome data. The cyan and magenta dots indicate pre- and post-synaptic sites.

Within the visual system is a donut-shaped structure called the ellipsoid body that uses visual input to encode a sense of direction — a kind of “compass system,” says Sung Soo Kim, assistant professor of biology at the University of California, Santa Barbara and part of the FlyWire Consortium. When Kim first opened his lab in 2019, he wanted to know which neurons from the visual system feed into this compass and where, in turn, those neurons receive input from.

He decided to train his students to use the FlyWire platform to map them. By 2021, he realized they had mapped 70 percent of their neurons of interest. He decided they should finish the job.

By mapping all of these neurons, they’ve been able to make predictions about which types of neurons feed different types of information into the ellipsoid body — predictions they are now testing with methods such as electroencephalography. Kim and his colleagues plan to detail their findings in a preprint this month.

“We are not doing this anymore in the darkness,” Kim says. “We can design our experiments informed by this anatomy.”

A

bout 65 percent of disease-related genes in people have homologs in fruit flies, Winding says, and many of those are related to brain development. Researchers who study neurodevelopment and conditions such as autism could mutate genes of interest in living flies, identify which neurons are affected by the mutation, and pinpoint their partners in the connectome to better understand the gene’s function, he says.

To explore how much variation exists between individual flies, researchers have begun to look at how the new connectome compares with the 2020 hemibrain. Some cell types are more consistent between the two brains than others, according to a companion FlyWire preprint also posted in June. One-third of the cells in the hemibrain could not be located in the new connectome, “cautioning against defining cell types based on single connectomes,” the investigators wrote.

“It’s a really nice first step into developing tools to compare brains,” Winding says. “As we get bigger and bigger brains, we’ll really have a foothold in what to do next.”

Dorkenwald says the consortium plans to publish more preprints in the coming months that more deeply analyze the connectome data.