Laser-Based Measurements in Cell Biology

report biophysical parameters of ﬂuorescently labeled molecules within living cells. Photobleaching techniques, ﬂuorescence lifetime imaging, and ﬂuorescence correlation methods can measure kinetic rates, molecular di V usion, protein–protein interactions, and concentration of a ﬂuorophore-bound molecule. This chapter provides an introduction to the ﬁeld of laser-based microscopy enabling readers to determine how best to match their research questions to the current suite of techniques. pulsed laser. The in vitro measurements demonstrate the fast residence times of in vitro proteins (di V usion constant of (cid:1) 80 m m 2 /s) versus the intracellular measurements resulting in a e V ective increase of about cytoplasmic viscosity by a factor of four (small complex (cid:1) 21 m m 2 /s, large complex (cid:1) 1 m m 2 /s, relative concentration ratio of 6:1, small to large). The intracellular measurements indicate that the Mad2 protein exists in two major complexes in mitotic cells (one component ﬁt dashed line, two component ﬁt solid line) seen by the poor one component ﬁt at long residence times. Only the small complex is present in nonmitotic cells, indicating a mitosis-speciﬁc assembly process.


I. Introduction
A. In this chapter, we review the imaging techniques and methods of molecular interrogation made possible by integrating laser light sources with microscopy. We discuss the advantages of exciting fluorescence by laser illumination and review commonly used laser-based imaging techniques such as confocal, multiphoton, and total internal reflection microcopy. We also discuss emerging imaging modalities based on intrinsic properties of biological macromolecules such as second harmonic generation imaging and coherent anti-Raman resonance spectroscopy. Super resolution techniques are presented that exceed the theoretical diVraction-limited resolution of a microscope objective. This chapter also focuses on laser-based techniques that can

I. Introduction
The use of laser sources in microscopy long predate their routine use today in modern cellular-imaging applications such as scanning confocal or multiphoton microscopy (MPM). As outlined in the chapters included in this volume, the first uses of lasers were as sources of high energy light for localized photodamage or microsurgery. Many of the same properties that make them excellent tools for spatially localized damage also provide for high-resolution microscopic imaging and biophysical measurement of molecular parameters within living cells. In part, the wide use of lasers in microscopic imaging, particularly in living cells, has been enabled through the revolution in fluorescent dyes and genetically encoded fluorescent proteins that provide molecular specificity (Shaner et al., 2005;Zhang et al., 2002). In this chapter, we discuss the properties of lasers that make them a unique light source for microscopic imaging and measurement. These properties, such as coherence, monochromaticity, and ultrashort pulses have enabled higher spatial resolution for cellular imaging. More conventional methodologies such as confocal or multiphoton imaging will be discussed alongside some of the recent developments in imaging technologies. Many of the characteristics of lasers also provide the ability to measure molecular parameters within living cells such as protein-protein interactions, diVusion constants, and reaction rates. Many of these measurements have long been considered the domain of in vitro systems but now can be interrogated in living cells. The application of lasers to these in vivo biophysical methods will also be discussed.

A. The Laser
The laser, originally based on its microwave counterpart the maser of the 1950s, has revolutionized the field of biological measurement. The name laser is an acronym for Light Amplification by Stimulated Emission of Radiation. In a laser, an active laser medium, also referred to as a gain medium, is placed within a resonant optical cavity (for review see Svelto, 1998). An external energy source such as a flash lamp, arc lamp, or light-emitting diode array pumps the electrons of the gain medium into an excited state. If the electrons were allowed to relax to the ground state, spontaneous emission would occur resulting in emitted photons of narrow spectral bandwidth, but of random phase. Instead, a laser uses the principle of stimulated emission to force the emitted photons into a tightly bound range of phase, frequency, and polarization. Photons traveling back and forth due to reflection from the mirrored walls of the optical cavity bombard the excited atoms of the gain medium. Instead of absorbing the photons, the atoms of the gain media release their potential energy by emitting a second photon similar to the first that travels along the same trajectory. The resonant cavity is designed so that a percentage of light passes through one of the mirrors. The escaped light is the laser beam and is typically collimated, coherent, monochromatic, of constant power, and polarized. This is in stark contrast to light emitted from conventional wide-field illumination sources (e.g., mercury or xenon arc lamps) that are incoherent, of wide spectral content, nonpolarized, and exhibit intensity fluctuations.

B. Coherency and Illumination Volume
Laser light is well suited for fluorescence microscopy because it is spatially and temporally coherent, properties that allow a very tight and bright focus in the specimen plane. Temporal coherency means that the wave is nearly perfectly correlated to itself in time at each point (Svelto, 1998). Spatial coherency means that all points within the wavefront are perfectly cross-correlated, or have similar phase independent of time. Consider, for example, a collimated laser beam incident on the back aperture of a microscope objective lens. The light travels as a coherent plane wave. The microscope objective transforms the plane wave into a converging spherical wave. Because of the coherence of the laser, the resulting focus is nearly diVraction-limited. The laser is ideal in that it minimizes the excitation volume, it provides a ''bright'' excitation field, and it excites fluorescent molecules with a time-invariant intensity (for the case of continuous wave lasers). Furthermore, using the techniques of Q-switching, mode-locking, or gain-switching, laser output can be transformed from continuous wave to pulsed output. Using pulsed lasers much higher peak irradiances can be achieved in the focal spot. The higher peak irradiances allow for nonlinear interactions, such as multiphoton absorption, between the sample and the light as described in sections below.

A. Laser Scanning Confocal Microscopy
The most common imaging technology based on laser illumination is confocal microscopy. In 1957, Marvin Minsky proposed a ''white light'' confocal microscope in which the specimen is scanned by the stage (Inoué, 1995). Later, stage scanning confocal microscopes using laser light sources and nonimaging (photomultiplier tube, PMT) detectors were developed (Wilson, 1980;Wilson et al., 1980) for scanning electronic devices. In the 1980s, laser scanning confocal microscopes (LSCMs) (Carlsson et al., 1985(Carlsson et al., , 1987 came into the forefront in which confocal fluorescent images of biological specimens were acquired by steering the laser across a stationary specimen. It was apparent that the out-of-focus light rejection and optical-sectioning capability of LSCM over conventional image collection permitted the visualization of individual biological structures never before visualized in fluorescence (White et al., 1987). In conventional laser-based epi-fluorescence, a steep cone of laser light converges to focus, and then diverges out of the specimen. As the cone narrows, irradiance increases with maximum intensity at the focus spot. In order to maximize quantum yield of the fluorescent molecules in the focus spot, illumination power is typically well above the fluorescence threshold (Tsien and Waggoner, 1995). As a result, fluorescence is emitted and subsequently collected by the objective lens from the true focal plane, as well as from out-of-focus objects. In LSCM, laser light is focused into the specimen and excites fluorescence, which passes through a dichroic filter and is reimaged onto the detection pinhole which is placed ''confocal'' to the laser focus spot (Carlsson and Aslund, 1987;Carlsson et al., 1985). As a result, fluorescence originating from the laser focus passes through the detector pinhole, while that originating from out-of-focus regions does not. Light passing through the pinhole is detected, and by raster scanning the focus spot through the specimen, an image can be constructed. Although the LSCM is rate limited and may adversely eVect live cells, it has been used to provide unique insights into neurobiology (Barger et al., 1995;Goodman et al., 1996;Nitatori et al., 1995), HIV (Amara et al., 1997), intracellular signaling (Bae et al., 1997;Hakansson et al., 1998), hepatocyte function (Buchler et al., 1996), intracellular transport (Lin et al., 1995), roles of intracellular calcium (Cheng et al., 1993), carcinogenesis (Aragane et al., 1998), and RNA processing (Spector et al., 1991), just to name a few.

B. Multipoint Confocal Laser Microscopy
One major concern with LSCM is that the high irradiances required to achieve high frame rates and suYcient signal-to-noise ratios adversely eVects cell viability. Additionally, because the quantum yield has a finite maximum value, LSCM is inherently rate limited (Gräf et al., 2005). As a solution, multiple confocal pinholes can be implemented and projected into the specimen simultaneously. As the number of pinholes increases, the dwell time per pixel increases proportionally for the same frame rate, and irradiance can be reduced. While attempts have been made to construct and commercialize multipinhole laser microscopes (MPLMs) using technologies such as liquid crystal spatial light modulators (Smith et al., 2000) or micromirror arrays (Fainman et al., 2001;Verveer et al., 1998), the Nipkow spinning disk has become the technology of choice (Conchello and Lichtman, 2005). The Nipkow disk is a spiral array of equally spaced pinholes and was first applied to microscopy by Petrán et al. (1968). An expanded laser beam illuminates the disk, while the tube lens together with microscope objective image the illuminated pinholes into the specimen plane. With each rotation of the disk, the entire field is scanned. The Nipkow disk suVers from poor excitation light eYciency, but has nonetheless been successfully commercialized using an arc lamp white light source (BD Biosciences, California) and has been used in a diverse range of studies including intracellular protein traYcking (Braunagel et al., 2004), nerve myelination (Maria, 2003), and calcium conduction in cerebral endothelia cells (Marrelli et al., 2003). The light eYciency of spinning disk confocal microscopes has been increased to that of LSCM by rotating a second disk containing microlenses coincident with each pinhole of the Nipkow disk. Currently, the CSU10 (confocal scanning unit 10) and the CSU22 from Yokogawa (Yokogawa, Japan) represent the state of the art and scan 20,000 microlenses and pinholes to achieve frame rates of up to 1000 frames per second (Nakano, 2002;Tanaami et al., 2002). For most applications, images are integrated over longer durations and the real benefit of the spinning disk technology over LSCM is cell viability achieved because the necessary irradiance, per pinhole, decreases as the number of pinholes increase for a given frame rate. The Yokogawa spinning disk system is widely used in cell science and has been critical in studies of the cytoskeletal role during cell migration (Adams et al., 2003;Grego et al., 2001;Salmon et al., 2002), the interplay of kinetochores and microtubules during mitosis (Tirnauer et al., 2002), mitotic spindle formation (Ovechkina et al., 2003;Fig. 1), and mitochondrial positioning in yeast (YaVe et al., 2003).

C. Total Inter nal Reflect ion Microscopy
Total internal reflection microscopy (TIRFM) is a method for imaging cell adhesions, cytoskeletal elements, and membrane components at the glass-cell interface. TIRFM induces total internal reflection between the glass and cell culture media interface sending an evanescent wave a few hundred nanometers deep into the sample with enough irradiance to excite fluorescence. The theory of TIRFM is well described by Reichert and Truskey (1990) and by Gingell et al. (1987). T he fir st r epo rt ed TIRFM system for cells (Axelrod, 1981) implemented a cell culture chamber consisting of a glass coverslip sandwich created by inverting a coverslip containing adhered cells above a second coverslip spaced by a thin membrane. A fused quartz cube was placed on the top coverslip coupled by glycerol. The cube serves to further increase the angle (measured from the optical axis of the objective lens) of laser light from a source suYciently juxtaposed to the optical axis so that the angle of light between the top coverslip and the cell culture media is greater than the critical angle for total internal reflection. Fluorescence excited by the evanescent wave is collected by the microscope objective and imaged onto a CCD. It is also common to perform TIRFM using only a high numerical aperture objective lens without the relatively complicated setup of the quartz cube-based system (Stout and Axelrod, 1989). Laser light is focused at the back focal plane of the objective, oV axis, and at an angle of convergence large enough to illuminate a reasonably large field (Axelrod, 2001a). There are several other schemes for creating TIRFM, reviewed by Axelrod (2001b). Because TIRFM only excites a very thin section of fluorescence at the cell adhesion level, it has enabled several key observations regarding the extent and density of focal adhesions (Mathur et al., 2003), guidance of microtubules to focal adhesions (Krylyshkina et al., 2003), heterogeneity of microtubule-based transport in motile cells (Wadsworth, 1999;Yvon and Wadsworth, 2000), and dynamics of secretory granules (Oheim et al., 1998;Rohrbach, 2000).

D. Nonlinear Imaging
The aforementioned imaging technologies rely on linear absorption of light by molecules to yield fluorescence. In this case, the probability of a fluorescent event is linearly dependent on the beam intensity incident on a fluorescent molecule. A number of imaging modalities are gaining acceptance in cell biology that rely on nonlinear absorption in which the probability of an event (e.g., fluorescence) is not simply proportional to laser intensity (Zipfel et al., 2003). The emergence of nonlinear imaging techniques is in no small part due to the increased availability of turnkey pulsed laser light sources. One such technology is MPM that was first reported by Denk et al. (1990). Two-photon absorption was first postulated by Gö ppert-Meyer (1931) in the 1930s and was first demonstrated shortly after the invention of the laser using a relatively long pulsed laser (ruby laser, $1 ms pulse duration) to excite a CaF 2 :Eu 2þ crystal (Kaiser and Garrett, 1961). The principle of two-photon absorption is that two photons of near equal energy can interact with a single molecule as if they were a single photon of twice the energy, if they arrive nearly simultaneously. MPM is an umbrella term that includes not only two-photon absorption, but higher modes of nonlinear absorption as well. In the field of biology, MPM has been demonstrated with continuous wave lasers (Booth and Hell, 1998;Kirsch et al., 1998), but in general ultrashort pulse lasers (100 fs or less) in the near infrared (NIR) are used (Curley et al., 1992). MPM provides the advantages of confocal optical sectioning, without the need of the detection pinhole. MPM excites with NIR light, which has the advantage over visible light that endogenous absorption by the cells/tissues and water is relatively low compared to the visible spectrum (Konig et al., 1995;Liang et al., 1996) allowing deeper penetration into tissue samples, multilayer cell cultures, and living tissue. MPM has been used to image pyramidal cells deep within the neocortex (Helmchen and Denk, 2005;Fig. 2), cancerous cell migration in intact tumors (Condeelis and Segall, 2003) and in tissue models (Wang et al., 2002;Wolf et al., 2003), angiogenesis in the developing brain of zebrafish embryos (Lawson and Weinstein, 2002), dynamic translation rates in living dendrites (Job and Eberwine, 2001), calcium dynamics in living tissues (Denk and Detwiler, 1999;Svoboda et al., 1997), and cell fusions in living Caenorhabditis elegans embryos (Mohler et al., 1998). Single-beam MPM systems suVer the same rate limitations as LSCM. Accordingly, several groups have developed multipoint MPM (MMM) to gain the speed and viability advantages found in spinning disk confocal microscopes (Egner et al., 2002). MMM have been developed based on beam splitting (Nielsen et al., 2001), a microlens array disk (Bewersdorf et al., 1998;Straub and Hell, 1998), and Yokogawa type spinning disk systems (Fujita et al., 1999).
Two nonlinear imaging modalities: second harmonic generation (SHG and third THG) and coherent anti-Stokes Raman scattering (CARS) are imaging methods that act on endogenous molecules, rather than through the introduction of fluorescent reporters. SHG microscopy, based on a 1978 publication first suggesting the imaging modality (Sheppard and Kompfner, 1978), was first applied to biological tissue by  who examined collagen structure in the rat tail tendon. SHG, or frequency doubling, is a nonlinear process requiring high irradiance illumination through a highly polarizable material with a noncentrosymmetric molecular organization (Campagnola and Loew, 2003). SHG emits light at half the wavelength of the excitation illumination so that two NIR photons from a femtosecond pulsed laser typically used in two-photon microscopy would create a single photon in the visible spectrum at twice the energy. Since SHG is dependent on the scale and the form of molecular ordering within the tissue Zipfel et al., 2003) and collagen is highly organized in the tendon, the method yields high contrast signals in that tissue. SHG has been shown to produce optically sectioned images of the plasma membrane (Campagnola et al., 2002;Peleg et al., 1999), tumor development in hamster cheek pouch mucosa (Guo et al., 1999), collagen in RAFT tissue models (Zoumi et al., 2002), and sarcomeres with the C. elegans nematode (Campagnola and Loew, 2003; Fig. 3). CARS is a type of multiphoton ''vibrational imaging'' cranial window with the dura mater removed so that micropipettes for cell labeling and electrophysiological recordings can be inserted (top). Pulsation of the exposed brain is reduced by covering the craniotomy with agar and a coverglass. Thinned skull (20-to 4-mm thickness) preparation (middle). Cellular structures are either prelabeled (e.g., with fluorescent proteins in transgenic mice) or stained through a tiny hole lateral to the thinned area. Chronically implanted glass window replacing the skull (bottom). Agar is used underneath the window for stabilization. (B) Example of deep two-photon imaging in mouse neocortex. Maximum intensity side projection of a fluorescence image stack, obtained in a transgenic mouse expressing Clomeleon, a genetically encoded chloride indicator, under the control of the Thy1-promoter, preferentially in deep layer 5 (L5) pyramidal cells. Data were taken with a 10-W pumped Ti:Sapphire oscillator using a 40, NA 0.8 water immersion lens (Zeiss). Note that nearly the entire depth of the neocortex can be imaged. Reprinted by permission from Macmillan Publishers Ltd., Nature (Helmchen and Denk, 2005), copyright (2006).
based on coherent anti-Stokes Raman scattering in which three electric fields: the pump, Stokes, and probe interact with molecules in a sample to generate the anti-Stokes signal (Cheng and Xie, 2004;Duncan et al., 1982;Tolles et al., 1977). Molecular specificity is achieved from characteristic molecular vibrations of endogenous species. CARS is still an experimental technology and has not yet moved into mainstream imaging. It has been used to image lipids in live cells , to determine the thermodynamic state (liquid crystalline or gel phase) of lipid membranes (Muller and Schins, 2002), to image neutral lipid droplets (LDs) in unstained live fibroblasts (Nan et al., 2003), and to measure real-time intracellular water concentrations in live cells (Potma et al., 2001). CARS has been used to image The sarcomeres in the body wall muscles are seen at the edges of the animal, as well as in a portion of the chewing mechanism. (B) An optical section further into the same animal, where only the chewing mechanism is observed with substantial SHG intensity. Reprinted by permission from Macmillan Publishers Ltd., Nature (Campagnola and Loew, 2003), copyright (2003).

Laser Microscopy in Cell Biology
intact tissues including lipid domains in the mouse ear in vivo (Evans et al., 2005), and lipid phase within axonal myelin in live spinal tissue explants (Wang et al., 2005).

E. Super Resolution Microscopy
Lasers can be used to stretch the resolution of microscopy beyond the diVraction limit of light. For example, TIRFM improves axial resolution beyond the diVraction limit because the evanescent field extent is smaller than the axial extent of the point spread function (PSF) of a focused laser. Using acoustic optical modulators to vary the angle of laser incidence in TIRFM with subsequent Laplace transformbased reconstruction, topographical resolutions of tens of nanometers have been achieved (Loerke et al., 2000).
In standing wave fluorescence microscopy two coherent plane waves, either from two apposing objective lenses, or from a single objective opposing a piezo-controlled mirror, interfere in the specimen plane to yield an axial resolution better than 0.05 mm (Bailey et al., 1993). The 4Pi confocal fluorescence microscope presented by Hell and Stelzer (1992b) uses two apposing 1.4 NA objectives to optically compensate for axial spreading of the PSF to achieve 110 nm axial resolution (vs the diVraction limit of $1 mm) in Nile Blue dye. In type A 4Pi confocal microscopy, both objectives focus a laser beam to a common focal point where constructive interference occurs and the resulting illumination PSF takes on a more spherical geometry thereby reducing axial stretching. In type C 4Pi confocal microscopy, resolution is further enhanced by interference of detected light on a common point detector. The 4Pi design has also been shown to improve two-photon microscopy (Hell and Stelzer, 1992a).

III. Measurement of Cellular State: Molecular Parameters
A central feature in fluorescent imaging is the ability to determine the position and relative levels of a specific molecule. Lasers, as described above, act to provide better spatial resolution and multiplexed detection than conventional light sources due to their coherence and monochromaticity. The ability to provide a high flux of photons to a small diVraction-limited volume can also be used to modulate fluorescent molecules at a specific location or measure the properties of fluorescent molecules. Below we survey a number of biophysical measurements techniques used in living cells that have been enabled or improved by the use of lasers.

General Principle
In this family of methods, a protein (or other cellular component) is labeled with a fluorophore and permitted to equilibrate in the cell. Through the use of a laser, a region of these fluorescent molecules can be excited suYciently so that they are no longer fluorescent, that is undergo photobleaching. These bleached molecules now represent a new species that can be distinguished from the existing fluorescent pool. The kinetics of these molecules can provide a quantitative insight into various physical processes such as diVusion, reaction, and flow. The experimental measurement described above is often referred to as fluorescence recovery (or redistribution) after photobleaching (FRAP).

Microscopy Configurations
Practical photobleaching is accomplished via high-intensity light source, such as a laser, that excites the fluorophore with a suYcient number of photons to make the probability of losing its fluorescence significantly large. Many LSCMs can be used as FRAP setups, by increasing the laser intensity in a small region for photobleaching and then lowering the power for imaging the subsequent recovery dynamics. Other setups also use a laser to achieve a diVraction-limited spot of high illumination that is used for single-point FRAP and imaging is then usually achieved by another light source such as an arc lamp. Scanning confocals can photobleach regions of arbitrary shape and size, via the use of the scanhead, permitting the bleaching of a subregion of a fluorescently labeled organelle such as the nucleus (Daigle et al., 2001). Single-point FRAP configurations can accomplish spatial-patterning FRAP through the use of galvanometers that move the laser or lenses that can change the focused beam into lines or more complex shapes.
Some experimental setups utilize lower intensity sources (e.g., mercury or xenon arc lamps) and arise to bleach molecules but require an extended time period to achieve a detectable bleached pool. These longer exposure times compete with intrinsic diVusive mechanisms that permit mixing of the ''bright'' and ''dark'' species during the photobleaching period and thus the FRAP dynamics become a complex mix of photobleaching and recovery kinetics. To measure dynamic processes with fast timescales (on the order of tens of seconds to many minutes), a laser is essential to provide a high-intensity flux of photons that can rapidly bleach, and therefore mark, the fluorescent molecules of interest.

Measurement Data and Interpretation
In a typical FRAP experiment, the recovery at the point of photobleaching is monitored. This measurement is often made via quantitative imaging but can also be carried out using low-intensity laser excitation to measure local fluorescence. The basic FRAP recovery curve consists of the initial fluorescence level, the bleaching event which reduces the amount of fluorescence and the recovery to a new steady state (Fig. 4). The key parameters in the recovery curve are the rate and extent of recovery. These parameters are obtained by fitting the recovery curve to an exponential recovery equation [Eq. (1)] F ðtÞ ¼ Að1 À e À1=t Þ ð 1Þ where F ðtÞ represents the normalized fluorescence of the region that incorporates the initial region fluorescence, fluctuations of the light source, and bleaching due to observation. Recovery curves often exhibit single exponential recovery kinetics with a single recovery time (t) and extent (A). In more complex kinetics, recovery curve fitting requires modeling the existence of multiple components of recovery [Eq. (2)] F ðtÞ ¼

Laser Microscopy in Cell Biology
where each component is parameterized by its time of recovery (the exponential decay term, t i ) and contribution to extent of recovery (A i ). The extent of recovery is a direct readout of the mobility of the fluorescent moiety in the region. A high extent of recovery indicates a transient interaction at some subcellular scaVold or complete diVusive recovery in a membrane or cytoplasmic compartment. Low extent of recovery indicates a significant immobile fraction (1 À A or P i A) that may turnover on timescales longer than that of the experimental measurement (Fig. 4). The time of recovery indicates the timescale of diVusion, reaction kinetics or flow, whichever is the dominant process at work. The interpretation of the recovery time depends greatly on the underlying biology under study. Two simple model systems are the in vitro fluid membrane (or in vitro solution) and the insoluble scaVold. Within simple in vitro soluble systems, a single exponential recovery rate can be readily converted into a diVusion constant, whereas at insoluble scaVolds, the recovery rate is directly related to the oV-rate (1/t) of the reaction between the fluorescent marker and the scaVold (Axelrod et al., 1976;Braeckmans et al., 2003;Bulinski et al., 2001;Lele and Ingber, 2006;Sprague et al., 2004). These two extremes represent the simplest systems in which we can find direct interpretation of the recovery time. In vivo cellular measurements can be complicated by the complex structure of membranes and the cytoplasm. However, in combination with a model of the underlying biology, FRAP experiments can reveal kinetic phenomena otherwise inaccessible via fluorescence imaging alone and provide a quantitative measurement of these kinetics.

Cellular Applications
Early work using FRAP measured the diVusion of fluorescent reporters within the plasma membrane and laid the groundwork for studies of kinetics and diVusion in many other systems (Axelrod et al., 1976;Jacobson et al., 1976). Since that time, FRAP has been widely employed within many areas of cell biology. Work by Luby-Phelps and Taylor (1988) provided the first measurements of the heterogeneity of diVusion in the cellular cytoplasm using fluorescently labeled dextrans introduce by microinjection. Membranous organelles, such as the nuclear envelope (Daigle et al., 2001), have been studied extensively to dissect the ability of proteins to freely diVuse within the membrane (Axelrod et al., 1976;Jacobson et al., 1976) or membranous compartments (Nehls et al., 2000). FRAP has also found tremendous utility in the study of nuclear and chromosome dynamics, in part due to the stable nature of some protein scaVolds found near or on chromatin such as chromosomes and kinetochores (Gerlich et al., 2006;Howell et al., 2000;Shah et al., 2004), DNA damage machinery (Bekker-Jensen et al., 2005) and nucleoli (Chen and Huang, 2001), and other more transiently bound nuclear components (Misteli et al., 2000). Similarly, studies related to the cytoskeleton (Bulinski et al., 2001;Pearson et al., 2003) and cytoskeletal-associated structures (Khodjakov and Rieder, 1999;Kisurina-Evgenieva et al., 2004;von Wichert et al., 2003) use FRAP as a central methodology to investigate dynamics of protein association and turnover.

Related Methodologies
Modified experimental setups can also monitor the loss in fluorescence that occurs far from a position undergoing constant photobleaching (Delon et al., 2006;Wachsmuth et al., 2003). Through constant photobleaching of fluorescent molecules in one compartment and simultaneously monitoring the loss in fluorescence in another, the dynamic exchange between the compartments can be established. These fluorescence loss in photobleaching or FLIP experiments can provide insight into diVusion and transport that occur between diVerent cellular compartments, for example cytoplasm and nucleus (Belaya et al., 2006;Shimi et al., 2004) or nuclear and subnuclear compartments (Chen and Huang, 2001).
Following the trajectory of bleached molecules by FRAP can also be accomplished by the use of photoactivation of fluorescence or photoconversion. FLAP or fluorescence localization after photoactivation (or photoconversion) follows the emergence of a new fluorescent species that is produced through the action of a laser or high-intensity light source. Fluorescence activation was originally performed by a high-intensity ultraviolet source that ''uncaged'' or cleaved a chemical moiety that prevented fluorescence (Mitchison, 1989). Such methods found widespread utilization in dynamical cell processes such as cell division (Mitchison, 1989) and cell motility (Theriot and Mitchison, 1991), although the ''caged'' fluorophores had to be introduced through cellular microinjection. The development of a photoactivatable green fluorescent protein mutant (PA-GFP) (Patterson and Lippincott-Schwartz, 2002) and the photoswitchable fluorescent protein Dronpa (Habuchi et al., 2005) have enabled genetic tagging of proteins and their subsequent activation within living cells without microinjection. Such proteins have been used in dynamic cellular processes such as cell division (Salic et al., 2004). In addition to the activation of fluorescence, a number of proteins have been developed that dramatically change their fluorescent spectra on exposure to ultraviolet wavelengths. Kaede (Ando et al., 2002), EosFP (Wiedenmann et al., 2004), and KikGR (Tsutsui et al., 2005) exhibit dramatic red-shifted spectral shifts in emission providing a spectrally distinct fluorophore after photoconversion. A simple photoconvertible fluorescent protein tag has also been developed by utilizing the eYcient energy transfer between concatenated cyan and yellow fluorescent proteins. Without perturbation, excitation of the cyan fluorescent protein produces yellow emission, but after bleaching of the yellow fluorescent protein undergoes a dramatic donor-dequenching producing primarily cyan emission (Shimozono et al., 2006). These proteins represent the state-of-the-art in fluorescence dynamics reporters and should enjoy great success in cell biological applications where dynamic measurements are required.

General Principles
While FRAP-based techniques developed from imaging-based modalities, fluorescence spectroscopy techniques have proved to be a driving force for the innovation in microscopy-based biophysical measurements in living cells (Lakowicz, 1999). One such example is fluorescence lifetime imaging (FLIM). The lifetime of a fluorophore is a result of stochastic nature of spontaneous emission of photons from a population of excited fluorophores. The excited fluorophores, when excited nearly simultaneously, do not synchronously emit singlet transition photons, but instead have a characteristic average timescale of transition. Rapid time-resolved measurements after short-pulsed excitation of an ideal fluorophore reveal a single exponential decay in emission intensity. This decay constant is the lifetime of the fluorophore. The lifetime of a fluorophore can be exquisitely sensitive to its surroundings providing a unique measure of local state. Moreover, the measurements of anisotropy (another form of lifetime measurement) as a result of polarized excitation and emission can provide quantitative measurements of rotational diVusion. FLIM techniques are still actively being developed as is the analysis, but its utility in the realm of energy transfer [Fluorescent (or Fö rster) resonant energy transfer, FRET] measurements, has already provided significant momentum in its use in cell biology.

Microscopy Configurations
As with cuvette-based fluorescence spectroscopy measurements, laser-based microscopy provides two methods for the measurement of fluorescence lifetimes. Frequency-domain measurements drive fluorophores with a signal rapidly modulated intensity and detect the modulated response at the detector. Driving the fluorophores with a modulated intensity with varying periods near the fluorescence lifetime results in an emission signal shifted in phase and modulated in amplitude due to the lifetime decay (Harpur et al., 2001). Frequency-domain methods can be implemented on LSCMs, but equivalently can also be done in wide-field microscopy by illuminating the entire microscope field with a modulated excitation. The ability to image by widefield provides a rapid method of measuring lifetimes on a pixel-by-pixel basis, at a cost of requiring expensive cameras for sensitive detection. The frequency-domain methods, however, can require complex excitation protocols and calculations in the presence of multiple lifetimes (Bastiaens and Squire, 1999).
To accurately resolve multiple lifetimes, time-domain measurements use a pulsed laser to rapidly excite the fluorophores in a small region of the microscope field and detect the emission profile in time. The time-resolved measurements of the fluorescence decay can be evaluated at each scanned position of the laser and the fluorescence lifetime image is thus built pixel-by-pixel. Such time-resolved measurements can also be made on an LSCM equipped with high-time resolution detectors. Here the confocal laser source would not be a continuous wave laser, but a pulsed source of femtosecond to picosecond duration. Many multiphoton scanning microscopes already have much of the hardware for fluorescence lifetime imaging through the use of the pulsed infrared source and nonimaging detectors (Becker et al., 2004).

Measurement Data and Interpretation
Frequency-domain measurements use the change in amplitude and phase that occurs as the input signal passes through the fluorescent samples. The calculations result in two lifetimes, one based on the change in amplitude and one based on the change in phase (outlined in French et al., 1998;Lakowicz, 1999). For modulation input periods close to the single fluorescence lifetime in the sample, these values will be similar, but when multiple species are present or if single fluorescent species have multiple lifetimes, these values diVer and computation of the lifetimes requires more involved calculations (Harpur et al., 2001). Time-resolved measurements involve fitting the exponential decay captured after the pulsed excitation of the sample. This decay can be directly fit with single or multiple exponentials to obtain the fluorescence lifetimes present in the sample (Lakowicz, 1999).

Cellular Applications
Lifetime measurements have enjoyed great success in the area of energy transfer measurements in living cells. FRET is a phenomenon by which excitation of a donor fluorophore results in the emission of acceptor fluorophore through a nonradiative mechanism. The length scale for this transfer must be in the range of 2-7 nm and decays with the sixth power of the increasing distance (Bastiaens and Squire, 1999). This strong distance dependence provides exquisite sensitivity for detecting proteinprotein (or other molecular) interactions within living systems. Intensity-based measurements of FRET can be problematic due to the relative concentrations of acceptor and donor fluorophores as well as cross-talk between fluorescent spectra (Jares-Erijman and Jovin, 2003), particularly when using fluorescent protein fusions (Pelet et al., 2006). One interesting result of energy transfer is the reduction in fluorescence lifetime of the donor fluorophore when undergoing energy transfer. Using a variety of measurement modalities, FRET detection via fluorescence lifetimes have become increasingly popular in cell biology applications (Caudron et al., 2005;Delbarre et al., 2006;Kalab et al., 2006;Peter et al., 2005;Ramdya et al., 2003;Fig. 5). Moreover, the possibility of multiphoton excitation provides for measurements made deep within tissues to detect protein-protein interactions within living organisms (Chen and Periasamy, 2004).
Lifetime imaging has also been applied to measurements of protein mobility through the use of polarization/anisotropy methods (Clayton et al., 2002;Dix and Verkman, 1990). These techniques provide measures of rotational mobility and are very sensitive to changes in protein size. Rotational correlation times of proteins are generally in the area of tens of nanoseconds. Unfortunately, GFP mutants with lifetimes in the range of $1-3 ns (Tsien, 1998;Volkmer et al., 2000) cannot provide the temporal resolution for such measurements. As a result, live cell anisotropy measurements for molecular mobility studies require nongenetically encoded fluorophores with long fluorescent lifetimes that need to be covalently coupled to the molecule of interest and directly introduced into the cellular environment. It should be noted, however, that GFP mutants do exhibit significant diVerences in fluorescence lifetimes, a property that can be exploited for imaging fluorescent proteins with large spectral overlap in emission (Harpur et al., 2001).

General Principles
Fluorescence correlation spectroscopy (FCS) has become an increasingly popular tool in cell biology due to its ability to discern changes in biological complex size, concentration, and composition in living cells (Medina and Schwille, 2002).    (Kalab et al., 2006(Kalab et al., ), copyright (2006.

Laser Microscopy in Cell Biology
FCS techniques use a laser to produce a small excitation volume ($0.1 to tens of femtoliters) from which the fluctuations in fluorescence are measured at high time resolution (approximately tens of nanoseconds). The origin of the fluctuations can be a result of photophysical processes, reaction kinetics, diVusion, and flow (Magde et al., 1972). Photophysical processes and reaction kinetics have timescales shorter than the average residence time of the fluorophore within the excitation volume, whereas fluctuations due to diVusion and flow are a result of fluorescent molecules leaving and entering the excitation volume. The time average of these fluctuations can also provide the absolute concentrations of fluorescent molecules. The ability to measure such a wide variety of biophysical processes and make absolute measurements of concentration has fueled interest in the use of FCS-based techniques in cell biology.

Microscopy Configurations
FCS measurements can be made using a variety of microscope configurations. Basic requirements are a laser to produce the small excitation volume, a high-sensitivity or high time resolution detector [e.g., avalanche photodiode (APD) or PMT], and hardware to convert the measured detection signal into a digital form (Schwille et al., 1999).
Confocal excitation with a continuous wave laser requires a pinhole at the detector to exclude out-of-focus excitation, similar to those already present in laser scanning confocals. Multiphoton excitation does not require the pinhole since only the volume of interest is excited making detection hardware simpler, at the expense of more complex and expensive ultrafast lasers (e.g., Ti:Sapphire femtosecond sources) (Berland et al., 1995;Wang et al., 2006).
Detection hardware can vary depending on the application. High-sensitivity applications, such as measuring low concentrations of fluorescent molecules ($10 pM), require the use of APDs, whereas less sensitive applications can use PMTs. High-time resolution are features of both detectors; however, many singlecolor FCS configurations use two detectors at half the intensity and crosscorrelation, to reduce noise and after-pulsing that may be present in the detectors. This can greatly increase the eVective time resolution of the instrument.
Correlation measurements to determine molecular photophysics or mobility can be accomplished online through fast hardware autocorrelators. Alternatively, photon-counting cards with high bandwidth can be used to record fluctuation time series directly and correlation can be done oZine. Direct fluctuation recordings can be used in related fluctuation spectroscopy methods such as photon-counting histogram analysis (Chen et al., 1999).

Measurement Data and Interpretation
A central feature of FCS measurements is the underlying model of fluorescence fluctuations. Within living cells dynamics of reaction kinetics are relatively slow, particularly when compared to average residence times within the subfemtoliter excitation volume. As a result, most fluctuations are due to the entry and exit of fluorescent species permitting the measurement of a diVusion constant (or equivalently the size of a complex) and concentration. The diVusion constant of the complex of interest can be determined both in the 3D space of the cytoplasm or 2D plane of the membrane. In both cases, the diVusion constant can be derived from fits to the autocorrelation function of the fluorescence fluctuations (Schwille et al., 1999). The equations to fit the autocorrelation decay curve, a measure of the residence time of the fluorescent molecules also contain the average number of molecules contained within the volume. Thus, the fits to the decay curve, even for a single fluorescent species, require two free parameters (Fig. 6). Another parameter derived from the microscope configuration is the excitation volume. This volume can be calibrated by using molecules with a known diVusion constant and at known concentrations and is a sensitive parameter in the fitting of the autocorrelation function. The ability to evaluate absolute concentrations within the cellular environment can be challenging, but relative concentrations can be evaluated with high confidence to follow the dynamic changes in complex formation and dissociation within PtK2 cells stably expressing ECFP fused to Mad2 (right curves and points). Excitation was accomplished with the 800-nm line of a Ti:Sapphire femtosecond pulsed laser. The in vitro measurements demonstrate the fast residence times of in vitro proteins (diVusion constant of $80 mm 2 /s) versus the intracellular measurements resulting in a eVective increase of about cytoplasmic viscosity by a factor of four (small complex $21 mm 2 /s, large complex $1 mm 2 /s, relative concentration ratio of 6:1, small to large). The intracellular measurements indicate that the Mad2 protein exists in two major complexes in mitotic cells (one component fit dashed line, two component fit solid line) seen by the poor one component fit at long residence times. Only the small complex is present in nonmitotic cells, indicating a mitosis-specific assembly process. (Wang et al., 2006). In addition, a number of elements such as excitation power, shape of the excitation volume as well as the geometry and obstacles within the cell must all be considered in evaluating the parameters that result from the fit of the autocorrelation function (Berland et al., 2003;Hess et al., 2002;Nagy et al., 2005). With these factors taken into account carefully, FCS can be a unique tool for measurements of cellular biochemistry.

Cellular Applications
FCS measurements in living cells are becoming increasingly popular, particularly with the introduction of commercial systems available as additions to scanning confocal microscopes. There has been a great deal of detailed measurement of GFP and its variants to understand the nature of the fluorophore both in vitro (Chen et al .,20 02 ) in the cellu lar environm ent ( Schwille et al ., 1999;W ang et al ., 2004 ;Fig. 6). Measurements of intracellular protein dynamics have made important insights into retroviral assembly (Larson et al., 2003), intracellular motility (Kohler et al., 2000), cytoplasmic structure (Weiss et al., 2003), and cell cycle control (Wang et al., 2006). On-going developments in fluorescent proteins and fluorophores (Kogure et al., 2006) and data analysis methods (Chen et al., 1999Muller, 2004) promise to provide new insights into cellular dynamics.

Related Methodologies
FCS methodologies are being actively developed and the list of related methodologies is changing rapidly (Breusegem et al., 2006). One technique that is particularly well suited to cellular studies is fluorescence cross-correlation spectroscopy (FCCS). Here two spectrally distinct fluorescent species are monitored in the cytoplasm or within the cellular membrane. Autocorrelation analysis of each color separately provides the concentration and diVusion constant parameters previously described. However, cross-correlation between the channels provides concentration and diVusion parameters on complexes containing both fluorophores. Much like FRET, the result is the ability to monitor protein-protein interactions but without the requirement for close apposition of the fluorophores. FCCS-based protein-protein interactions are detected on the basis of the correlated motions into and out of the excitation volume (Bacia et al., 2006). FCCS measurements provide the composition of the complexes, an element not available through single color FCS. Recent work has used FCCS to monitor receptorligand interactions (Larson et al., 2005), calcium-signaling (Kim et al., 2005), and in vivo protease activity (Kogure et al., 2006).
Many FCS/FCCS instruments are configured for single-point measurements making spatial diVusion maps and spatial cross-correlation measurements impossible. Developments have incorporated position information into FCS measurements providing spatial and temporal correlation profiles (Digman et al., 2005;Sisan et al., 2006;Skinner et al., 2005;Ries and Schwille, 2006). Cross-correlation in time and space will provide novel methods of analyzing cellular complex dynamics and reaction kinetics all within the living cell.
New developments in FCS and FCCS through novel fluorescent proteins that permit single color excitation dual-color emission FCCS or optical configurations for interleaved dual-color excitation (Thews et al., 2005) as well as position scanning are moving the field toward simpler instruments to extract molecular dynamics and protein-protein interactions as they occur in living cells.

IV. Summary
We have presented an overview of the leading and emerging technologies in laserbased imaging and molecular state measurement. While several of these techniques had been proposed in the years preceding laser development, the widespread commercialization and distribution of continuous wave and pulsed laser systems has brought molecular characterization into the biology research laboratory. The transformation of laser systems from the intimidating research platform architecture into the ''friendly,'' closed system ''turnkey'' architecture is a clear indication that lasers are becoming ubiquitous tools in the biological laboratory. The availability of these turnkey laser systems has enabled the cell biologist to bring a host of techniques that permit imaging and molecular measurements to the study of the biochemical basis of cellular function.