Monitoring and quantifying dynamic physiological processes in live neurons using fluorescence recovery after photobleaching

Authors

  • Kevin Staras,

    1. School of Life Sciences, University of Sussex, Brighton, UK
    Search for more papers by this author
  • Dan Mikulincer,

    1. Department of Physiology and Cell Biology, Faculty of Health Sciences, Zlotowski Center for Neuroscience, Ben-Gurion University of the Negev, Beer-Sheva, Israel
    Search for more papers by this author
  • Daniel Gitler

    Corresponding author
    1. Department of Physiology and Cell Biology, Faculty of Health Sciences, Zlotowski Center for Neuroscience, Ben-Gurion University of the Negev, Beer-Sheva, Israel
    • School of Life Sciences, University of Sussex, Brighton, UK
    Search for more papers by this author

Address correspondence and reprint requests to Daniel Gitler, Department of Physiology and Cell Biology, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel. E-mail: gitler@bgu.ac.il

Abstract

The direct visualization of subcellular dynamic processes is often hampered by limitations in the resolving power achievable with conventional microscopy techniques. Fluorescence recovery after photobleaching has emerged as a highly informative approach to address this challenge, permitting the quantitative measurement of the movement of small organelles and proteins in living functioning cells, and offering detailed insights into fundamental cellular phenomena of physiological importance. In recent years, its implementation has benefited from the increasing availability of confocal microscopy systems and of powerful labeling techniques based on genetically encoded fluorescent proteins or other chemical markers. In this review, we present fluorescence recovery after photobleaching and related techniques in the context of contemporary neurobiological research and discuss quantitative and semi-quantitative approaches to their interpretation.

Abbreviations used
FRAP

fluorescence recovery after photobleaching

Movement is typically gauged by observing the displacement of individual elements. However, this approach becomes problematic when the objects are either too small or too intermingled to be resolved, as is true for most molecules of biological interest. In this case, a biological sample with a large amount of macromolecular movement would still appear as a motionless continuum. One informative strategy to reveal mobility is to create a disturbance in the sample and monitor the subsequent dynamics. For example, a drop of colorless water added to the center of an unstirred dye solution reveals the intrinsic mobility of the dye molecules over time as they invade the colorless region. Fluorescence recovery after photobleaching (FRAP), a methodology used to measure the mobility of fluorescently labeled molecules, proteins or small organelles (Reits and Neefjes 2001), draws on these same basic principles. A subregion of a biological sample containing mobile fluorescent elements is rapidly and irreversibly photobleached by targeted irradiation with intense excitation light. This produces a transient distinction between bleached molecules within the target region and intact fluorescent molecules outside it. Subsequently, fluorescent moieties move back into the bleached region (influx) and bleached moieties move out (efflux). This can be directly observed as macroscopic recovery of fluorescence within the photobleached area (Fig. 1a–c). Quantitative or semi-quantitative measurement of the kinetics and extent of recovery reveals the nature of the mobility of the moiety of interest. FRAP encompasses a group of related methodologies that differ in the time scale of the examined phenomena, the initial distribution of the moiety of interest, and the quantitative accuracy of the answer being sought. A complementary and inverse approach involves the use of photoactivatable (Patterson and Lippincott-Schwartz 2002; Tsuriel et al. 2006) or photoswitchable fluorescent proteins (Chudakov et al. 2007), or of caged fluorescent compounds (Santamaria et al. 2006). In these, fluorescent moieties are formed by focal irradiation and their dispersal is monitored and quantitatively analyzed.

Figure 1.

Outline of a Fluorescence Recovery After Photobleaching experiment. (a) The cytoplasm of a cell is filled with a mobile fluorescent dye (green). (b) Enlargement of the area demarcated by a dashed square in (a) reveals that the dye is composed of mobile molecules moving randomly throughout the imaged area (left, green circles). The molecules within the dashed circle (corresponding to black spot in a) are bleached by intense light (center, black circles). Recovery (right) ensues as a result of the continuous movement of fluorescent molecules into the spot, and of bleached molecules out of it. Depicted here is an intermediate and not yet complete state of recovery. (c) Graphical representation of fluorescence intensity within the bleaching spot. The sudden drop in intensity refers to the rapid bleaching of the fluorophores; it is followed by recovery. The rate of recovery depends on the instantaneous rate of movement of the molecules, as well as on the size and shape of the bleached area. A black diamond marks the timing of the right frame in (b). If the molecules are freely mobile, the intensity recorded after recovery (the asymptote) reflects the remaining fluorescence within the whole cell. If the total volume is very large in comparison to the bleaching spot, the overall loss of fluorescence is negligible. (d) Depiction of the recovery phase when a subset of the molecules is immobilized (red outline). (e) In the case an immobile subset exists, the asymptote of the recovery curve does not approach the pre-bleach fluorescence, but rather reflects just the mobile fraction.

FRAP has been used extensively in cell biological research, for example to study the mobility of cytosolic and membrane proteins (Peters et al. 1974; Wang et al. 2008; Berkovich et al. 2011), to examine the structure of membranes (Teissie et al. 1978), to probe the structure and directionality of the cytoskeleton (Iliev and Wouters 2007; Fridman et al. 2009), to probe transcription of mRNA and its processing (Aizer et al. 2008), and to view the movement of various organelles (Mitra and Lippincott-Schwartz 2010; Nadezhdina et al. 2010). Likewise, FRAP and homologous techniques have been widely exploited in neurobiological research, for example to examine mobility of neurotransmitter receptors or transporters (Axelrod et al. 1976; Eriksen et al. 2009; Jaskolski and Henley 2009), polarity of the axonal skeleton (Edson et al. 1993; Gauthier-Kemper et al. 2012; Koskinen et al. 2012), diffusional properties of calcium buffers (Schmidt et al. 2003, 2005, 2007), axonal transport and mobility of vesicles (Gaffield et al. 2006; Kamin et al. 2010; Staras et al. 2010; Fornasiero et al. 2012; Orenbuch et al. 2012; Scott and Roy 2012), and the stability of synaptic structure (Tsuriel et al. 2006, 2009).

This review presents an outline of FRAP approaches and similar related techniques, with specific reference to neurobiological applications. It includes a general description of instrumentation, details on the assumptions made in FRAP experiments, approaches to their analysis, and an explanation of the type of information that can be readily deduced.

The FRAP experiment

The initial step in a FRAP experiment is the introduction of a dye, or the fluorescent labeling of the structure/protein of interest in live cells (Fig. 1a). Exogenous probes can be micro-injected into cells as free dye or pre-conjugated to a protein of interest (Schmidt et al. 2005, 2007; Santamaria et al. 2006). Alternatively, they may be introduced as acetoxymethylester dyes (Blatter and Wier 1990) or through physiological processes. An example of the latter is the use of styryl dyes, a family of amphipathic proteins used extensively for labeling recycling synaptic vesicles (Betz and Bewick 1992; Gaffield et al. 2006). Genetically encoded fluorescent proteins [such as the Enhanced Green Fluorescent Protein (EGFP)] are widely used for FRAP experiments, although due consideration should be given to the fact that their large size means they could affect the mobility of the target, and are thus less applicable to the direct determination of protein diffusion. Finally, it is important to select a fluorophore which can be readily photobleached by strong illumination yet is sufficiently resistant to bleaching by the weaker illumination required to subsequently monitor its distribution.

The FRAP experiment itself is divided into three separate phases (Fig. 1b and c): (i) the pre-bleach phase, (ii) bleaching, and (iii) post-bleach recovery. The pre-bleach imaging phase is intended to provide readout of the initial fluorescence within the observation volume, and to establish that the biological and experimental systems are stable. The bleach phase uses intense excitation light to bring about localized photobleaching within a defined and usually small compartment. In the post-bleach phase, fluorescence recovery within the bleached region is monitored. While the kinetics of recovery provides a measure of the instantaneous speed of the mobile moieties that can help to establish the type of movement executed by the fluorescent moiety, the extent of recovery also provides information about the presence of immobile moieties (Fig. 1d–e). More detailed information concerning the analysis of FRAP experiments is provided below.

Assumptions and requisites

To quantitatively analyze FRAP experiments the following conditions should be fulfilled:

  1. For quantitative analysis, the precise shape of the bleaching spot should be known. The simpler the geometry of the spot and its boundaries, the simpler the formula that describes the recovery phase (Soumpasis 1983; Brown et al. 1999; Tsibidis 2009).
  2. Bleaching should be irreversible, but does not need to be complete [i.e., not all fluorescence in the spot needs to be bleached (Brown et al. 1999)]. See however (Sinnecker et al. 2005).
  3. The bleaching process should not in itself influence the mobility of the moiety of interest and should not harm the biological processes within the sample (Bancaud et al. 2010). Bleaching may form free radicals and locally heat the sample, representing a real concern for any FRAP experiment. Appropriate tests to ensure that normal physiological processes are not compromised should be performed.
  4. The bleaching step should be brief and significantly faster than the recovery process to minimize substantial recovery during the bleaching step itself.
  5. Imaging-related bleaching outside of the bleaching phase should be minimal, or at least directly quantifiable to allow for correction.
  6. It is imperative that the bleached area remain in focus for the duration of the experiment.

Effect of sample geometry

The geometry of the sample can have a significant impact on the ready interpretation of results. Flat two-dimensional objects are especially suitable for FRAP experiments, and were the first to be examined (Peters et al. 1974). Even though the point-spread function of an optical system is a complex three-dimensional volume, a two-dimensional sample structure permits sharp-edged bleaching at the focal plane of the objective (Brown et al. 1999). Furthermore, the whole bleached area can be conveniently imaged at once. Finally, the dimensionality of diffusion/transport equations can be reduced (Teissie et al. 1978). While FRAP within three-dimensional volumes is possible, a two-photon microscope is needed to achieve a spot-like bleaching volume (Brown et al. 1999; Schmidt et al. 2007). This is based on the fact that the diffraction-delimited oval excitation volume produced by two-photon systems are qualitatively different than hourglass bleaching volumes produced by conventional lasers, or the larger bleaching volumes achievable with arc lamps. Finally, if photobleaching is performed within thin elongated processes, such as axons and dendrites, often the analysis of diffusion or transport can be reduced to one dimension (Schmidt et al. 2007), assuming that axial mobility can be neglected (Zador and Koch 1994; Gabso et al. 1997). While only simple geometries were discussed above, it should be borne in mind that complex geometries can have strong effects on diffusion. For example, it was shown that the presence of spines significantly retards diffusion within dendrites (Santamaria et al. 2006), contributing to compartmentalization.

Another factor to note is the influence of the initial distribution of the fluorophore. The fluorophore may be distributed homogeneously throughout the cytoplasm of a cell (Schmidt et al. 2007), or it may be concentrated in discrete areas, such as the presynaptic terminal (Orenbuch et al. 2012). In the former case, bleaching is performed only within a subregion of the sample. In the latter, bleaching of a subregion of a terminal is applicable (Gaffield et al. 2006), but so too is bleaching of entire presynaptic terminals. In this case, recovery will reflect the movement of fluorophores over long distances from outside the boundaries of the target terminal (Darcy et al. 2006a; Orenbuch et al. 2012) (Fig. 2).

Figure 2.

Fluorescence Recovery After Photobleaching of whole presynaptic terminals. (a) Labeled vesicles (green circles) are transported in and out of presynaptic terminals through the axon. Fluorescence is bleached in a single presynaptic terminal (large blue circle). Thereafter, transport continues, so that intact vesicles enter the terminal and bleached vesicles (black circles) leave it. (b) Depiction of Fluorescence Recovery After Photobleaching of synaptophysin-EGFP in live synapses of wild type (WT) hippocampal neurons. The presynaptic terminals appear as green spots. A white arrow indicates a single terminal which was bleached between the left and center panels. The small right panels illustrate gradual recovery of fluorescence (1, 6, 11.5, 16, 20.5, 26, 31, 36, and 41 min after bleaching). Note that recovery is not complete within the time scale of this experiment. (c) Graph representation of (b). Symbols represent the normalized fluorescence intensity measured from bleached terminals in WT and synapsin triple knockout (TKO) neurons. Both traces were fit with bi-exponential functions using identical time constants (red lines); the lines are extrapolated to emphasize the lack of full recovery. This is interpreted as indicating that although the instantaneous rate of vesicle movement in neurons from both genotypes is similar, a larger fraction of the vesicles is immobilized in WT neurons. (b–c) are adapted from figure 4 of (Orenbuch et al. 2012).

Instrumentation

The expected rate of fluorescence recovery and its decomposition into different time scales profoundly affects the planning and execution of FRAP experiments. The bleaching phase needs to be shorter than the faster kinetic components of recovery, whereas the duration of the post-bleaching phase should be long enough to reliably determine both the slower kinetic components and the final extent of recovery. Not surprisingly, these requirements dictate the properties of the instrumentation. We illustrate this with two contrasting examples where very fast versus very slow processes were assessed.

When fast recovery is expected, bleaching is performed by an intense and temporally discrete burst of excitation light. Moreover, sampling of emitted fluorescence is performed at a very high-frequency. Because the recording period is quite brief, focal stability is usually not a concern. This approach is illustrated by the experimental setup that was used to measure FRAP of fluorescent parvalbumin in Purkinje neurons, an event that occurs in substantially less than 1-s (Schmidt et al. 2007). To produce an intense diffraction-delimited bleaching spot lasting 1–6 ms with a transition time of 1 μs, a Pockels cell was placed in the excitation path of a two-photon microscope. This optical device attenuates the intensity of polarized light by electronically rotating the beam prior to passage through a polarizer. The oval shape and the size of the bleaching volume were defined by the two-photon effect. To permit fast acquisition, the mirrors in the scan head were held in a parked position. Imaging at a frequency of 250 kHz continued for 1.5 s after the bleaching pulse, sufficient time for completion of recovery.

A contrasting example is the case of FRAP experiments used to look at the exchange rate of fluorescently labeled structural synaptic proteins, involving recovery over many hours (Tsuriel et al. 2006, 2009). In this case, the prolonged nature of recovery means that it proved acceptable to have a bleaching step lasting for several seconds. In these experiments, regions of interest were used to define a number of target synapses and an acousto-optic tunable filter was used to specifically subject the regions of interests to intense illumination during several iterations of frame scanning. The bleaching volume was determined by the intersection of the axonal volume and the waist of the hourglass shape of the single-photon laser beam. Acquisition was performed from optical planes whose depth was determined by the confocal pinhole. Sampling of recovery was performed at a very slow rate, every 5, 10 or 60 min. As a result of the length of the recovery period concerns about focal drift had to be addressed. To minimize the potential for drift, the temperature of the sample and the environment were stabilized, and focus drift was corrected automatically before acquiring each time point. Furthermore, z-stacks spanning a range of focal planes were acquired at each time-point, so that in-focus frames could be used for analysis. A different approach for focal stability involves the detection of the reflective surface of the coverslip with infrared light, using it as a reference point for focus correction (such as the Nikon Perfect Focus system, Nikon Instruments, Tokyo, Japan). This approach is suitable only in applications where such surfaces are present, for example when imaging cultured cells.

Both aforementioned examples utilized the diffraction-limited spot of a laser-scanning confocal microscope both for bleaching and acquisition. A different approach is to use an epi-fluorescence arc lamp in a wide-field system, where targeted bleaching is achieved by restricting the area of illumination, for example by narrowing the field diaphragm (Axelrod et al. 1976). Alternatively, an external light source may be used (Roy et al. 2011). These approaches benefit from the advantage of simplicity, but typically at the price of a larger bleaching area which exhibits diffuse edges, unlike the gaussian edges of the confocal spot (Berkovich et al. 2011).

Controls, corrections, and analysis

Controls

Control experiments are necessary to assess whether assumptions made in planning the FRAP experiment are valid. Here, we discuss two types of controls; first, those aimed at establishing that recovery does indeed stem from movement, and second, those designed to test for possible unintended effects on the biological system.

In many instances it is possible to compare FRAP in live and chemically fixed cells, with the assumption that fixation immobilizes the fluorescent marker. In this way, it is possible to assess whether a decrease in fluorescence, which is interpreted as bleaching, is indeed irreversible. This type of control can also address the possibility of other dynamic changes in fluorescence properties; for example, photoactivation or entry into dark states (Malkani and Schmid 2011). Of course, it is also important to consider that the fixation steps themselves may have confounding effects that make interpretation of the results difficult; for example, the generation of an autofluorescence signal (Clancy and Cauller 1998). Another type of control experiment is recommended to establish that the physiological function of the biological preparation is not compromised by the bleaching process. This type of control needs to be separately developed for each experiment, and depends on the availability of a suitable assay that can determine whether cellular functions are unaffected by bleaching. For example, in experiments involving mitosis, the timing of the cell cycle can be calibrated in control cells to verify that photoperturbation does not inhibit cell-cycle progression (Bancaud et al. 2010). In another example, it was shown that bleaching of styryl dyes contained within synaptic vesicles in hippocampal synapses did not deleteriously affect synaptic function by determining that recycling of fresh styryl dye was unchanged in synapses that had been previously subjected to bleaching (Darcy et al. 2006a).

Corrections

In some instances technical limitations prevent the fulfillment of some of the aforementioned assumptions for a FRAP experiment. In such cases, certain corrections can be implemented. For example, if it is not possible to keep the bleaching phase sufficiently brief, substantial recovery may occur within this period, thus distorting the time course of recovery; compensations for this eventuality have been proposed (Kang et al. 2009; Yang et al. 2010). Likewise, if sampling during the prebleach and recovery phases results in significant bleaching, its extent can usually be quantified and can be corrected during offline analysis (Endress et al. 2005; Tsuriel et al. 2006; Bancaud et al. 2010; Wu et al. 2012).

Analysis

The rigor of analysis is ultimately derived from the overall objective of the experiment. In some cases, a semi-quantitative approach may be deemed sufficient. For example, when the aim is to compare the overall mobility of a molecule/structure of interest under different conditions, direct comparison of the recovery traces or of raw time constants may suffice. However, when the aim is to specifically determine a diffusion coefficient (Schmidt et al. 2007), or to establish the type of movement being executed by the moiety of interest [e.g. isotropic, stick-and-diffuse, caged, anomalous or directional (Yeung et al. 2007; Sprague and McNally 2005; Berkovich et al. 2011)], then quantitative analysis is called for. Much information can be extracted from fitting of the recovery trace with various mobility models (for examples see Bancaud et al. 2010; Sprague et al. 2004), taking into account the geometry of both the bleach spot and the sample. For example, binding of dye-bearing molecules with organelles or the plasma membrane (Berkovich et al. 2011) or the presence of obstructions that constrain diffusion (Santamaria et al. 2006) can result in significantly slower apparent diffusion, but their presence can be inferred by careful analysis. More importantly, the nature of the constraints (Sprague et al. 2004) and the binding and unbinding constants (Berkovich et al. 2011) can be extracted. A useful reference in this type of study is the determination of the diffusion of free dye (Sprague et al. 2004). Furthermore, determination of the asymptote of the recovery trace can reveal whether a fraction of the labeled moieties is immobile (Schmidt et al. 2005; Bancaud et al. 2010; Orenbuch et al. 2012), after correcting for the decrement in the total quantity of fluorescent material in the sample that is directly attributable to the bleaching step. The rationale for this statement is that bleached immobile moieties prevent incoming mobile fluorescent ones from associating with a limited number of binding sites, effectively and proportionally reducing the extent of fluorescence recovery (Fig. 1d and e). In some cases, redistribution of the moiety of interest may involve active transport, a fact that can (although not always) result in directional movement (Iliev and Wouters 2007; Scott et al. 2011), which is not expected in the case of diffusion. Another handle on this eventuality is that manipulating the cytoskeleton may arrest transport (Darcy et al. 2006a), but will usually not retard isotropic diffusion.

For the purpose of presentation, FRAP traces are typically normalized by the level of fluorescence measured prior to bleaching, after subtracting the background fluorescence (Bancaud et al. 2010) (Eqn 1); this facilitates visual and statistical comparison of related experiments.

display math(1)

where F(t) is the fluorescence at time t, F(PB) is the fluorescence prior to bleaching and BG is the background value. Note that F(0), the fluorescence just after bleaching, is not necessarily 0.

If the extent of bleaching does not influence the time course of recovery [but see (Brown et al. 1999)], it is helpful to subtract the fluorescence remaining just after bleaching (Eqn 2), thus rescaling the calculated value of recovery to be within a range of 0–1.

display math(2)

When the kinetics of recovery, but not its extent, are unchanged by an experimental perturbation (Campbell and Knight 2007), normalization by the end-point (Eqn 3) can illustrate that even though the proportion of mobile moieties differs, their instantaneous rate of movement does not (Orenbuch et al. 2012).

display math(3)

where F(∞) is the fluorescence at the endpoint (Saxton et al. 1984).

We stress that quantitative analysis of a FRAP experiment depends strongly on the specific implementation of FRAP. Consequently, we do not strive to provide here a full mathematical derivation of FRAP analysis. Readers interested in this subject are referred to excellent publications that provide examples which can be further tailored to the specific requirements of novel experimental conditions (Brown et al. 1999; Sprague et al. 2004; Weiss 2004; Sprague and McNally 2005; Santamaria et al. 2006; Jonsson et al. 2008; Berkovich et al. 2011). These examples address the geometry of the bleach spot, the sharpness of its edges, the dimensionality of the diffusion model, interactions with cellular structures, and obstruction to free diffusion.

Alternatives to FRAP

Several viable alternatives to FRAP exist, which address mobility using different principles; some of these are described briefly below.

Photo activation

This method relies on photoactivatable fluorescent proteins, which in their initial state do not have a fluorescence emission (Patterson and Lippincott-Schwartz 2002; Subach et al. 2009). Brief irradiation at an appropriate wavelength causes a chemical change in the fluorophore, so that it becomes strongly fluorescent. Activation is performed in a confined area and the time-dependent dispersion of fluorescence is recorded (Tsuriel et al. 2009; Roy et al. 2011). A similar approach involves the use of photoswitchable proteins, which irreversibly change their emission properties after a ‘switching’ illumination step (Chudakov et al. 2007; Staras et al. 2010). While photoactivation is essentially a mirror-image of FRAP, some differences between them should be noted. To achieve photoactivation, typically near-ultraviolet light is briefly applied. Although irradiation at this wavelength can be more biologically damaging, this fact is usually offset by the fact that the intensity of irradiation needed to achieve photoactivation is significantly weaker than that typically used for bleaching. Another difference refers to the signal to noise ratio. In FRAP, the initial signal is weak, and increases with time, while in photo activation the inverse is true, resulting in an inversion in the signal to noise ratio between the two methods (Tsuriel et al. 2006). This should be considered when deciding on the specific methodology to use.

Super-resolution and single-molecule localization

In many cases, it is feasible to use live-cell super-resolution techniques to directly view the mobility of individual organelles or of protein complexes. For example, stimulated emission depletion microscopy was used to measure the mobility of styryl-dye loaded vesicles (Westphal et al. 2008; Kamin et al. 2010). Similarly, three-dimensional tracking of single vesicles loaded with quantum dots was achieved at a resolution of approximately 20 nm to examine activity-dependent changes in vesicle mobility (Park et al. 2012). Another important example is the direct visualization of the mobility of glutamate receptors into and out of post-synaptic contacts by nanoscopic single-molecule tracking and localization (Cognet et al. 2006).

Fluorescence correlation spectroscopy

In a fluorescence correlation spectroscopy experiment, a small volume is illuminated within a labeled sample. Labeling needs to be sufficiently sparse to ensure that at any given time, very few fluorophores enter or exit the illumination volume. By analyzing the fluctuation in the fluorescence readout, much can be deduced concerning the mobility of the labeled moieties (Gennerich and Schild 2000, 2002). This method has been employed, for example, to examine mobility of synaptic vesicles within the presynaptic terminal (Yeung et al. 2007). A related technique, Fluorescence fluctuation spectroscopy has been employed in a similar capacity (Jordan et al. 2005).

Applications in neuroscience

FRAP and related techniques have been used to study numerous questions of fundamental neurobiological interest. In this section, we explore some of these. The section concludes with a discussion of a particularly powerful application of FRAP for evaluating inter-synaptic vesicle movement in small central synapses.

Mobility of neurotransmitter receptors and transporters

Neurotransmitter receptors are known to be highly concentrated in the post-synaptic specializations formed opposite the active zone. Early work revealed that acetylcholine receptors found on the surface of developing muscle are very restricted in their movement (Axelrod et al. 1976). Later, FRAP and other techniques expanded this view by demonstrating that receptors, as well as transporters, move on the surface of the neurons in and out of synaptic areas, in addition to being internalized by endocytotic mechanisms (Jacob et al. 2005; Rasse et al. 2005; Bats et al. 2007; Lin and Huganir 2007; Eriksen et al. 2009).

The cytoskeleton and axonal transport

The cytoskeleton of axons and dendrites has been extensively investigated with FRAP. Its longitudinal organization and dynamic nature is especially amenable to examination by this technique. Moreover, FRAP was instrumental in exploring the differences in cytoskeletal stability in different locales and under various physiological conditions (Edson et al. 1993; Takeda et al. 1995; Star et al. 2002; Iliev and Wouters 2007). Another related subject, which has been addressed exquisitely with FRAP and photoactivation approaches, is the nature of axonal transport mechanisms of various classes of neuronal proteins (Roy et al. 2011; Tang et al. 2012).

Mobility of calcium and its buffers

Because of its involvement in a myriad of processes, the dynamics of calcium ions within neurons is of intense interest (Augustine et al. 2003). Research on this subject has revealed that the properties of endogenous calcium buffers are critical determinants of these dynamics, since they strongly influence the concentration and distribution of calcium ions (Gabso et al. 1997). The diffusive properties of two of these buffers, parvalbumin and calbindin D28k, have been studied by FRAP (Schmidt et al. 2003, 2005, 2007). A quantitative approach revealed that parvalbumin is freely diffusible, while the mobility of calbindin D28k is restricted by intermolecular interactions.

Synapse tenacity

Although it has long been known that the synapse is plastic, the extent of exchange of its components had not been well-characterized. This topic was addressed directly in a series of FRAP studies (Tsuriel et al. 2006, 2009). It was shown that different structural components of the synapse (post-synaptic density proteins, active zone proteins, vesicle-associated proteins etc.) are replaced at vastly different rates. Moreover, while some of these rates of movements were influenced by neuronal activity patterns, others were not.

Intra- and inter-synaptic vesicle mobility

Close examination of the presynaptic terminal reveals within it a large collection of vesicles that are packed at a high density in the vicinity of the active zone (Palay 1956). In central synapses and many peripheral junctions, this vesicle clustering is consistent with the idea that vesicles are immobilized, probably by interactions with cytoskeletal elements and by cross-interactions between the vesicles themselves through vesicle-associated proteins such as the synapsins (Gitler et al. 2004b; Denker and Rizzoli 2010; Pechstein and Shupliakov 2010). Nonetheless, vesicles also have some intra-terminal mobility, as revealed by FRAP, fluorescence correlation spectroscopy and fluorescence fluctuation spectroscopy (Li and Murthy 2001; Jordan et al. 2005; Gaffield et al. 2006; Yeung et al. 2007), perhaps reflecting the mobilization of vesicles from the cluster to the active zone upon demand (Kamin et al. 2010). Conversely, in some specialized terminals, such as lizard cone photoreceptors, FRAP experiments reveal that their vesicles are highly mobile with movement dynamics approaching the rate of free diffusion (Rea et al. 2004). Recent work in hippocampal neurons, using a combination of FRAP, photoconversion of styryl dyes, and electron microscopy (Darcy et al. 2006a,b; Staras et al. 2010) has observed that a significant number of functional releasable vesicles are also transported actively between distinct terminals (Staras and Branco 2010). This observation significantly extends our view of the presynaptic terminal, by adding the synapse-spanning vesicle ‘superpool’ (Westphal et al. 2008; Staras et al. 2010) to the known list of functional vesicle pools (Denker and Rizzoli 2010; Alabi and Tsien 2012). The discovery of the superpool also raised the question of the regulation of inter-synaptic transport of vesicles (Fig. 2). Because deletion of the synapsins is known to cause a pronounced decrease in the vesicle population within presynaptic terminals of hippocampal neurons (Gitler et al. 2004a; Siksou et al. 2007), it was hypothesized that in the absence of synapsins more vesicles are mobile, and that consequently, they should redistribute into the axon. Indeed, FRAP of the entire vesicle population of presynaptic terminals revealed that vesicle mobility was enhanced by the deletion of the synapsins, and in agreement, 3D electron-microscopic reconstruction of the terminals and adjacent axonal segments found more vesicles in the axons of neurons devoid of synapsins (Fornasiero et al. 2012; Orenbuch et al. 2012). Moreover, extrapolation of the FRAP traces illustrated that deletion of the synapsins enlarged the fraction of mobile vesicles (Orenbuch et al. 2012) (Fig. 2b–c). Finally, synapsins were found to specifically affect the mobility of resting pool vesicles, by comparing FRAP experiments using the styryl dye FM1-43, which labels recycling-pool vesicles, and of synaptophysin I-EGFP, which labels all synaptic vesicles (Orenbuch et al. 2012). Interestingly, the list of proteins participating in the regulation of inter-synaptic vesicle traffic was recently expanded to include α-synuclein, a protein associated with Parkinson's disease (Scott and Roy 2012).

Concluding remarks

FRAP is a group of related techniques useful for probing mobility of sub-resolution particles, be they proteins or organelles, in living functional cells. The ability to genetically manipulate cells to introduce a host of different fluorescent labels into a myriad of structures and proteins has substantially enhanced the utility of FRAP. Advances in instrumentation, such as the introduction of secondary scanners and of ultrafast and sensitive super-resolution acquisition, will no doubt amplify the applicability of FRAP in traditionally quantitative research fields like neuroscience.

Acknowledgements

This work was supported by grant 1427/12 of the Israel Science Foundation (DG). The authors do not report any conflict of interest.

Ancillary