Scientists have developed a new genetic method for labeling cells that transforms the tangles of neurons within the brain of a fruit fly into fantastic rainbows of color. Images created by the new technique, called dBrainbow, are visually arresting – but their real power lies in the information they contain.
dBrainbow, which was developed by Julie Simpson and her team at the Howard Hughes Medical Institute’s Janelia Farm Research Campus, offers researchers a wide palette of colors with which to simultaneously label individual neurons. The vibrant hues help scientists “paint” different types of neurons so they can trace their paths in the brain and understand how they wend their way in relation to other neurons. The tool will enable scientists to generate higher resolution maps of the fly brain, as well as learn how different cellular lineages contribute to neural circuits. The technique is described in the March 8, 2011, issue of the journal Nature Methods.
“The more people who use it, the more robust a tool it is.”
Julie H. Simpson
dBrainbow expands the reach of a strategy that researchers from Harvard University first introduced in mice in 2007. The technique, dubbed 'Brainbow' by its creators, Jeff Lichtman and Joshua Sanes, uses a rainbow of fluorescently labeled proteins expressed in developing brain cells and combining to produce more than 100 distinct hues.
Brainbow has enabled researchers to see the mouse brain, and peripheral nerves, in unprecedented detail. A version of Brainbow has since also been applied to visualize neurons in the zebrafish brain. Now, dBrainbow will enable new investigations of the brain of the fruit fly Drosophila melanogaster, one of the most studied organisms in biology and a powerful system for learning about the neural basis of behavior.
In her Janelia Farm laboratory, Simpson is working to understand how individual neurons control behavior in the fly. Flies offer a slate of behaviors for study – they eat, taste, sleep, and mate --but rely on
fewer neurons than more complex animals to produce those behaviors. With roughly 100,000 neurons, the connections between brain and behavior are more accessible in the fruit fly than in mice and humans, which have millions of neurons. .
Nonetheless, pinpointing the precise neurons responsible for driving a fly to eat or sleep, for example, is not trivial. One of the challenges is that most strategies used to genetically direct neurons to produce a particular protein – such as the fluorescent markers that make them visible under a microscope -- target groups of cells, not individual cells. "So you have a super cool phenotype, the ability to target 100 neurons, and you guess that only 10 of those are important for the behavior," Simpson says. "How do you dissect which neurons are really relevant in these complicated anatomical patterns?"
A couple of techniques developed in the past decade have let scientists target individual neurons or groups of cells in Drosophila. For example, a technique called mosaic analysis with repressible cell markers, or, developed by HHMI investigator Liqun Luo at Stanford and Tzumin Lee, now a Janelia Farm group leader, is used to label neurons that have a common origin. The powerful strategy has uncovered new genes involved in brain development. To get a comprehensive view of the brain, however, requires many separately labeled brain preparations. Scientists typically spend a lot of time aligning these separate images using computational tools to get the full picture.
The Brainbow method overcomes those problems, but applying the technique to flies required developing tools and methods distinct from those that had been fine-tuned in mice. Most Drosophila imaging works using only two colors — green and red fluorescent proteins — but to do Brainbow in any animal, you need a lot of bright and stable colors, Simpson says. The main challenge was picking, combining and optimizing these fluorescent labels so that they could be seen easily using standard microscope equipment. The group tried every fluorescent label possible, taking several years of trial and error, tweaking and troubleshooting. "I completely underestimated how hard it would be to get it all finished," she says.
A separate team, led by Iris Salecker at the National Institute for Medical Research in London, reports similar findings in the same issue of Nature Methods, nicknaming their approach 'Flybow.' "It turns out to be quite helpful because we made very different choices at every branch point," Simpson says. "So I think the tools will be beautifully complementary and that people should use them both. And I was very pleased that we were able to get our papers into the same journal at the same time."
To verify that dBrainbow works, Simpson's team looked at one of the most anatomically well characterized parts of the fly brain: projection neurons that process the sense of smell. "We picked an area of the brain that has been really well studied and where the MARCM analysis had been done. We knew what the right answer would be," she says. Indeed, Simpson's team was able to recapitulate characteristics of the projections in a single brain what previously took many.
One variation of the dBrainbow fly allowed Simpson's team to look at populations of neurons containing octopamine, a common brain signaling molecule that controls aggression, ovulation, and learning. Previous studies of these neurons were unable to resolve different classes of octopaminergic neurons, partly because the neurons show different projection patterns in different flies. Using dBrainbow, it was easier to distinguish the different types of neurons within the octopaminergic population.
Another benefit to the method is that it can be used to label neurons from a common origin in the same color. "So it lets you look at whole lineages in relation to other whole lineages," Simpson says. For instance, her team labeled three separate neuron lineages in a single fly and looked at their projections. The group realized in their particular experiment that two neurons from the same lineage might go to the same place but that two neurons from different lineages will go to different places. "It's very obvious to see that if you have each lineage in a different color — you can see that they just don't overlap," Simpson notes.
Simpson's group plans to make minor tweaks to the technique, but most future work will focus on unraveling the brain circuitry that drives complex behaviors like grooming and eating. In the new study, Simpson's team mapped individual motor neurons to muscles that control proboscis extension. The overall goal is to understand how the fly takes in sensory cues, integrates it with previous experiences and responds with the appropriate behavior.
The tool will likely benefit many others. Janelia Farm group leaders James Truman and Tzumin Lee are already using dBrainbow to understand how different neuron lineages work together to make functional circuits and compose the brain, and Simpson says it’s also being used by other labs for studies of sleep and development. "We've been sending out the flies and constructs to anybody who's asked for it all along," she says. "I don't see any reason why we should not be generous with the reagent."
And sharing will ultimately make the tool better, too, Simpson says. "The more people who use it, the more robust a tool it is," she says.