For the First Time, Researchers Capture an Entire Live Brain on Video

At least 80 percent of this entire baby zebra fish brain is observable down to the single-cell level.

Things are moving as fast as a synaptic flash in the science of brain-mapping lately. Just last month, scientists in Japan revealed stunning video footage from “inside” the brain of a larval zebra fish as it hunted for food. The images, though beautiful and profound, were isolated to the optic tectum—a region of the brain devoted to processing visual cues.

According to recent findings published in the scientific journal Nature Methods, a team of scientists this side of the Pacific have taken the next giant leap forward: For all intents and purposes, they have created a live activity map for an entire zebra-fish brain—the first time a whole-brain map has ever been recorded for any creature.

The new research, conducted by scientists at the Howard Hughes Medical Institute, comes at a particularly exciting time. Just last month, President Obama’s proposed a national Brain Activity Map (BAM) project in hopes of doing for the brain what the Human Genome Project did for genetics. Brain size and structure grows ever more complex as one travels up the food chain—and hence ever more challenging to map. And the difference in complexity between human and larval zebra fish brains can hardly be overstated. (The baby zebra fish’s brain comprises about 100,000 neurons; a human brain, by contrast, contains around 90 billion.)

But what we’re seeing here is exactly the sort of imagery scientists hope to recreate for human brains in the next decade. The video footage above makes that future feel mouth-wateringly palpable.

To capture the fireworks we see in the video above, neurobiologist Misha Ahrens and microscopist Philipp Keller first used gene therapy techniques so that the fish’s neurons would express synaptic activity as a synaptic flash. This is a technique similar to the one used in the Japanese experiments. The big difference-maker here was the technology.

Video of a larval zebra fish's brain from last month's Japanese study, restricted to the optic tectum. Source: Current Biology, Muto et al.

The difficulty in most brain-imaging studies comes from trying to capture so much super-high-resolution information in anything close to real time. Even with simple brains, it requires capturing tens of thousands of neuronal flashes in resolution at the single-cell level every second. That requires a camera that’s fast enough at a molecular level to catch what’s going on, but that can also penetrate the brain at multiple levels and capture the images accordingly. (Imagine trying to photograph thin slices of a brain, like what a CT scan provides, but doing a bunch at the same time—otherwise, you’re just photographing the surface).

To do that, the Virginia scientists used a technique called “high-speed light-sheet microscopy” (the light sheet is like the slicer). The technique has been used before, but never so well. “We redesigned the volumetric imaging strategy, eliminated electronic overhead and integrated new detectors,” the researchers explain in the study. In doing so, they “increased the sustained volumetric imaging rate approximately tenfold over previously reported performance” (my emphasis), catching almost everything that happens in the brain every 1.3 seconds (see graphic, below).

Source: Ahrens and Keller, Nature Methods (2013)

Using those techniques, the researchers say, they were able to capture activity from about 87 percent of all the neurons in the baby zebra fish’s brain (some were lost because they were shadowed by the eyes or lost to light refraction); about 92 percent of that was captured at resolution down to the single-cell level, meaning we can see about 80 percent of the entire brain's activity on a neuron-by-neuron basis—an estimate they call “conservative.”

“We see the big picture without losing resolution,” Keller told Nature.

It's difficult to overstate just how profound isolating single sections of the brain in real time is. But figuring out how those sections work in tandem with other parts always required some inference and hypothesis. Piecing together, for example, how a brain apprehends its food in the optic tectum is one thing, but observing if and how it worked with other parts of the brain remains more elusive.

Here we can see it all happening at once. The baby zebra fish in this experiment, though alive, were held very still, so it may be a while before we can go beyond observing one in vivo to also observing one at this depth in situ. But there's no telling what we'll learn once scientists manage to start recording whole-brain responses to controlled stimuli. 

Before we get too excited over its implications for the human brain, it’s worth noting that larval zebra fish present researchers with one distinct and major advantage for this sort of early-stage research: Their heads are transparent. Getting inside a human’s head, by contrast, presents some obvious logistical and ethical challenges. Just a few weeks ago, scientists at Stanford captured video of the neurons firing inside a live mouse’s hippocampus as navigated it an enclosed space. That’s encouraging, but they also mounted the camera inside the mouse’s head, and the resolution is nothing like what we see in this new zebra fish study.

Taken together, though, it’s easy to see we’re well on our way toward building the brain’s version of the human genome—what some are calling the “ connectome.” Let's keep this train rolling.