Josh P. Roberts has an M.A. in the history and philosophy of science, and he also went through the Ph.D. program in molecular, cellular, developmental biology, and genetics at the University of Minnesota, with dissertation research in ocular immunology.
It can be tough to get a clear microscope image through more than a single layer of cells. All that extra material scatters and absorbs light, a reality that complicates imaging of thick samples such as embryos, nematodes and brain tissue. The problem is even greater when imaging live cells over time—to make a movie, for instance; in that situation, researchers have the added complication of sample photodamage caused by continuous exposure to energetic light.
One solution is two-photon microscopy (TPM). From neurobiology to immunology to embryology and beyond, researchers increasingly are finding that TPM may hold the key to their most challenging imaging problems.
Stack ’em up
In many ways TPM (also called multiphoton microscopy, or MPM) is like confocal microscopy. In most applications, laser light is used to excite fluorophores, with the resulting fluorescence captured by a detector. Images are built up by scanning, point by point, across a field, generating multiple thin, in-focus optical slices of tissue.
Confocal microscopy irradiates the sample throughout the z axis, with a single photon of the correct wavelength needed to excite a fluor. But many of those data are discarded: To create sharp images, a pinhole is used to eliminate scattered or out-of-focus emitted light, enabling the instrument to record a single focal plane at a time.
TPM, on the other hand, “excites only at the focal point” in the first place, eliminating the need to throw away any out-of-focus signal, explains Joseph Wolenski, a research scientist and director of the Confocal Microscope Core Imaging Facility in the Department of Molecular, Cellular and Developmental Biology at Yale University.
TPM does this by requiring two photons of longer wavelength (say 976 nm) and lower energy to impinge on a fluor that is normally excited by a shorter-wavelength (say 488 nm), higher-energy photon. That two-hit event is rare, and it occurs only in the focal plane—hence, the sharper focus.
Wolenski explains that the way a lens focuses light looks a lot like an hourglass. And the only place in the optical path where two photons are likely to coincide is at the waist of the hourglass, “where the cross-sectional photon flux is very high,” he says. “So you have all these billions and billions of photons going through the sample, but they only cause excitation at the beam waist, where the sample is in focus.”
Go deep
Because biological tissue is not uniform, it’s difficult to calculate exactly how much, and in which directions, it will scatter light. Nonetheless, there is an inverse and exponential proportionality between wavelength and light scatter, meaning that shorter wavelength light is far more likely to scatter than light of longer wavelength. Thus, it’s possible to excite fluorophores—and recover the signal—from far deeper in tissue using two low-energy photons as opposed to one of shorter wavelength.
“The unique niche that this method fills is being able to image in living tissue—deep in the tissue,” says Ellen Robey, professor of immunology and pathogenesis at the University of California, Berkeley, who uses the technique to study T-cell development in thymus explants and sometimes in live animals. Whereas confocal is limited to perhaps 50 μm, “with two-photon, you can go much deeper, say 300 μm.” Researchers also can image for longer periods without killing the cells or damaging the tissue, Robey adds. “The only way you can get that kind of information about what cells are doing in their native tissue environment, with a time component, is two-photon.”
In TPM, no significant excitation occurs outside of the focal point—again, because two-hit events are rare—so photobleaching and phototoxicity are limited to when the laser is scanning across that point in that plane. This is in contrast to confocal microscopy, in which “even though you’re throwing away out-of-focus light to get a better confocal image … the laser itself is still irradiating the sample above and below that sample spot,” explains Wolenski.
What are you looking at?
Partially because of the difficulty of labeling deep into tissue, most biologists using TPM employ genetically encoded tags such as green fluorescent protein (GFP), says John Rafter, worldwide application specialist at Bruker, which last year acquired life-science fluorescence-microscopy products supplier Prairie Technologies.
Robey’s lab, for instance, uses transgenic mouse strains in which specific cell populations are fluorescently labeled. Her lab also reintroduces (adoptively transfers) into mice cells such as lymphocytes that have been labeled ex vivo. But labels can do more than just identify particular cell types. For example, genetically encoded calcium indicators can provide information about when the cell is receiving a signal through its antigen receptor, or they can be used as a proxy for neural activity.
TPM also can be used to activate proteins or uncage probes. The burgeoning field of optogenetics is built on using photons to impact genetically altered ion channels in neurons. “You introduce light and basically gate open the cells,” Rafter explains. “It’s very powerful to be able to effect change instead of just observing the system.”
Multiphoton microscopy can even be applied to endogenous fluorescent molecules such as hemoglobin, NAD(P)H and FAD, to examine processes like blood flow and cellular metabolism—sometimes combining three photons to stimulate the more energetic transitions [1]. More specialized, non-turnkey techniques include combining TPM with light-sheet microscopy [2], fluorescence resonance energy transfer (FRET), fluorescence correlation spectroscopy (FCS), fluorescence recovery after photobleaching (FRAP) [3] and harmonic generation.
Cost
TPM is for neither the faint-hearted nor the budget-challenged. It requires a million times more photons to impinge upon a single acceptor molecule within a 10-18 second interval than is necessary in single-photon fluorescence microscopy. The light source of choice is thus an ultrafast, high-powered, mode-locked (pulsed), tunable cavity Ti:sapphire laser, which can cost $180,000 or more. These are class-4 lasers—the highest and most dangerous class—with shielding requirements typically mandating the system be housed in its own room.
A base model, turnkey two-photon microscope will cost an additional $125,000 on top of the laser. “By the time you add components, you’re starting at around $225,000 [for the microscope],” says Rafter. A complete setup could cost "probably ... close to a half-million dollars," Robey estimates. “But if you’re sharing that with multiple labs, that’s not a ridiculous amount of money.”
Nearly every fluor used in confocal microscopy can be used in TPM, with identical emission spectra, and can be multiplexed in the same way. In addition, “you can even excite UV fluors like DAPI and Hoechst, which are very common DNA probes,” Wolenski points out. “So if you have a two-photon laser, you probably don’t need a UV laser—and UV lasers are actually very expensive.”
TPM certainly won’t solve every microscopy woe. But for some questions it offers the best—and perhaps the only—light path to the end of the tunnel.
References
[1] Chen, Y, et al., “Recent advances in two-photon imaging: Technology developments and biomedical applications,” Chin Opt Lett, 11:011703. [Article]
[2] Mahou, P, et al., “Multicolor two-photon light-sheet microscopy,” Nat Methods, 11:600-1, 2014. [PubMed ID: 24874570]
[3] Piston, DW, Fellers, TJ, Davidson, MW, “Fundamentals and applications in multiphoton excitation microscopy,” available online at http://www.microscopyu.com/articles/fluorescence/multiphoton/multiphotonintro.html.