Microscopic marvels: Seeing the system

“We don’t pretend to have invented new physics,” says Ernst Stelzer modestly, standing next to an equally modest layout of lasers, mirrors and lenses. “It has all just been plain common sense.” Stelzer has been applying his common sense to microscopes ever since he arrived at the European Molecular Biology Laboratory in Heidelberg, Germany, in 1983 as a fresh-faced physics PhD student, and stayed on to head a team developing three-dimensional light microscopy. A quarter of a century in the field has now led to single plane illumination microscopy (SPIM), the technique with a modest face but an extravagant view: beautiful and unprecedented moving images of whole organisms as they grow one cell division at a time. When Stelzer started out in the 1980s, a method called confocal fluorescence microscopy was beginning to show huge potential. The technique exploited the labelling of molecules in cells with a fluorescent tag to build up a threedimensional image. But it still relied on traditional light microscopy procedures by which the biological sample is attached to a two-dimensional slide and a beam of light passes through the lens and through the entire sample, stimulating fluorescent emission that is detected by a camera. Two problems with this approach bothered Stelzer. First, the squashed, two-dimensions of the slide were an unnatural environment for cells that live in threedimensional tissue. And second, fluorescent microscopy slowly destroys the very thing that researchers are trying to make visible. The undiscriminating beam illuminates and triggers fluorescence from the entire sample, not just the focal plane. Yet photons are damaging to cells, which — apart from those “This idea had been around for a century, but biologists didn’t realise its potential.” — Ernst Stelzer P. K EL LE R, A . S C H M ID T in skin, eyes or other body surfaces — receive little daylight. Light also ‘bleaches’ the fluorescent tags, which limits the number of times that the sample can be observed. Stelzer determined to develop a microscope that could visualize living biological samples, for long time periods, in conditions that approximate normal physiology. To optimize the use of light, he turned a laser-light source at 90° to the camera and lens. In 2002 , he added the element that is central to SPIM — mechanics that create a micrometre-thin sheet of light in a single plane that sweeps gradually through the specimen (see graphic). “This idea had been around for a century,” he says, “but biologists didn’t realize its potential for high-resolution.” Only the fluorescent tags in this plane of light are excited, and the camera efficiently captures the photons emitted. The specimen is then rotated and the procedure repeated along half a dozen or so additional axes. Eventually all the planar images are merged together computationally into a three-dimensional whole. By eliminating excess illumination and dye photobleaching, the whole process can be repeated every two minutes or less for more than 24 hours . The specimen to be viewed — intact, in three dimensions — passes the time in a tiny, transparent cylinder filled with agarose gel to dampen any movement and perfused with physiological levels of gases such as oxygen and carbon dioxide . Stelzer, working with colleague Joachim Wittbrodt, has shot spectacular movies of the first day in the life of a zebrafish embryo, a species loved by microscopists for its transparent skin (P. J. Keller et al. Science 322, 1065–1069; 2009). As development is triggered, a single oval cell undulates, seeming to squeeze out additional complexity. More