Two-photon Absorption Microscopy
Over the last decade, nonlinear optical microscopies have evolved to become as important as linear optical microscopies. The most developed variety (in the imaging community) is two-photon fluorescence (TPF) microscopy. Molecules can absorb two photons in the near-IR (e.g. at 800 nm) and access states at energies that would require UV photons in a single photon absorption event. High spatial resolution is made possible by the quadratic dependence on the laser intensity; the fluorescence signal is generated in the laser's focal volume, which can be as small as a femtoliter. Raster-scanning the focal spot in specimens allows for 3D image formation. Ultrashort (<1 picosecond) laser pulses are often used in the experiments to achieve high peak power with relatively low tissue damage from average power dissipation. The intrinsic contrast of TPF comes from two-photon induced autofluorescence which originates from naturally endogenous fluorophores and protein structures such as NADH, flavins, elastin, collagen, porphyrins and melanin. However, TPF microscopy has some obvious limitations. Only species that fluoresce are visible and the fluorescence has to be strong enough to get out of the sample before it gets absorbed by other molecules. The most common applications of multiphoton imaging in biology (e.g. neuroimaging) use fluorescent genetically encoded reporters, such as green fluorescent protein (GFP). Endogenous contrast is more difficult to achieve, since only species that fluoresce are visible. Thus, for example, while NADH can be seen, the oxidized version NAD+ cannot, so flavoprotein is used as a surrogate. Fluorescence reabsorption is a particularly challenging problem for pigmented lesions, and as longer wavelengths are used to minimize scattering and absorption of pigments in imaging, less fluorescence can be found from endogenous fluorophores. Every molecule may have two-photon allowed electronic transitions, but many molecules do not fluoresce from their electronic excited states (due, for example, to internal conversion, or other nonradiative relaxation effects). TPA would provide more general contrast than TPF, because many important biomarkers exhibit this effect with long-wavelength excitation. The sensitive TPA measurement methods developed in our lab allow us to use the TPA of endogenous chromophores as a means of providing intrinsic contrast.
The two-photon absorption microscope (TPAM) in our lab was built on a home-built two-photon fluorescence laser scanning microscope (TPFM) by adding TPA measurement capabilities. Amplitude modulated laser pulses were sent into the microscope. As shown in figure 1, this microscope acquires TPA images using transmitted light while TPF images are acquired at the same time in epi-mode. TPA signals are detected by either the loss modulation method or the two-color pump-probe method (recent progress has demonstrated that we could also acquire TPA images in epi-mode too). A lock-in amplifier is used to extract TPA signals and limits the image acquisition speed. The typical scanning speeds for our TPA imaging setup is about 100 ms/line.

As a main pigment in skin tissue, melanin plays an important role in the photo-protection of skin from UV radiation. However, melanocytes may become altered or deranged due to environmental factors; for example, sun exposure may cause DNA damage to melanocytes and induce a melanoma. Imaging pigmentation changes may provide invaluable information in discovering these malignant transformations in their early stages, and in turn improve the prognosis of patients. We have demonstrated that two-photon absorption microscopy can provide remarkable contrast for imaging melanin in skin tissue by either measuring two-photon absorption (TPA) at one wavelength or excited state absorption (ESA) at two different wavelengths.
Blood vessel structure and hemodynamics on the microscopic level provide important information in a variety of biomedical applications, for example angiogenesis in tumors and cerebral oxygen delivery in brain function. Established imaging technologies such as MRI, CT, PET, and ultrasound can measure features of blood vessels including blood flow, blood volume, and vessel permeability, but most of these methods involve external contrast agents, and they often lack the resolution to resolve individual capillaries. It is also hard to measure tissue oxygenation directly with these methods. Among optical methods, NIR absorption and diffusing light reflectance can measure total hemoglobin concentration and the oxygenation level but the resolution is usually limited to a few mm due to the high scattering of light by tissue. Techniques such as confocal and two-photon fluorescence microscopy are partially resistant to scattering with inherent 3D sectioning capabilities, but hemoglobin is a non-fluorescent molecule and therefore cannot be imaged directly. An indirect way of imaging microvasculature in deep tissue is to inject an exogenous fluorescent dye into the blood. Other multiphoton microscopy methods such as CARS and THG also show promising results in hemoglobin imaging, but they have not been applied to deep tissue imaging yet. We have recently developed a highly sensitive two-photon absorption (TPA) measurement method that has the same advantage as two-photon fluorescence and can be easily adapted to imaging applications, but does not require fluorescence detection. Recently we have demonstrated sensitive two-color TPA imaging of melanin with micrometer resolution. We have also applied the same technique to hemoglobin imaging; both oxy-hemoglobin and deoxy-hemoglobin can have sequential TPA (or excited state absorption). We have demonstrated this in the ex vivo and in vivo imaging of blood vessels in mouse ears. It is also possible to differentiate oxy- and deoxy-hemoglobin through their different excited state dynamics by employing different pump and probe wavelength combinations. This is an active goal of our current research.






