Scientists at Harvard Medical School have combined activity and morphological images to map neuronal circuits in the visual cortex with a homemade electron microscope. The method could help researchers map all of the brain’s active circuits.
R. Clay Reid, a professor in the Department of Neurobiology at Harvard Medical School, led the research team that used two-photon microscopy and electron microscopy to capture the images for the so-called neuron conversation map. The team found that multiple neurons can talk simultaneously to one target neuron in the brain.
“We can see what neurons did for a living in the brain,” said Reid.
To capture the fine details of the neuronal contact points, Reid’s team needed the high resolution that can only be provided by an electron microscope. But high-quality electron microscopes cost well over $250,000, and Reid’s lab did not have the budget to purchase one. So during the summer of 2006, Reid, along with one graduate student and two technicians, built their own for about $100,000 by upgrading a second-hand electron microscope with some spare parts that they purchased.
“We engineered it just well enough to work,” said Reid. “We were competing against groups that had much larger equipment budgets and highly engineered approaches.”
To capture the talking neurons, Reid’s team mounted a mouse on a photon microscope and loaded its brain cells with Oregon Green dye, which binds to calcium and detects active neurons. The team showed the mouse a video filled with visual cues—moving shaded bar patterns—which stimulated the visual cortex. Using two-photon microscopy, his team captured the talking neurons in two dimensions and then took additional pictures above and below the plane to produce a 3-D image of neuron activity.
Using his electron microscope, Reid and his team collected 45 nm–thick slices of the mouse’s brain. For the next two years, Reid’s team stacked over 3.5 million images, each 11 megabytes, together to produce a final 10-terabyte 3-D morphological image of the visual cortex.
With help from the collaborators at Carnegie Mellon University, Reid’s team merged both 3-D images to produce one containing activity and morphological details. Reid located the interneuron through the neuronal contact point and, by using a 3-D modeling program called TrakEM2, worked backward to determine all of the afferent neurons connected to it.
Reid believes the rise of computing power and cheap storage space has removed a major block in the use of electron microscopy in neurobiology. “If the computers existed 40 years ago, people could have done it back then,” said Reid. “Disk storage has increased roughly at a million fold. That’s what’s made it possible.”
The paper, “Network anatomy and in vivo physiology of visual cortical neurons,” was published 10 Mar. 2011 in Nature.