Image sequence of a crawling Drosophila larva with GFP-labeled muscles and central heart tube. Image reproduced from Bouchard et al.1, Nature Publishing Group.

Imaging neuronal activity or other dynamic processes in awake animals is challenging owing to their movements, which can lead to artifacts during imaging. To overcome this problem, Elizabeth Hillman and her collaborators at Columbia University developed a high-speed volumetric imaging technique inspired by advances in light-sheet microscopy.

The idea of applying light-sheet microscopy to dynamic samples initially seemed crazy to Hillman because typical samples for light-sheet microscopy are dead, and thus there is generally no need for fast live imaging. The design of light-sheet microscopes requires the alignment of two or even four objectives, imposing sample constraints. Hillman and her colleagues came up with a microscope design that gets by with a single objective. The light sheet exits the objective from the edge and illuminates the sample at an oblique angle. The same objective collects the emitted light from the illuminated plane. With these modifications, “you don't need to prepare your sample, you don't need to hold your sample,” says Hillman, making it possible to image live animals.

The increase in imaging speed necessary for live imaging required further adaptations compared to established fast light-sheet microscopy implementations. Traditional setups can reach a volumetric imaging speed of 1 Hertz by moving the imaging plane with the help of an electrically tunable lens or piezoelectric translation. Hillman and her colleagues achieved a higher imaging speed of 20 volumes per second by keeping the stage and the objective stationary. To acquire image volumes, they swept the light sheet through the sample by reflecting the light from a slowly rotating polygonal scanning mirror. Coincidentally, the light returning from the sample was similarly deflected, thereby maintaining the alignment of incoming and outgoing light paths. In light of these two modifications, the researchers named their technique swept, confocally aligned planar excitation (SCAPE) microscopy.

“The system we have now is so versatile and usable that you can literally just stick anything under there,” says Hillman. Together with her collaborators, she demonstrated the capabilities of the SCAPE microscope by imaging neuronal activity in head-fixed, awake mice and in crawling Drosophila larvae. The high imaging speed allows monitoring of calcium activity in large volumes, which was not possible with imaging techniques that are traditionally used for calcium imaging in animals, such as laser scanning confocal imaging.

Hillman thinks that, in comparison to other fast imaging techniques, SCAPE microscopy has two main advantages. First, neither the objective nor the sample is moved, which makes the technique exceptionally fast but also does not disturb the sample in any way. And similarly to other light-sheet microscopy techniques, only a single plane is illuminated at any given time, thereby lowering phototoxicity, which is important in live imaging.

Although the current implementation of the SCAPE microscope works well for calcium imaging, the group is continuously improving the system, especially because it does not yet reach the resolution that is theoretically possible. “I am pretty confident that we can reduce the number of lenses in the system and actually have it work better,” says Hillman.