A new method to map activity and anatomy in the zebrafish brain

The larval zebrafish brain is small and accessible enough that it can be studied as a whole to determine how this entire brain generates behavior. Using high-throughput brain imaging and automated analyses, HFSP Long-Term Fellow Owen Randlett and colleagues have developed new methods to create and analyze brain-wide activity maps made from freely behaving zebrafish. By imaging over 1600 fish, they have identified areas in the brain that respond to various behavioural and pharmacological stimulations, cell types that may underlie these signals, and potential functional associations among brain regions.

HFSP Long-Term Fellow Owen Randlett and colleagues
authored on Mon, 12 October 2015

The holy grail for neuroscientists is to understand how the symphony of activity in our 100 billion neurons acts to control our thoughts and actions. The NIH BRAIN Initiative aims to make this possible by developing tools and techniques to record neural activity across the entire human brain, and to create an anatomical census of all of the neurons and their connections. Due to the complexity and scale of the human brain, these goals are only likely to be achieved in the distant future. Yet, within this project the larval zebrafish brain is recognized as an important milestone since it is the only vertebrate for which it is currently possible to record activity from all neurons using calcium imaging. This is possible because zebrafish larvae are very small, only 5 mm long, and they only have about 100,000 neurons. This is roughly a million times smaller than the human brain. Despite such a modest endowment of brain hardware, zebrafish must live independently in the world. At less than a week old, they need to seek out and hunt their own food, navigate in their environment and avoid predators. This balance of a small and accessible brain, yet a complex repertoire of behaviours, make the larval zebrafish a powerful model system for understanding how the vertebrate brain functions as a whole.

Figure 1: The brain in sagittal projections. Top view is a multicolor projection of brain anatomy, bottom view is a functional map showing the areas of the brain that are engaged while fish are hunting and feeding.

Unfortunately, the ability to record activity from all neurons alone will not allow us to understand how the circuits in the brain are working. For this, it is essential to understand activity in the context of neuroanatomy. Are the active neurons excitatory or inhibitory? Which cell type are they, and which other cells are they connected to? Such details are critical to constrain possible models, and to understand the organization and function of neural circuits. Currently, our ability to record activity from zebrafish neurons vastly outweighs our anatomical knowledge of this brain.

 

Figure 2: A multicolor projection of brain anatomy in top view.

In a recent paper, Owen Randlett, Florian Engert, Alex Schier and colleagues have created new methods to analyze the anatomy of the larval zebrafish brain. This was done by building a digital reference brain atlas using image registration of whole-brain image volumes. They aligned nearly 900 individual brains stained for various anatomical features to a common template. The result is the Z-Brain atlas, a “virtual brain” where different features can be visualized, overlaid and quantitatively compared in the same physical space. This reference atlas sets the framework for the collaborative accumulation of comprehensive anatomical and functional data defining the larval zebrafish brain. The Z-Brain can be viewed and downloaded at the project website.  

With their atlas in hand, they next sought a way to easily and rapidly create activity maps. For this they used immunostains of phosphorylated-ERK, which marks active neurons in the brain. By registering fish stained with phosphorylated-ERK to the Z-Brain, they automatically created maps identifying the areas in the brain that are responding to stimuli. This method is particularly powerful because it uses freely behaving animals and so it can be applied in nearly any context. In this paper, they describe areas of the brain that respond to drugs, visual and aversive stimuli, and that are activated during hunting and feeding.

By creating activity maps within the Z-Brain, they are able to analyze the anatomical features underlying activity. This can generate precise hypotheses as to the identity of the relevant neurons in a circuit. Using this approach, they discovered cell types that may mediate turning behaviour induced by whole-field motion, and a grouping of hindbrain neurons that are activated by aversive stimuli.
 

Reference

Whole-brain activity mapping onto a zebrafish brain atlas. Randlett O, Wee CL,  Naumann EA, Nnaemeka O, Schoppik D, Fitzgerald JE, Portugues R, Lacoste MB, Riegler C, Engert F*, Schier AF*. Nature Methods. 2015 doi:10.1038/nmeth.3581

Link to Nature Methods article

Z Brain Atlas