Photochemical tools to study dynamic biological processes

Abstract

Light-responsive biologically active compounds offer the possibility to study the dynamics of biological processes. Phototriggers and photoswitches have been designed, providing the capability to rapidly cause the initiation of wide range of dynamic biological phenomena. We will discuss, in this article, recent developments in the field of light-triggered chemical tools, specially how two-photon excitation, “caged” fluorophores, and the photoregulation of protein activities in combination with time-resolved x-ray techniques should break new grounds in the understanding of dynamic biological processes.

A prerequisite to study the dynamics of biological processes is to achieve a spatiotemporal control of the studied phenomenon. This requires putting this process under the control of a conditional trigger. Light might be an ideal orthogonal external trigger since, if judiciously selected, it will not affect the biological environment.

Phototriggers and photoswitches provide the capability to rapidly cause the initiation of a wide range of dynamic biological processes (Fig. 1). The first strategy refers to caged compounds that release biological effectors after photolysis (Corrie and Trentham, 1993; Adams and Tsien, 1993; Marriott, 1998; Pelliccioli and Wirz, 2002; Goeldner, 2005; Mayer and Heckel, 2006; Ellis-Davies et al., 2007). The second refers to synthetic photochromic systems (a “photoswitch” molecule), which conditionally activate a protein by attaching a photoisomerizable ligand (Kramer et al., 2005, Volgraf et al., 2006; Gorostiza and Isacoff, 2008; Sakata et al., 2005).

Figure 1. Photoregulation of functional proteins using a phototrigger or a photoswitch.

Caged compounds are biological effectors, which are rendered biologically inert (or caged) by chemical modification with a photoremovable protecting group (Adams and Tsien, 1993; Marriott, 1998; Bochet et al., 2002; Pelliccioli and Wirz, 2002) (Fig. 2). Illumination results in a localized concentration jump of the biological active effecter, turning on the biological process. Ideally, the photoremovable protecting groups should be water soluble, stable to hydrolysis, their photodeprotection should take a single reaction pathway and occur with a high efficiency (high product quantum yields together with high absorbance), the photolytic side-product(s) should not interfere with the ongoing photolysis and should be biologically inert, and finally the release of the biological effector should be fast, in agreement with the time scale of the analyzed biological event.

Figure 2. Several photolabile groups used for the caging of biologically active compounds (LG=leaving group).

The idea of using light to unleash biologically active compounds from inert precursors (uncaging) was introduced over 30 years ago, using o-NB [o-nitrobenzyl] (Fig. 2) esters of cAMP (Engels and Schlaeger, 1977) or ATP (Kaplan et al., 1978) to induce, through the action of light, rapid enzyme activation or ion channels activities. The o-nitrobenzyl photoremovable derivatives were by far the mostly developed series, even though their overall properties might not be optimal, in particular in the near-UV and visible range.

This concept was applied to investigate neurological processes using caged neurotransmitters (Callaway and Yuste, 2002; Hess, 2005; Gillespie et al., 2005), in cell biology using caged second messengers and cellular signaling molecules (Adams et al., 1988; Ellis-Davies and Kaplan, 1994; Hagen et al., 2005; Pavlos et al., 2005; Walker et al., 1987; Dinkel and Schultz, 2003; Li et al., 1998), in gene expressions using caged small molecule inducers (Cambridge, 2005) and caged RNA (Ando et al., 2001; Mikat and Heckel, 2007), and in structural biology by combining caged compounds with time-resolved analytical techniques (IR, x-ray, etc.) (Barth, 2005; Bourgeois and Royant, 2005; Bourgeois and Weik, 2005). Recent efforts prompted the development of photoremovable groups that have increased photochemical efficiencies in the near-UV (Papageorgiou, et al., 2005; Specht et al., 2006; Shembekar et al., 2007) and visible range (Hagen et al., 2008) and∕or displaying faster photorelease kinetics (Furuta, 2005; Givens and Yousef, 2005; Hagen et al., 2008) to allow more sophisticated applications.

On the other hand, precise light-induced protein activity regulation has recently been performed using synthetic optical switches (Kramer et al., 2005, Volgraf et al., 2006; Gorostiza and Isacoff, 2008, Sakata et al., 2005). Therefore, the protein of interest was modified by attaching photoisomerizable ligands. Theses ligands tethered via an optical switch can function in two ways. They can conditionally block the active site of an enzyme or the pore of a channel or they can reversibly present ligand to a protein binding site and conditionally trigger the protein activity (Fig. 1).

All those cutting edge applications of the concept of “caging” groups have recently been thoroughly reviewed (Pelliccioli and Wirz, 2002; Goeldner and Givens, 2005; Mayer and Heckel, 2006; Ellis-Davies et al., 2007; Lee et al., 2009). Therefore we will focus the discussion, in Secs. 1, 2, 3, on recent developments and applications in this field: such as two-photon excitation process, caged fluorophores, and the photoregulation of protein activities in combination with time-resolved x-ray techniques. Two-photon excitation permits a tremendous increase in the spatial resolution of the released biomolecule, including fluorescent signals associated with cellular proteins to investigate events in living cells, while the last approach should allow new insights into the understanding of protein dynamics.

TWO-PHOTON UNCAGING

The challenging problems encountered in life sciences, and in particular in neurosciences, need new powerful techniques to trigger and record cellular signals. Indeed, there is a strong demand for new methods and techniques with an increased spatiotemporal resolution and∕or allowing complex patterned illuminations (Dedecker et al., 2007; Papagiakoumou et al., 2008; Lutz et al., 2008). Two-photon excitation can intrinsically provide such a fine spatiotemporal control, for example, during a photochemical reaction (photolabile protecting groups: uncaging) (Ellis-Davies, 2007) or fluorescence (imaging) (Denk et al., 1990; Zipfel et al., 2003). Indeed, a molecule can reach an excited state not only by the absorption of a single photon of energy E=hv, but also by the simultaneous absorption of two photons of half energy E^′=hv∕2 (Fig. 3a). The interaction between the electric field of the excitation light (E) and the electronic cloud of a molecule induces a charge redistribution so that the molecular dipole moment (μ) is modified. The resulting μ is usually described as a development of increasing powers of the electric field E, as described in Eq. 1 (Terenziani et al., 2008). The static dipolar moment of the molecule is μ₀, and αE represents the effects related to the polarizability (linear optical properties). The terms β⋅E² and γ⋅E³ lead to the first and second order molecular nonlinear optical properties, while β and γ are, respectively, the first order and second order molecular hyperpolarizabilities,

$$\mu ={\mu }^0+\alpha \cdot E+\beta \cdot E^2+\gamma \cdot E^3\dots .$$ (1)

The cubic term γ⋅E³ deals in particular with the two-photon absorption phenomenon. The excitation probability of a molecule (P) is in this case dependent of the squared light intensity [Eq. 2]

$$P=1∕2\cdot {\sigma }_2\cdot I^2$$ (2)

This quadratic dependence of P versus I is crucial for the spatial localization of this nonlinear optical phenomenon. The excitation will occur only where the light intensity is maximum (Fig. 3b), typically at the focal point of an optical system and only with a pulsed laser as excitation source (yielding to spatial and temporal compression of photons by a lens system and a pulsed laser, respectively). The parameter σ₂ is called “two-photon absorption cross section”; it indicates the efficiency of the two-photon absorption by a molecule, and is generally expressed in Göppert-Mayer unit (GM) in honor of Maria Göppert-Mayer who predicted theoretically this phenomenon in 1931 (Göppert-Mayer, 1931) (1 GM=10⁻⁵⁰ cm⁴ s photon⁻¹ molecule⁻¹). The σ₂ is directly correlated with the imaginary part of the second order hyperpolarizability γ (Andraud et al., 2000). It is important to note that some physicists symbolize this parameter as δ (Lee et al., 2005; Albota et al., 1998), notation also adopted for uncaging. In addition to an intrinsic fine 3D localization of the excitation, this nonlinear process is obtained with low energy wavelength (typically with IR light for molecules which absorb classically in the UV). This induces limited phototoxicity for cells, tissues, or organs, and an increased penetration depth (tissues transparency windows from 700 to 1,000 nm), for example, in highly scattering tissues such as brain slices (Helmchen and Denk, 2005; Zipfel et al., 2003). This type of excitation shows an increasing interest, in particular, in the photorelease of active small molecules (uncaging). Nevertheless, efficient photoremovable groups used upon one-photon excitation are rarely efficient upon two-photon processes, and exhibit low two-photon uncaging action cross sections (δ_u⋅Φ_u), typically <0.1 GM (Goeldner, 2005). This value δ_u⋅Φ_u is matching to ε⋅Φ in one-photon experiments, taking into account both the efficiencies of the excitation and of the photochemical reaction. Thus, these classical cages cannot be used in new patterned illuminating systems, such as holographic photolysis microscopes or two-photon macrophotolysis. The laser power needed for two-photon uncaging would be too high, leading to cellular damages, as endogenous chromophores (such NAD(P)H, flavins, and porphyrinic systems) could also absorb and generate reactive oxygen species, thus inducing an oxidative stress. (Kiskin et al., 2002). Collaborative efforts between chemists, physicists, and biologists have led to the design and the synthesis of new photolabile protecting groups with increased two-photon uncaging action cross sections. The molecular engineering of the two-photon uncaging cross sections was performed on classical platforms such as MNI [4-methoxy-7-nitroindolinyl], o-NB, or DMNPB [3-(4,5-dimethoxy-2-nitrophenyl)-2-butyl] (Fig. 2). Classical strategies known in optimization of organic molecules for nonlinear optics were used. At this point, it is important to remember that the two-photon absorption process occurs through second order nonlinear optical properties. This is of crucial importance for the molecular engineering of new systems. Contrary to first order nonlinear optical properties (such as second harmonic generation), to be active in two-photon absorption, molecules do not need to be noncentrosymmetric. The simplest active systems upon two-photon excitation are linear (or one-dimensional) and consist in a more or less rigid conjugated central core surrounded by conjugated systems bearing electron withdrawing or donating groups. The modifications used to improve two-photon absorption are classically the lengthening of the conjugated system and the increase of the electron donating or withdrawing ability of one part of the molecular system (He et al., 2008). These modifications have led to new caging platforms with increased two-photon uncaging cross sections, such as CDNI [4-carboxymethoxy-5,7-dinitroindolinyl] (Fig. 3c) (Ellis-Davies et al., 2007), modified nitrobenzyls (Aujard et al., 2006) or nitrophenethyl derivatives (Gug et al., 2008a). Recently, symmetrical architectures (Fig. 3c) were described with ultra-efficient glutamate two-photon uncaging cross section, over 3 GM within a large wavelength range (750–820 nm) in vitro (Gug et al., 2008b).

Figure 3. (a) Two-photon versus one-photon processes.

(b) Fluorescence localization in one-photon (all light pathway) and two-photon (focal point) excitation of an organic fluorophore (Hayek et al., 2006) (c) Dissymmetric donor-acceptor (CDNI-Glu) and symmetric acceptor-acceptor (BNSF-Glu) caged glutamates with high two-photon uncaging action cross sections.

These new efficient two-photon photolabile protecting groups can be conjugated to various biologically active molecules, such as neurotransmitters, gene effectors, or combined to photocleavable ions chelators for calcium uncaging for example (Ellis-Davies, 2007; Nikolenko et al., 2007; Neveu et al., 2008; Gug et al., 2008a, 2008b). They should play in the next years a critical role to understand the neuronal circuitry in brain slices and to study synaptic integration. They would also permit experiments in small living animals, since the two-photon excitation allows either an increase in depth penetration of the uncaging signal (over 750 μm) or a decrease in the excitation power compared with classical cages developed for one-photon applications (Kerr and Denk, 2008; Helmchen and Denk, 2005; Lutz et al., 2008; Papagiakoumou et al., 2008).

In addition, two-photon excitations would be of great interest as it can induce the generation of reactive oxygen species in combination with a sensitizer, which could be used in FALI (fluorophore assisted light inactivation of proteins), with an increased spatiotemporal control known as microFALI (Tanabe et al., 2005; Jacobson et al., 2008).

Another interesting outcome for tracking molecular and cellular dynamics is fluorophore uncaging of fluorescently labeled proteins. The development of new caged fluorophores specifically engineered for two-photon excitation represents again a crucial issue. They would allow new development in emerging microscopies with nanometer resolution, such as PALM (photoactivated localization microscopy), which takes advantage of the intrinsic 3D resolution of the two-photon excitation (2P-PALM) (Hess et al., 2006; Folling et al., 2008, Gould et al., 2009).

CAGED FLUOROPHORES

Since the introduction of the green fluorescent protein as a genetically encoded fluorophore 15 years ago (Chalfie et al., 1994), AFPs (autofluorescent proteins) have been employed by a plethora of researchers in both in vivo and in vitro applications (Chudakov et al., 2005). Genetically encoded PAFPs (“photoactivatable” fluorescent proteins) make up a small category of fluorescent proteins (Lukyanov et al., 2005) but allow more sophisticated use over AFPs (Hell, 2007). The fluorescence characteristics of photoactivatable proteins can be controlled by irradiation at a specific wavelength, intensity, and duration. This provides unique possibilities for the optical labeling and tracking of living cells, organelles, and intramolecular molecules in a spatiotemporal manner (Lippincott-Schwartz and Patterson, 2008). But all those genetically encoded fluorescent proteins have two intrinsic limitations, especially for trafficking studies, first their big size and a restricted number of fluorophores (Eisenstein, 2006), which do not necessarily display the “best” photophysical properties in terms of brightness and stability. To overcome these limitations the specific labeling of fusion proteins with synthetic small molecule photoactivated fluorophores (also called caged fluorophores) has been developed.

The general strategy in masking fluorescence is to perturb the electronic structure by attachment of the photoremovable group to render the molecule either nonfluorescent or very weakly so. Photoactivation removes the protecting group and abruptly switches on the fluorescence of parental dyes. Desirable properties for caged fluorophores include fast release of the fluorophore and a large enhancement of the fluorescence in response to brief irradiation, the fluorophore being selected for its photophysical properties: high brightness and weak photobleaching at selected wavelengths. More recently, the fluorophore formation can result from a photochemically induced intramolecular rearrangement of a nonfluorescent precursor, creating the fluorophore in situ (Gagey et al., 2007). Caged fluorophores initially described on fluorescein and rhodamin derivatives (Theriot and Mitchison, 1991; Ottl et al., 1998) have been reviewed (Mitchison et al., 1998) and prompted innovative cell biology applications (Theriot and Mitchison, 1991). Subsequently, 7-hydroxy coumarin derivatives were caged as o-nitrobenzyl ether derivatives (Zhao et al., 2004). These molecules allowed a 200-fold fluorescence enhancement after UV photolysis and displayed an increased “uncaging”cross section over previously reported caged fluorophores. Recently, novel orthonitrobenzyl fluorescein derivatives (caged TokyoGreens) displaying a larger fluorescence enhancement were described (Kobayashi et al., 2007). We have recently reported a new class of caged 7-hydroxycoumarin fluorophore (Orange et al., 2008) using an DMNPB (orthonitro phenethyl ether) as a photoremovable protecting group, which was initially described for the caging of carboxylic acids (Specht et al., 2006). This probe displayed a high uncaging cross section at 365 nm and showed a rapid enhancement of fluorescence after irradiation.

Nevertheless, all those caged fluorophores will have a limited use in cell biology, especially because these blue-emitting fluorescent dyes are excited at “toxic” visible light (400–430 nm). Therefore it becomes urgent to develop new red-emitting caged fluorophores, which are excited at wavelength, where tissue turbidity and cellular autofluorescence are low (Billinton and Knight, 2001).

Indeed, even with the best photoactivable fluorophore in hand, the biggest limitation for visualizing molecular events in living cells remains its specific recognition of the cellular component of interest, i.e., a cellular protein. This limitation can be overcome by using small genetically encoded peptide tags and complementary small organic affinity probes (Griffin et al., 1998) with sufficient specificity and affinity to be used in living cells. To be of practical use, such an approach has to provide a highly specific labeling reaction both in vitro and in living cells (Gronemeyer et al., 2005; Johnsson et al., 2005).

Therefore, the development of specific and small tags for the labeling of proteins in combination with more efficient caged fluorophores should become important research tools for tracking the spatiotemporal dynamics of molecular movements in cell biology (Fig. 4).

Figure 4. Cellular protein trafficking studies using caged fluorophores and small genetically encoded peptide tags.

X-RAY METHODOLOGIES USING CAGED COMPOUNDS

The dynamic behavior of proteins is one of the fundamental properties of living systems. Recent developments in x-ray source technology provide new opportunities for the rapid imaging of macromolecule in three dimensions within crystals (Neutze et al., 2000, 2004). This emerging field in structural biology has nowadays two distinct sections.

The first one is the “pump-probe” crystallography, whereby a reaction is rapidly triggered at room temperature using a “pump” (mostly light) and the structural changes within the protein are followed in real time using an x-ray “probe” (Moffat, 1998). The Laue crystallography (a technique that uses a white beam x-ray probe) has been the only method used for such studies (Bourgeois et al., 2003, Bourgeois and Royant, 2005; Bourgeois and Weik, 2005; Schotte et al., 2003; Baxter et al., 2004; Anderson et al., 2004; Rajagopal et al., 2005; Ihee et al., 2005), due to the rapid x-ray diffracting data collection on a nonrotating static crystal (Bourgeois et al., 1996) provided by this technique. The most impressive success using this methodology is the 100 ps studies on the photodissociation of the carbon-monoxide molecule bound to the heme of myoglobine (Srajer et al., 2001; Schotte et al., 2003; Hummer et al., 2004). But mostly all pump-probe dynamic x-ray studies have been done on direct light activation on built-in photosensitive moieties.

Usually, natural chromophores, such as retina of bacteriorhodopsin (Oesterhelt and Stoeckenius, 1971), 4-hydroxycinnamoic acid from photoactive yellow protein (Hoff et al., 1994), or photosensitive coordination bonds, such as iron-carbone monoxide bonds in heme containing proteins (Austin et al., 1975), have been used. Therefore the application of the caging concept to this field as light activated triggers was largely put forward as one of the most elegant applications of high-brilliance synchrotron machines in structural biology.

But even if the Laue technique allows x-ray exposure times as short as 100 ps, the time resolution for such studies would be as high as permitted by the photoactivation process (down to a few of nanoseconds, for the faster photolabile groups (Furuta, 2005; Givens and Yousef, 2005). Therefore, the biggest limitation for real-time resolved crystallography studies employing caging groups is to build up a very efficient and fast phototrigger to be able to collect a real movie of the protein activity.

The second section of time-resolved crystallography is the use of physical techniques to control the rates of reactions within crystals (Hajdu et al., 2000). Therefore, the biological reaction is stopped (usually by lowering the temperature), and monochromatic x-ray diffraction data may be collected. Recently, so-called “kinetic crystallography” studies were used to accumulate and visualize unstable intermediate in crystals (Stoddard, 2001; Petsko and Ringe, 2000; Schlichting, 2000). In principle, the easiest way for the triggering of a protein response in a crystalline state consists of letting the natural effectors diffuse into the crystal just by soaking. But this is usually a very slow process, which therefore prevents the synchronous built-up of intermediate states.

The most frequently used protocol to trigger the formation of intermediate states in protein crystallography again relies on direct light activation of built-in photosensitive moieties. But for the vast majority of proteins that do not contain a naturally photoactivatable group, caged compounds represent an ideal way to trigger protein activity by UV-visible light (Schlichting and Goody, 1997). Especially, the ability of such compounds (i.e., caged substrates) to be photolyzed at cryotemperature (Specht et al., 2001; Ursby et al., 2002) (and below the dynamical transition temperature of the studied protein) in combination with appropriate temperature profiles, represent a nice way to accumulate protein intermediates, which can then be trapped by cooling the crystal to 100 K (Bourgeois and Royant, 2005; Bourgeois and Weik, 2005). This technique does not require the use of efficient caging groups and therefore a large number of proteins may, in principle, be studied. This methodology was, for example, applied to structural changes in rapid enzymes, such as cholinesterase (Colletier et al., 2007), where the catalytic reaction has been described to occur at the bottom of a deep narrow gorge. A “kinetic” crystallography study was developed to structurally address the product traffific in acetylcholinesterase, where UV-laser-induced cleavage of a photolabile precursor of an enzymatic product analog (e.g., caged arsenocholine) was performed in a temperature-controlled x-ray crystallography regime, to allow the structural monitoring of the traffic of the enzymatic product in acetylcholinesterase. The results were consistent with the idea that choline, the enzymatic product, exits the enzyme catalytic side either via the gorge or via an alternative “backdoor” trajectory.

Related to those methodologies are the emerging possibilities for time-resolved structural probes of photoregulated proteins without using x-ray diffraction, in particular, using x-ray scattering (Neutze et al., 2001; Georgiou et al., 2006; Ihee et al., 2005b; Kim et al., 2006) and x-ray absorption spectroscopy (Saes et al., 2003) from noncrystalline samples. A major drawback of this technique is that all structural information concerning local-conformational changes within the protein, overlap with other signals, and it is therefore impossible to extract specific structural dynamics concerning one site within the protein. Therefore, a prerequisite for success in those experiments has been the presence of heavy elements in the studied system. The increased x-ray flux available from an XFEL (x-ray free electron laser) should also make it possible to recover, with confidence, smaller signals over a higher background, and hence improve the general application of those techniques. The noncrystalline sample used here, opens this field to the study of “larger” protein movements, which may lead to crystal disorder or even rupture, ruining the diffracting power of the sample. For example, time-resolved x-ray scattering methodology could be used for the study of membrane proteins catalyzing vectorial transport (Andersson et al., 2008).

Again, success relies on direct light activation of built-in photosensitive moieties; therefore, the extension of these techniques to proteins that are not responsive to light requires the design of adapted photochemical triggers.

Photochemical triggering of biological events can be achieved either from caged biological effectors or caged proteins (Goeldner, 2005). Clearly, it is more advantageous to cage directly a protein because a stoichiometric irreversible modification leads to a stable inactivated enzyme or receptor that requires a minimum amount of photons for its reactivation. Only a few examples of protein caging are found in the literature, due to the difficulty in targeting a suitable caging group to the desired site of the protein (Loudwig and Bayey, 2005; Lawrence, 2005). This has been successfully achieved by selective modification of strategic cysteine residues, either natural (Arabaci et al., 1999; Marriott and Heidecker, 1996) or incorporated by site-directed mutagenesis (Bayley et al., 1998). The extension of protein caging to residues other than cysteines is more difficult and requires a targeted site-directed modification (Turner et al., 1987; Loudwig et al., 2003). Alternatively, unnatural amino-acid mutagenesis enabled direct incorporation of photolytic precursors of various amino acids at a precise position in the sequence (Heckler et al., 1984; Beene et al., 2003; Noren et al., 1989; Stromgaard et al., 2004) and could here find a nice application. But the major actual drawback for the specific incorporation of a non-natural photoactivatable amino-acid into protein is the very limited amount of functional protein produced using this technique. Therefore the development of new, fast, and efficient tools for the triggering of biological response is still needed.

For example, we are developing a new strategy to understand, at the atomic level, the mechanism by which membrane transporter proteins can actively transport substrate using x-ray scattering methodologies. The triggering of the targeted transporter protein, which needs to undergo an important structural rearrangement to catalyze the transport of specific molecules or ions across an membrane bilayer (Abramson et al., 2003; Faham et al., 2008; Forrest et al., 2008), should be possible by using both site-directed cysteine mutants of the studied transporter protein in combination with cysteine-reactive intramolecular photocleavable crosslinkers (Omran and Specht, 2009) (Fig. 5).

Figure 5. Proposed mechanism for the photoregulation of transporter protein activity using photocleavable crosslinkers.

This class of membrane proteins needs to undergo an important structural rearrangement to catalyze the transport of specific molecules or ions across membrane bilayer. Therefore, they could be locked in one conformation by using both site-directed double cysteine mutants of the protein and intramolecular crosslinking. The UV photolysis of such modified transporters should rapidly and efficiently cleave the crosslinker, leading to the triggering of their activities.

Application of efficient photoswitches for the triggering of protein activities in combination with time-resolved x-ray methodologies should also provide very elegant tools for protein dynamics studies. Specially, the very rapid kinetics of activation and the reversibility of such tools are very attractive for “real” time-resolved studies.

SUMMARY

In the past 30 years, more and more sophisticated photochemical tools have been developed for the triggering of biological processes. Phototriggers (Corrie and Trentham, 1993; Adams and Tsien, 1993; Marriott, 1998; Pelliccioli and Wirz, 2002; Goeldner and Givens, 2005; Mayer and Heckel, 2006; Ellis-Davies et al., 2007) and photoswitches (Kramer et al., 2005, Volgraf et al., 2006; Gorostiza and Isacoff, 2008) provide the unique capabilities to rapidly cause the initiation of a wide range of dynamic biological processes (Lee et al., 2009). A further improvement was recently given by two-photon excitation. Especially the two-photon photo-labile protecting groups, that should play in the next years a critical role to understand the neuronal circuitry in brain slices, to study synaptic integration, or even for experiments in live small animals, since the two-photon excitation allows a fine spatiotemporal control of a photochemical reaction and a deeper penetration of light (over 750 μm).

The developments of new red emitting caged fluorophores sensitive to two-photon excitation combined to the development of specific and small tags for the in vivo labeling of proteins in living cells (Gronemeyer et al., 2005; Johnsson et al., 2005) should provide unique possibilities for trafficking studies of cellular proteins.

Applications of phototriggers and photoswitches on structural biology should also break new ground in the understanding of protein dynamics, especially using the new opportunities provided by the more powerful out coming x-ray sources (Neutze et al., 2004). But this still depends on the development of very fast and efficient photochemical tools for the triggering of proteins activities.