A group of researchers has completed the most advanced brain map to date, that of an insect, a historic achievement in neuroscience that brings scientists closer to a true understanding of the mechanism of thought, as published in the magazine Science in its latest issue.
According to the Europa Press agency, the international team led by Johns Hopkins University (United States) and the University of Cambridge (United Kingdom) produced an astonishingly detailed diagram that traces every neural connection in the brain of a larval fruit fly, a archetypal scientific model with brains comparable to humans.
The work will likely inform future brain research and inspire new machine learning architectures. “If we want to understand who we are and how we think, part of it is understanding the mechanism of thought,” says lead author Joshua T. Vogelstein, a Johns Hopkins biomedical engineer who specializes in data-driven projects like connectomics, the study of the connections of the nervous system – and the key to that is to know how neurons are connected to each other.”
The first attempt to map a brain – a 14-year study of the roundworm begun in the 1970s – resulted in a partial map and a Nobel Prize. Since then, partial connectomes have been mapped in many systems, including flies, mice, and even humans, but these reconstructions often represent only a small fraction of the total brain.
Complete connectomes have only been generated from several small species with a few hundred or thousands of neurons in their bodies: a roundworm, a sea squirt larva, and a marine annelid larva.
This team’s connectome of a young fruit fly, the larva of Drosophila melanogaster, is the most complete and extensive map of an insect brain ever made. It includes 3,016 neurons and all the connections between them: 548,000.
“50 years have passed and this is the first brain connectome. It is a flag in the arena that we can do it,” Vogelstein points out. “Everything has worked until we got to this.”
Mapping entire brains is difficult and time consuming, even with the best modern technology. To get a complete picture at the cellular level of a brain, it is necessary to divide it into hundreds or thousands of individual tissue samples, all of which have to be analyzed with electron microscopes before the laborious process of reconstructing all those pieces, neuron by neuron, in a complete and accurate portrait of a brain.
It took over a decade to do that with fruit fly breeding. The brain of a mouse is estimated to be a million times larger than that of a baby fruit fly, which means that the possibility of mapping anything resembling a human brain is unlikely in the near future.
The team purposely chose the fruit fly larva because, for an insect, the species shares much of its fundamental biology with humans, including a comparable genetic base.
In addition, it has rich learning and decision-making behavior, making it a useful model organism in neuroscience. For practical purposes, its relatively compact brain allows it to be imaged and its circuits reconstructed in a reasonable amount of time.
Even so, the work took 12 years at the universities of Cambridge and Johns Hopkins. Imaging alone took approximately one day per neuron.
The Cambridge researchers created the high-resolution images of the brain and scanned them manually to find individual neurons, rigorously tracing each one and relating their synaptic connections.
Cambridge turned the data over to Johns Hopkins, where the team spent more than three years using the original code they created to analyze brain connectivity. The Johns Hopkins team developed techniques to find groups of neurons based on shared connectivity patterns, and then looked at how information might spread through the brain.
In the end, the entire team graphed each neuron and each connection, and ranked each neuron by its role in the brain. They found that the most active circuits in the brain were those going to and from neurons in the learning center.
The methods developed by Johns Hopkins are applicable to any brain-wiring project, and their code is available to anyone trying to map an even larger animal brain, Vogelstein said, adding that despite the challenges, scientists are expected to face the mouse, possibly in the next decade. Other teams are already working on a map of the adult fruit fly brain.
Co-first author Benjamin Pedigo, a Johns Hopkins doctoral candidate in biomedical engineering, hopes the team’s code can help reveal important comparisons between adult and larval brain wiring. As connectomes are generated from more larvae and related species, Pedigo hopes his analysis techniques will provide a better understanding of variations in brain wiring.
Work with fruit fly larvae showed circuitry features strikingly reminiscent of prominent and powerful machine learning architectures. The team hopes that continued study will reveal even more computational principles and may inspire new artificial intelligence systems.
“What we’ve learned about the fruit fly code will have implications for human code,” Vogelstein said. “That’s what we want to understand: how to write a program that drives a human brain network.”
