When Cornell University physicist Chris Xu first began working with multiphoton fluorescence microscopy in the 1990s as a graduate student at Cornell, the technique was brand-new—recently developed by scientists Winfried Denk, Watt Webb, and Jim Strickler. Although the potential of using three photons was known and explored early on, it was inefficient because of the laser technology available at the time. As a result, two-photon fluorescence microscopy became the standard.
Multiphoton fluorescence microscopy (MFM) creates high-resolution images with incredible selectivity, imaging only what is labeled with a fluorescent dye. MFM has the advantage of being able to image more deeply into tissue than confocal fluorescence microscopy. It's a technique that is especially popular with neurobiologists because it allows them to image individual synapses and can be performed on brain slices or in live, intact animals.
"If you really want to see what the brain does, you have to be at cell-level resolution," says Kevin Eliceiri, an advanced light microscopy expert and director of the Laboratory for Optical and Computational Instrumentation at the University of Wisconsin, Madison. "Multiphoton microscopy is the only technique where you can get some depth with resolution and see cells actually moving."
To perform MFM, researchers begin by introducing fluorophores, chemical compounds that emit light upon excitation, into the specimen as a dye. Then a focused laser beam scanned in a raster pattern sends long-wavelength photons that excite an electron in the fluorophore into a higher energy state when multiple photons are simultaneously absorbed: two photons for two-photon microscropy or three photons for three-photon microscropy. As the electron decays, it emits a fluorescence signal which is captured by the microscope. In MFM, the excitation wavelength is longer than the emission wavelength—the opposite of traditional fluorescence microscopy.
But as exciting as the technique is, it has had its limitations—specifically, how deep it can go due to scattering and refractive index. If researchers used longer wavelengths of photons, they could image more deeply because the photons would be less susceptible to scattering. But when the wavelength approaches the 1300-nm range, the water in a specimen begins to absorb the photons and the photons are rendered useless. So the question of whether one could image more deeply by using three photons instead of two had gone largely unanswered—until now.
Xu’s work has shown that although there is this water absorption window, three photons of fluorescence is strong enough so that in many cases you can still see clearly. What's more, longer wavelengths don't seem to damage cells as previously thought.
New developments in laser technology have been key to making three-photon microscopy efficient and workable. After studying multiphoton imaging as a graduate student, Xu worked in fiber optics in the telecom industry, providing him with valuable experience in laser technology. "We created our own laser system," says Xu. "With this system we don't need to use a lot of power, only 20 milliwatts, to get superb image quality. This is a lot lower power than the two-photon laser."
The laser Xu created has a high pulse energy and a low pulse repetition rate, making it more efficient for three photons than the standard two-photon titanium sapphire laser. He started with a fiber laser that generated a wavelength of 1550 nm, a standard wavelength used in the telecom industry. Then he used a photonic crystal rod to shift the laser's wavelength to 1700 nm, while scaling up the energy by about 100,000 times by forming a wave called a soliton, which keeps its shape when it travels at a constant rate.
The rod is glass with a special structure around it so that when the laser beams into it, it will maintain its spatial profile. "I knew that the bigger the rod, the more fiber, the more energy. So I wondered, how big can we make it? How far can we push in this direction?" recalls Xu. So he went to discuss his idea with coauthor Frank Wise, also a physicist at Cornell University. “‘I have this wonderful idea,’” Xu told Wise. “I need a big, long rod to form a soliton, so I can make this energy very useful.' [Wise] said, 'Yeah, I have a rod like that,' and literally the next day we just got it, put out a pulse, and yes, the energy was up."
Aside from Xu, others are also revisiting three-photon fluorescence microscopy. In a paper published in the Journal of Microscopy in February 2012, Gail McConnell, a physicist at the University of Strathclyde in Glasgow, UK, and colleagues reported that their three-photon microscope can produce images safely and effectively as well (2).
Like Xu, McConnell developed a laser for her three-photon microscope—in this case, a bi-directional pumped optical parametric oscillator. While it’s a fiber laser that emits at a wavelength of 1500 nm like Wu’s laser, McConnell’s has an additional feature: it’s tunable. So she can adjust the wavelength, giving her some flexibility.
Although her group has done some unpublished research looking at mammalian cells, their published paper imaged plant cells, showing that the cells were not damaged by the increased power. "We could image for hours," she says, "compared to seeing damage after about five minutes with a titanium sapphire laser." The next step, says McConnell, is to improve the resolution still further—something her team is actively working toward.
Three, Four, or More
If three photons work great, why not four? Well, with four photons, you need more photons and brighter lasers, which results in a greater potential for cell damage. Considering the technique is typically used to image live animals, that becomes problematic. "There may be circumstances that allow you to do that,” Xu says. “You would have to redo the optimization. And we haven't reached the depth limit with three photons yet."
Overall, a lot of potential has yet to be explored in the development and optimization of three-photon microscopy, according to Xu. "We can develop better contrast agents, better optics, and perhaps use a little less power to get the same signal, as well as optimize the system to collect the photons. We may be able to do an order of magnitude better on each of these things, which is a factor of 1,000 when added together—not a trivial amount." For now, three photons seems to strike the best balance between absorption and scattering.
Excitement about both Xu and McConnell's work is palpable. "Both have really shown the great potential of three-photon microscopy," says Eliceiri. "This work points out the importance of looking at other wavelengths, and doing a survey of what's there. But this shows the real promise of going very, very far red." With this new tool able to produce high-resolution, three-dimensional images of individual neurons deep in the brain, scientists will hopefully be able to learn much more about how the brain works.
1. Horton, N. G., K. Wang, D. Kobat, C. G. Clark, F. W. Wise, C. B. Schaffer, and C. Xu. 2013. In vivo three-photon microscopy of subcortical structures within an intact mouse brain. Nat Photon advance online publication (January).
2. Norris, G., R. Amor, J. Dempster, W. B. Amos, and G. McConnell. 2012. A promising new wavelength region for three-photon fluorescence microscopy of live cells. Journal of microscopy 246(3):266-273.