Two-photon excitation (TPE) is a nonlinear optical process first predicted theoretically by Maria Göppert-Mayer in 1931.1 Its application to fluorescence microscopy was pioneered much more recently by Denk, Strickler and Webb.2 In TPE, a fluorophore is excited via near simultaneous absorption of two photons, each having half the energy (twice the wavelength) required for the transition from the ground to the first singlet excited state (Figure 1). The prerequisite for near-simultaneous absorption and the timescale of molecular light absorption (~10–16 seconds) dictates the use of specialized excitation sources; in current instruments, this is typically a mode-locked Ti-Sapphire laser delivering infrared light pulses of femtosecond duration at high repetition rates.3 Two-photon excited fluorescence has a characteristic dependence on the square of the excitation light intensity; doubling the excitation intensity quadruples the fluorescence signal. In contrast, fluorescence derived from conventional one-photon absorption exhibits linear dependence on excitation light intensity.
Figure 1. Excited-state energy diagram showing two-photon excitation (1), followed by nonradiative vibrational relaxation (2) and spontaneous fluorescence photon emission (3). In conventional fluorescence detection systems, excitation is achieved by absorption of a single photon of energy (hυEX); processes (2) and (3) are essentially the same.
There are many practical benefits to using TPE, given the transparency of tissues to infrared excitation light:
However, in addition to requiring specialized (and therefore fairly expensive) excitation sources, TPE produces photodamage and photobleaching effects within the confined excitation volume that are often more acute than those produced by laser scanning confocal microscopy.7–10 The advantages of TPE primarily relate to imaging of live specimens. Accordingly, neuroscience—specifically structural and functional imaging of the nervous system—is the largest field of current applications (Table 1). In addition to providing benefits for fluorescence microscopy, TPE offers advantages in other biophotonic techniques such as fluorescence correlation spectroscopy,11 controlled photoablation,12 photodynamic therapy 13 and activation of “caged” compounds.14 To learn more about the technical foundations and applications of TPE microscopy, researchers should consult the growing collection of available review literature.15–20
Figure 2. An experiment illustrating ordinary (single-photon) excitation of fluorescence and two-photon excitation. The cuvette contains a solution of the dye safranin O, which normally emits yellow light when excited by green light. The upper lens focuses green (543 nm) light from a CW helium–neon laser into the cuvette, producing the expected conical pattern of excitation (fading to the left). The lower lens focuses pulsed infrared (1046 nm) light from a neodymium–YLF laser. In two-photon absorption, the excitation is proportional to the square of the intensity; thus, the emission is confined to a small point focus (see arrow), which can be positioned anywhere in the cuvette by moving the illuminating beam. Image contributed by Brad Amos, Science Photo Library, London.
One-photon and two-photon excitation of a given fluorophore generally result in identical fluorescence emission spectra, as the originating excited state and the photon emission process are the same (Figure 1). However, two-photon excitation spectra differ from their one-photon counterparts to an extent that depends on the molecular orbital symmetry of the fluorophore (greater difference for higher symmetry fluorophores).18 Consequently, most two-photon excitation spectra are blue-shifted and broader compared to the corresponding one-photon spectra plotted on a doubled wavelength axis. Simply stated, a fluorophore with a one-photon excitation peak at 500 nm will probably have a two-photon excitation maximum at <1,000 nm (Figure 3).
Because two-photon excitation spectra are relatively broad, multiplex detection schemes in which two or more fluorophores are excited at a single wavelength and discriminated on the basis of different emission spectra are relatively easy to implement (some examples are included in Table 1). The two-photon absorption cross-section (σ) in units of GM (for Göppert-Mayer; 1 GM = 10–50 cm4 seconds) quantifies the efficiency of TPE for different fluorophores and is plotted on the y-axis of excitation spectra (Figure 3). There are several published collections of two-photon excitation spectra and cross-sections that provide guidance on compatibility of dyes and probes with excitation sources.21–25 Excitation wavelengths used in selected published TPE microscopy applications are listed in Table 1.
Figure 3. Two-photon excitation spectra of Alexa Fluor 350, Alexa Fluor 488, Alexa Fluor 568 and Alexa Fluor 594 dyes. The y-axis units are products of fluorescence quantum yields (ΦF) and two-photon absorption cross-sections (σ). Data courtesy of Warren Zipfel, Cornell University.
TPE has added a new spectral dimension to fluorescence microscopy. Probes such as fura-2 (for Ca2+), SBFI (for Na+), monochlorobimane (for glutathione), and DAPI (for nuclear DNA) were previously of limited utility in confocal microscopy due to their requirements for ultraviolet excitation; now, however, these probes have a new lease on life (Table 1). Furthermore, in situ imaging of small endogenous fluorophores, such as serotonin and NADH, that are almost inaccessible to one-photon excitation has now become practicable.26 Organic fluorophores and fluorescent proteins typically have two-photon absorption cross-sections in the range 1–100 GM. However, fluorophores with smaller cross-sections (e.g., NADH, σ < 0.1 GM) can still generate sufficient TPE fluorescence for imaging purposes.26 At the opposite extreme, Qdot nanocrystals have cross-sections exceeding 10,000 GM, promising even further expansion of TPE imaging, particularly in the area of in vivo applications.27,28
|Alexa Fluor 488 phalloidin||720 nm or 830 nm||Imaging F-actin organization in pancreatic acinar cells||J Biol Chem (2004) 279:37544–37550||A12379|
|Alexa Fluor 594 hydrazide||810 nm||Ca2+-insensitive, neuronal tracer *||Neuron (2002) 33: 439–452; www.stke.org/cgi/content/|
|Amplex Red reagent||750 nm or 800 nm||Detection of reactive oxygen species (ROS) associated with amyloid plaques||J Neurosci (2003) 23:2212–2217||A12222, A22177|
|CFSE, CMTMR||820 nm||Tracking T and B lymphocytes and dendritic cell motility patterns in intact mouse lymph nodes †||Science (2002) 296: 1869–1873; Proc Natl Acad Sci U S A (2004) 101: 998–1003||C1157, C2927|
|CM-H2DCFDA||740 nm||Detection of localized reactive oxygen species release in cardiomyocytes ‡||J Biol Chem (2003) 278: 44735–44744||C6827|
|DAPI, Hoechst 33342||740 nm||Imaging DNA in nuclei and isolated chromosomes||Micron (2001) 32:679–684; Histochem Cell Biol (2000) 114:337–345||D1306, D3571, D21490, H1399, H3570, H21492|
|DiD||817 nm||Intravital imaging of mouse erythrocytes||Proc Natl Acad Sci U S A (2005) 102:16807–16812||D307, D7757|
|FM 1-43||840 nm||Monitoring synaptic vesicle recycling in rat brain slices||Biotechniques (2006) 40:343–349||T3163, T35356|
|Fluo-5F §||810 nm||Imaging Ca2+ concentration dynamics in dendrites and dendritic spines||Neuron (2002) 33:439–452; www.stke.org/cgi/content/|
|Fura-2||780 nm||Detection of GABA-mediated Ca2+ transients in rat cerebellar Purkinje neurons||J Physiol (2001) 536:429–437||F1200, F1201, F1221, F1225, F6799, F14185|
|Lucifer yellow CH||850 nm||Identification of gap junctions in rat brain slices||J Neurosci (2003) 23:9254–9262||L453, L682, L1177|
|Laurdan||800 nm||Detection of ordered membrane lipid domains||Proc Natl Acad Sci U S A (2003) 100:15554–15559; J Cell Biol (2006) 174:725–734||D250|
|Monochlorobimane||780 nm||Imaging glutathione levels in rat brain slices and intact mouse brain||J Biol Chem (2006) 281:17420–17431||M1381MP|
|MQAE||750 nm||Fluorescence lifetime imaging (FLIM) of intracellular Cl– concentrations in olfactory sensory neurons||J Neurosci (2004) 24:7931–7938||E3101|
|Oregon Green 488 BAPTA-1||880 nm||Imaging spatiotemporal relationships of Ca2+ signals among cell populations in rat brain cortex||Proc Natl Acad Sci U S A (2005) 102:14063–14068||O6806, O6807|
|Qdot 525, Qdot 585, Qdot 655 nanocrystals||750 nm||Multiplexed immunohistochemical analysis of arterial walls **||Am J Physiol (2006) 290:R114–R123||Q11441MP, Q10111MP, Q11621MP, Q11421MP|
|SBFI||760 nm||Imaging of intracellular Na+ gradients in rat cardiomyocytes||Biophys J (2004) 87:1360–1368||S1262, S1263, S1264|
|TMRE||740 nm||Mitochondrial membrane potential sensor ‡||J Biol Chem (2003) 278:44735–44744; Circulation (2006) 114:1497–1503||T669|
|X-rhod-1||900 nm||Simultaneous imaging of GFP-PHD translocation and Ca2+ dynamics in cerebellar purkinje cells||J Neurosci (2004) 24:9513–9520||X14210|
|* Used in combination with fluo-4, fluo-5F or fluo-4FF to obtain ratio signals that are insensitive to small changes in resting Ca2+ and are independent of subcellular compartment volume. † Multiplexed (single excitation/dual channel emission) combination of CFSE and CMTMR. § Techniques also applicable to fluo-4 and fluo-4FF indicators. ‡ Multiplexed (single excitation/dual channel emission) combination of TMRE and CM-H2DCFDA. ** Multiplexed (single excitation/dual channel emission) combination of Qdot 585 and Qdot 655 nanocrystals. PHD = pleckstrin homology domain.|
For Research Use Only. Not for use in diagnostic procedures.