Entire fly brain imaged at nanoscale resolution

Editorial

Rebecca Pool

Friday, July 20, 2018 - 12:15
Image: The fruit fly brain's 100,000 neurons can now be traced in detail (coloured threads) using a dataset of some 21 million images. [Z. Zheng et al./Cell 2018]
 
Researchers from the US-based Howard Hughes Medical Institute’s Janelia Research Campus have used a custom-built electron microscopy platform to image the entire brain of a fruit fly to synaptic resolution.
 
Results are publicly available and the latest data offers a new tool for researchers racing to map these connections.
 
As lead neuroscientist Davi Bock points out, the fly's poppy seed-sized brain is surprisingly sophisticated.
 
“[Fruitflies] can learn and remember. They know which places are safe and dangerous. They have elaborate sequences of courtship and grooming," he says.
 
“The entire fly brain has never been imaged before at this resolution that lets you see connections between neurons throughout the entire brain," he adds. "Any time you look at images with higher resolution and more completeness, you’re going to discover new things."
 
Using imaging data from a transmission electron microscope, scientists have reconstructed a subset of neurons in the female fly brain. The work revealed a new cell type that sends information to Kenyon cells in a memory centre of the brain. [Z. Zheng et al./Cell 2018]
 
Fruit flies have around 100,000 neurons, while humans have some 100 billion.
 
To image these cells, Bock and colleagues turned to serial section TEM, using two high-speed FEI Tecnai Spirit bioTWIN TEMs to collect 21 million images from 7062 brain slices.
 
A female fly brain was dissected, fixed, stained, infused with resin, carbon coated and then serial sectioned using a Leica UC-6 ultramicrotome.
 
The 7062 slices were between 35 and 40 nm thick, and took around 3 weeks to produce.
 
To image the slices, the researchers used high-speed Fairchild megapixel and sciMOS cameras as well as a custom-built single-axis fast stage and an automated transport and positioning system.
 
Using the custom-built systems, the slices could be rapidly moved in eight micron increments, allowing the researchers to quickly capture images across adjacent regions of the brains.
 
In this way, the researchers could image an entire brain slice in less than seven minutes, five times faster than a previous high-throughput TEM camera array, TEMCA1.
 
The millions of images Bock’s team collected and stitched together offer an in-depth look at the fly brain.
 
Bock and colleagues have traced the paths of neurons that reach out to the mushroom body, a region involved in memory and learning.
 
These cells, called olfactory projection neurons, have been well described previously, using light microscopy.
 
Manually tracing the outlines of these neurons and all their wirelike projections let Bock’s team confirm the quality of their image data.
 
Olfactory projection neurons send messages to neurons called Kenyon cells, which in turn, communicate with different sets of neurons.
 
Until now, researchers hadn’t identified Kenyon cells’ conversation partners in a region of the mushroom body called the calyx.
 
Bock’s team pinpointed some of these neurons, as well as a previously unknown brain-spanning neuron that also relays information to Kenyon cells.
 
According to Bock, the olfactory projection neurons also appeared to be more tightly bundled together than scientists had thought, suggesting an orderly structure in something once believed to be largely random.
 
“We think [this information] will tell us something about how the animal learns; how it associates odours with a reward or punishment,” he says.
 
The data were acquired over a period of ∼16 calendar months, and according to the researchers, assuming a 2000-hour work year (50 weeks at 40 hours per week) for ∼16 months, it took ∼2,666 microscopist hours to acquire the ∼106 TB whole-brain dataset.
 
Users can download the fly brain dataset at http://temca2data.org.
 
Research is published in Cell.
 
Website developed by S8080 Digital Media