Caitlin Smith has a B.A. in biology from Reed College, a Ph.D. in neuroscience from Yale University, and completed postdoctoral work at the Vollum Institute.
Imaging cells, tissues or even whole animals in three dimensions helps researchers put physiology into biological context. For human subjects, 3D modalities such as PET and CT provide volumetric scans of entire organs or even the body. But microscopes, too, can report 3D information, by layering planar images into so-called “z-stacks.”
One common approach to creating a z-stack is confocal microscopy, but another approach is rapidly gaining popularity. Light-sheet fluorescence microscopy (LSFM, also called selective plane illumination microscopy, SPIM), enables developmental biologists, for example, to image an entire living, developing embryo with minimal harm or photodamage, for extended time periods. Although this type of microscopy is still largely accomplished with home-built microscopes, at least one commercial system is now available. Here we put this intriguing new technique under the microscope.
How it works
Light-sheet microscopy is distinct from other light-microscopy techniques in several important ways. Like other methods, it is fluorescence-based, using a laser to illuminate a fluorescently labeled sample. But instead of illuminating the entire sample (as in widefield or epi-illumination microscopy) or focusing the light into a point (as in confocal laser scanning microscopy), LSFM projects a literal light sheet onto a single plane of the specimen, which is imaged from a perpendicular direction. As that sheet is moved up and down, LSFM sections the sample optically, producing a series of 2D images. But unlike a confocal laser scanning microscope, a laser sheet microscope sections the specimen all at once, not line by line.
To do this, LSFM systems use two objectives: one in the illumination path to focus the laser light into a sheet and the other in the detection path to capture the fluorescence signals from the sample. These signals are collected through the second objective by a CCD camera, which sends the data to a computer.
Because LSFM only illuminates one plane and collects data from the entire plane at once, it causes less photodamage overall compared with other types of microscopy (in which an area greater than the collection area must illuminated with each sampling event). Another unique feature of LSFM is sample positioning. The microscope is oriented horizontally, so the sample is embedded in a substance like agarose and then suspended from above in the end of a capillary tube.
LSFM is faster—imaging more than 200 frames per second, according to John Allen, a microscopist at the National High Magnetic Field Laboratory at Florida State University—and can image a larger area compared with confocal microscopy. “The camera has [a] large field of view—it’s huge compared to a confocal,” says Scott Olenych, product marketing manager of imaging products at Carl Zeiss Microscopy. “It would take a lot more time to image with a confocal.” One reason is that LSFM systems detect the entire illuminated plane in one take. This makes them faster than a confocal laser scanning system, which scans the sample point by point with a focused beam of light, pausing to collect the data each time. “A z-stack that takes 10 minutes on a confocal system takes about 30 seconds on a light sheet microscope,” says Olenych. The larger field of view is especially good for studying larger samples, such as tissues or even whole mouse brains.
Increasingly, LSFM is used to study large samples of tissue that have been fixed, fluorescently labeled and cleared of unnecessary structures that can comprise imaging by scattering light or hindering detection during microscopy. LSFM is also ideal for developmental biology. “The development of model organisms, including Drosophila, medaka fish, zebrafish and more can be monitored for several days on end, starting at the single-cell stage,” says Allen.
Philipp Keller, a group leader at the Howard Hughes Medical Institute’s Janelia Farm Research Campus, has used advanced light-sheet microscopy, including his homegrown, four-objective SiMView technique and custom computational methods to study neural development in live Drosophila and zebrafish embryos, with subcellular resolution.
One area in which confocal microscopy still bests LSFM is spatial resolution. In LSFM, “the spatial resolution is your widefield illumination, about 300[nm to] 400 nm or maybe a little smaller,” says Olenych. “But confocal [spatial resolution] can be much smaller, because you can use oil objectives, which puts you in the 200-nm range.” Another challenge with LSFM is data storage—the experiments tend to produce mammoth datasets. “Imaging big objects, very quickly, over long periods of time—it’s tricky to manage what data you want to keep, store and analyze,” says Olenych.
Commercial fledgling
Carl Zeiss recently launched perhaps the first commercial LSFM system, the Lightsheet Z.1. Looking less like a microscope than a microarray scanner – the system is a sealed box with no external eyepieces or stage -- the Lightsheet Z.1 features a single-photon laser light source and 20x W-PlanApoChromat 1.0 NA objective with a 0.4–2.5x optical zoom; 5x, 40x and 63x objectives are also available. (see system specs here (PDF))
According to Olenych, a unique feature of the Z.1 is that users can rotate the sample to any angle. This, along with the system’s speed, means you can easily image different z-stacks with the sample oriented in almost any position. By putting all these z-stacks together, the computer can build a multiview, 3D reconstruction of the sample. “This helps to improve resolution on parts of the sample that might have had less light due to angles or rotation,” says Olenych.
Zeiss Researchers in the lab of Michael Davidson, director of the Optical Microscopy Division of the National High Magnetic Field Laboratory at Florida State University, eagerly await their first Lightsheet Z.1 system, says Allen. “We plan to use it to characterize protein dynamics in 3D cell-culture systems using a variety of genetically expressed fluorescent proteins.”
DIY laser sheet microscopy
Before LSFM systems were commercially available, researchers like Keller needed to build their own. Many still do, because it is relatively inexpensive and the systems are less complex to construct than one might expect. In fact, an online community called OpenSPIM was created to support scientists who wish to build their own systems. An open-access platform, OpenSPIM provides instructions and advice for researchers building or improving their own SPIM/LSFM platforms.
Much of the innovation in LSFM comes from home-built varieties, born as researchers try to solve particular experimental problems. Most LSFM systems, including the Zeiss Lightsheet Z.1, use a single-photon laser for illumination. But improvements are in the works, aimed at improving the resolution (such as with a Bessel beam laser) and the depth of light penetration into the sample surface (such as with a multi-photon laser). Eric Betzig, for instance, also at Janelia Farm, developed a variation of LSFM that uses Bessel beams to image living cells with 3D isotropic resolution at hundreds of image frames per second.
Olenych predicts that the interest level in LSFM will grow quickly as more researchers see what it can do. This in turn will help to further the technique’s development, as well as innovation by the community of home-built LSFM microscopists. “They are always pushing in different directions, for example using different types of illumination or super-resolution,” he says. “That’s often where new ideas come from.” With researchers reaching out for new solutions, LSFM is likely to yield astonishing new results.
Image: Carl Zeiss Microscopy Lightsheet Z.1