System components presented another challenge. For instance, the original design used a relatively large, expensive (about $250,000, according to Hell), high-energy Ti:sapphire laser as a STED beam. The excitation laser was also large. And both, in Hell's design, operated in the open. To shrink things down and make them safer for general use, the Leica team decoupled the lasers from the system with fiber optics, which had the dual benefit of enabling them to pack components more closely together and increasing system reliability. “If the distance the light travels is short, it is more stable,” Gugel explains.
Also requiring re-engineering was the phase filter, the component that shapes the STED beam into its characteristic torus. Hell's design employed a “programmable spatial light modulator,” made of liquid crystal. As light bounces off this surface, explains Gugel, “it modifies the wavefront to make the doughnut. But [the device] is huge and also quite expensive, and so not an option for us.” Instead, Gugel, an optical physicist, designed (and patented) a new phase filter, which exploits the wavelength changes light undergoes as it passes from air to glass to metal.Lessons Learned
As Leica's engineers labored on STED, they relied on lessons learned from that earlier collaboration with Hell that produced the Leica TCS 4Pi. Though the company sold several units, according to Szellas, the 4Pi was expensive and challenging to use. It imaged samples jointly from above and below, and displayed images from the side (that is, in the x-z plane) rather than from above—something microscopists were not used to. It also employed mathematics to clean up and produce the final picture.
In contrast, operation of a STED microscope is more or less the same as standard confocal microscopy, Szellas says. “What I learned from the 4Pi microscope is that it is very important that the microscope is very easy to use and stable,” says Gugel, who also was involved in the development of the TCS 4Pi.
Leica launched the TCS STED in 2007 with a price of about $1 million. The physical product, says Hell, was a “neatly engineered and compact” system bearing “most of the physical-technical ingredients that we had in the laboratory … [plus] a lot of software that made it very easy to use.” In 2009, the company released a second system called the TCS STED CW. Whereas the original STED system requires dyes (such as ATTO-647) that emitted deep in the red end of the spectrum, the new system is compatible with more popular green fluorophores and fluorescent proteins, thereby simplifying live-cell imaging.
Swapping the expensive Ti:sapphire laser for less-expensive continuous wave lasers and with a simplified setup, the CW version costs about half as much as the first-generation instrument, says Szellas. By the end of this year, he estimates several dozen CW systems will be installed worldwide, plus some 20 Ti:sapphire-based units.
For the end-users, the result of Leica's STED development was technology democratization. According to Szellas, the system made super-resolution “usable for the daily ,research of life science researchers … rather than being accessible only for experienced biophysicists.” One of those first-gen system adopters was Silvio Rizzoli, a group leader at the European Neuroscience Institute in Goettingen.
Rizzoli, who has published 15 papers using STED, did his postdoctoral research at the Max Planck Instutute (though not in Hell's laboratory), and used the original STED installation. Or rather, he obtained data on the instrument—he never actually ran the system himself. “There was always a physicist that would place the sample on the microscope,” Rizzoli recalls. “Sometimes it took weeks of setting up the scope to take a few pictures.” That's because, as an open system, it required significant tweaking to align all the beams to make it work correctly. As a mere biologist, he says, “I didn't use it directly.”
Rizzoli's lab studies the biology of neural synaptic vesicles, and now, using a TCS STED instead, he can run his own experiments. “What we found was that when you fuse the vesicle to a membrane, the [vesicle cargo] molecules do not separate but stay together,” he explains. “You cannot resolve these molecules in a confocal microscope; they appear as a clump. But in STED, you see clusters of synaptic molecules together.” After a several-hundred-thousand-dollar upgrade to handle two-color experiments, the scope, says Rizzoli, is “quite a solid instrument,” and also “much more versatile.” As a full-featured confocal, it also supports applications such as two-photon microscopy.