Primary Light-Induced Reaction Steps of Reversibly Photoswitchable

Subscriber access provided by SUNY DOWNSTATE

Article

Primary Light-Induced Reaction Steps of Reversibly Photoswitchable Fluorescent Protein Padron0.9 Investigated by Femtosecond Spectroscopy Arne Walter, Martin Andresen, Stefan Jakobs, Jörg Schroeder, and Dirk Schwarzer J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/jp512610q • Publication Date (Web): 23 Mar 2015 Downloaded from http://pubs.acs.org on March 25, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Primary Light-Induced Reaction Steps of Reversibly Photoswitchable Fluorescent Protein Padron0.9 Investigated by Femtosecond Spectroscopy Arne Walter†, Martin Andresen†, Stefan Jakobs†,‡ , Jörg Schroeder§, Dirk Schwarzer*† †

Max Planck Institute for Biophysical Chemistry, Göttingen, Germany

‡

Department of Neurology, University Medical Center of Göttingen, Germany

§

Institute for Physical Chemistry, Georg-August University of Göttingen, Germany

ABSTRACT: The reversible photoswitching of the photochromic fluorescent protein Padron0.9 involves a cis-trans isomerization of the chromophore. Both isomers are subject to a protonation equilibrium between a neutral and a deprotonated form. The observed pH dependent absorption spectra require at least two protonating groups in the chromophore environment modulating its proton affinity. Using femtosecond transient absorption spectroscopy we elucidate the primary reaction steps of selectively excited chromophore species. Employing kinetic and spectral modelling of the time dependent transients we identify intermediate states and their spectra. Excitation of the deprotonated trans species is followed by excited state relaxation and internal conversion to a hot ground state on a timescale of 1.1-6.5 ps. As the switching yield is very low

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 36

(Φtrans→cis = 0.0003±0.0001) direct formation of the cis isomer in the times-resolved experiment is not observed. The reverse switching route involves excitation of the neutral cis chromophore. A strong H/D isotope effect reveals the initial reaction step to be an excited state proton transfer with a rate constant of kH = (1.7 ps)-1 (kD = (8.6 ps)-1) competing with internal conversion (kic = (4.5 ps)-1). The deprotonated excited cis intermediate relaxes to the well-known long lived fluorescent species (kr = (24 ps)-1). The switching quantum yield is determined to be low as well, Φcis→trans = 0.02±0.01. Excitation of both the neutral and deprotonated cis chromophore is followed by a ground state proton transfer reaction partially re-establishing the disturbed ground state equilibrium within 1.6 ps (deuterated species: 5.6 ps). The incomplete equilibration reveals an inhomogeneous population of deprotonated cis species which equilibrate on different timescales.

1

INTRODUCTION Reversibly photoswitchable fluorescent proteins (RSFPs) are structurally similar to the green

fluorescent protein (GFP), but may be reversibly photoswitched between metastable fluorescent ‘bright’ and non-fluorescent ‘dark’ states1. They exhibit a GFP-like fold, namely a beta-barrel enclosing an alpha helix containing the autocatalytically formed chromophore. RSFPs with ‘positive’ switching characteristics are switched to a fluorescent 'bright' state when irradiated with fluorescence-inducing light, while ‘negative’ switching RSFPs are switched to a nonfluorescent 'dark' state2. RSFPs have been used for a number of applications, including protein tracking3, data storage 4,5 and, most prominently, for super-resolution microscopy 5-7.


Page 3 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Several studies utilizing X-ray crystallography revealed a light-driven cis-trans isomerization of the chromophore as the molecular key event in the switching process

1,8,9

. In all reported

RSFPs so far, the bright state adopts a cis-configuration, but fluorescent proteins with transconfiguration have been described10. Hence, a specific chromophore configuration is not sufficient to explain the occurrence of fluorescence. Rather, chromophore torsion, its stabilization within the protein matrix and protein flexibility in general have been discussed to co-determine the fluorescence yield11-13. In addition, the protonation state of the chromophore is of key importance. For some RSFPs, cis-trans isomerization seems to be coupled with proton transfer; indeed, also pH-induced isomerization has been shown14. For most GFP-like proteins, only the deprotonated chromophore fluoresces with high yield, while for the neutral (protonated) species fluorescence is weak. The weak fluorescence of the neutral chromophore in many cases is caused by an excited state proton transfer (ESPT) and subsequent emission by the deprotonated form15. Still, many of the mechanistic details of reversible photoswitching are discussed controversially including the role of the protonation state of the amino acid residues in the vicinity of the chromophore 1. The well-investigated green fluorescent RSFP Dronpa exhibits negative switching characteristics 16. The bright state was identified to be the deprotonated cis-isomer 8. As an initial step for switching from the dark to the bright state ESPT has been discussed for years17, but was recently excluded by Warren et al.

18

in favor of direct trans to cis isomerization followed by a

ground state proton transfer. Padron is a derivative of Dronpa differing in only a few amino acids (T59M, V60A, N94I, P141L, G155S, V157G, M159Y, F190S), but shows opposite (i.e. positive) switching characteristics for which only V157G, M159Y are essential2. At thermal equilibrium, Padron



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 36

adopts the dark state. X-ray studies on the variant Padron0.9 identified the bright state to have cis- and the dark state to have trans-configuration12. Similar to Dronpa both isomers absorption spectra exhibit pH-dependencies with clearly distinguishable absorption bands at 400 nm and 500 nm which are assigned to the protonated and deprotonated chromophore, respectively12. Faro et al.19 succeeded to cryo-trap two different fluorescent intermediates along the on-switching pathway of Padron, showing first the isomerization from trans- to cis-configuration at 100 K and a subsequent relaxation at 120 K, both in the deprotonated state. An increase in temperature to 180 K eventually led to a protonation equilibrium. Recently the short-time dynamics of Padron was investigated by Fron et al. 20 who found that the excited cis-neutral form decays with a time constant of 1 ps presumably associated with ESPT leading to an unstable intermediate, tentatively assigned to the excited deprotonated cis state which isomerizes with a time constant of 14.5 ps to the trans state. This model does not account for the intense fluorescence observed after excitation of the neutral cis-form as shown in the fluorescence excitation spectrum of Ref. [12]. Excitation of the deprotonated trans form was proposed to be followed by a quantitative isomerization to an unrelaxed cis ground state with a time constant of 5.2 ps and subsequent stabilization to the relaxed cis isomer within 0.7 ps. However, this quantitative isomerization from the trans- to the cis-form is inconsistent with the switching cycles presented by Brakemann et al.

12

which suggest a significantly lower trans to

cis quantum yield. In this paper using femtosecond pump-probe spectroscopy we complement these former measurements and present a comprehensive study of the short-time dynamics of Padron0.9's ground and exited states. To this end all possible configurations of cis / trans and protonated / deprotonated chromophors were investigated by shifting the photo-stationary equilibrium to the


Page 5 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


desired isomer, choosing a suitable pH, and exciting selectively the desired chromophore choosing the proper wavelength. By using both spectral and kinetic modelling, we propose an as simple as possible model for the complex dynamics and involved states taking into account all available data. From this analysis we derive species spectra and the dominant relaxation pathways. As a major result we prove ESPT to be the initial reaction step after excitation of the neutral cis-chromophore. The isomerization of cis and trans Padron0.9 cannot be observed directly by our time-resolved method because both switching quantum yields are too low.

2

EXPERIMENTAL

2.1

Sample Preparation.

Samples were prepared from concentrated 5 mM solutions of Padron0.9 in PBS buffer. For experiments at pH 4 and pH 10 an aliquot of 50 μl was diluted in 6 ml 5 mM aqueous citrate buffer and phosphate buffer, respectively. Deuterated samples at pH 10 were prepared by dissolving disodium hydrogen phosphate in D2O. An aliquot of 50 μl concentrated Padron0.9 solution was diluted in 6 ml of this 5 mM buffer solution leading to a deuteration grade of >98%. The samples were stored overnight before usage. The solution was pumped in a closed loop through a flow cell with a cross section of 2 mm x 2 mm equipped with two CaF2 windows. To ensure a fresh sample in the 150 μm focus of each pump laser pulse we adjusted the flow rate to 1 ml/s. Part of the loop was an illumination chamber where the solution was irradiated by a collimated high power 1 W LED at 400 nm or 495 nm to shift the photo-stationary equilibrium to either the trans- or the cis-form. LED irradiation started half an hour before the actual measurement and remained on during the whole



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 36

experiment. Under our experimental conditions the optical density at the excitation wavelength of 387 nm for the cis (trans) solution was 0.17 (0.12) at pH 4 and 0.1 at pH 10. Excitation at 505 nm was applied for solutions at pH 10 where the optical density was 0.18 (0.26). 2.2

Laser System.

The laser system is based on a 1 kHz Ti:sapphire oscillator/regenerative amplifier producing 150 fs pulses at 773 nm with pulse energies of 1 mJ. This output beam was split to generate pump and probe pulses. Pump pulses to photoexcite Padron0.9 at 387 nm and 505 nm, respectively, were either generated by second-harmonic-generation (SHG) or by non-collinear optical parametric amplification (NOPA). These pulses were focussed to a diameter of 150 μm to excite the Padron0.9 solution with pulse energies of < 200 nJ. This intensity was sufficiently low to prevent multiphoton processes as was proven in separate experiments with two times higher pump intensities which showed no change in the kinetics. The probe pulse was a white-lightcontinuum generated by focussing a small portion of the 773 nm pulse into a CaF2 crystal of 4 mm length. This continuum was then split into a reference and a probe beam. The latter was superimposed with the pump beam at the sample. The relative plane of polarization was set to 54.7° to eliminate rotational effects. Reference and probe continuum were each detected with a 256 element linear diode array attached to a spectrograph. The measured transient spectra were afterwards corrected with respect to the shift of temporal overlap of pump and probe pulse due to the group delay dispersion within the white-light-probe-continuum. Also, spontaneous emission and stray light signal caught by the array were subtracted. 2.3

Kinetic and Spectral Modelling.

Our approach was to use an as simple as possible kinetic model to describe the time evolution of our spectra. Since the contribution of one species to a spectrum at a given wavelength is the


Page 7 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


product of concentration and extinction coefficient such a fitting procedure can easily be arbitrary. To avoid that it is necessary to reduce the number of degrees of freedom. This was done in several steps in which species spectra were determined and represented by sums of Gaussian functions. Once determined, the spectrum of a species was left unchanged when proceeding to the next step of fitting. The analysis started by modelling the ground state absorption spectra of the protonated and deprotonated cis and trans configurations as well as the stimulated emission spectrum. Using these spectra we modeled the time dependent spectral evolution, which we analysed in the order of increasing complexity. To denote the chromophore states we will use C for the cis and T for the trans isomer. The superscripts minus and H indicate the respective protonation state and an asterisk is used for excited states.

3

RESULTS AND DISCUSSION

3.1

Steady State Characterization

Ground State Absorption Spectra. Figs. 1a and b show the pH dependent absorption spectra of cis and trans Padron0.9. All spectra are normalized to the protein absorption at 280 nm (not shown). Each series of spectra is dominated by two absorption bands associated with the neutral (CH at 397 nm and TH at 384 nm) and the deprotonated (C- at 503 nm and T- at 506 nm) chromophore, respectively. The pH dependence of the cis-chromophore (Fig. 1a) clearly indicates a protonation equilibrium centered at around pH 6. Interestingly, at higher pH the deprotonation of the cis-isomer remains incomplete. This behavior is in contrast to most other related RSFPs, where absorption at 400 nm completely vanishes at high pH21, and cannot be explained by a single protonation equilibrium. Instead one has to assume another protonating



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 36

group in the vicinity of the chromophore which changes its proton affinity such that even at pH 10 a CH species can coexist with C-. Brakemann et al.12 had already assumed a close Histidine residue to account for the unusual titration behaviour of Padron0.9. Following the approach of Gayda et al.22 we denote this protonating group with X such that four species, CHXH, CHX-, C-XH, and C-X- form a pH dependent equilibrium.

Figure 1. pH dependent absorption spectra of cis- (a) and trans- (b) Padron0.9. Open circles are measured spectra, solid lines are fits with a model describing the protonation equilibrium of the Padron0.9 chromophore interacting with two protonating amino acids in the protein environment. For cis-Padron0.9 this model is shown in (c) with fitted pK values (corresponding results for trans-Padron0.9 are shown in parenthesis). Species highlighted with equal color have identical absorption spectra. The fractional populations derived from the modelling for cis- and trans-Padron0.9 are shown in (d) and (e), respectively. Solid lines represent the sum of species with identical absorption spectra. For details see the text and Supporting Information.


Page 9 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


However, closer inspection of Fig. 1a reveals two other subtleties: a broadening of the Cspectrum at pH 10 and a blue shift of the protonated species at pH < 5 that for pH 4 peaks at 385 nm. The former effect can be attributed qualitatively to a difference in the absorption spectra of C-XH and C-X- within the four species model. This model, however, cannot explain the blue shift at low pH. At first sight one might argue that at low pH, regardless of LED-irradiation an isomerization to the thermodynamically favourable trans species occurs as the pH 4 spectra of Fig. 1a and b appear very similar. However, this can be ruled out as there are quantitative differences: 1. the absorption coefficients differ significantly (the cis band is 20% stronger than the trans band), and 2. as shown in Fig. 1b there is residual absorption of T- at 506 nm whereas in Fig. 1a there is no indication of a deprotonated species at pH 4. Furthermore, protein solutions buffered at pH 4 and switched to the cis (trans) form by LED irradiation at 500 nm (400 nm) showed after neutralization the respective distinctive cis (trans) spectrum. This clearly proves that the pH 4 spectrum in Fig. 1a (1b) corresponds to the cis (trans) chromophore. Therefore, we attribute the pH 4 and pH 6 spectra of Fig. 1a to protonated cis-species which are subject to a varying electrostatic interaction with yet another protonating group Y of the protein environment, thus forming the reaction scheme presented in Fig. 1c. The equations describing the fractional populations of all eight species are presented in the Supporting Information. In a global analysis we modelled the pH dependent absorption spectra of the cis-species using the scheme of Fig. 1c. The previous discussion showed that at least four spectra are needed to account for spectral features of Fig. 1a. We attribute these to the species CHXHYH, C-X-Y-, and the non-distinguishable CHXHY-/CHX-YH/CHX-Y- and C-XHY-/C-X-YH/C-XHYH, respectively. Each species spectrum was fitted by a sum of Gaussian functions, two for the CH and three for



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 36

the C- bands. The final outcome of the modelling is in satisfactory agreement with experiment as shown in Fig. 1a by full lines. Species spectra (Fig. S1a) and corresponding fitting parameters (Table S1) are presented in the Supporting Information. In Fig. 1d the pH dependencies of the species fractional populations are plotted. One can see that for pH between 4 and 10 there exists a distribution of CH and C- species with different protein environment. In our time-resolved experiments CH and C- can be distinguished by choosing the proper excitation wavelength. Here Fig. 1d shows that the pure protonated species CHX-Y- can be investigated at pH > 7.5 and the pure deprotonated species C-XHY- between pH 7 and 9. In the kinetic analysis of our timeresolved spectra we assume no additional inhomogeneity. This is also supported by the fluorescence excitation spectrum of cis-Padron0.9 shown by Brakemann et al.12 which for both protonation states shows the same shape as the corresponding absorption spectrum. In the trans-isomer pH dependent spectra (Fig. 1b) the presence of protonating groups of the protein environment manifests itself in a more subtle way. The spectra show a protonation equilibrium at pH 4-5. However, the transition appears much sharper than for the cis-isomer and cannot be described by a single equilibrium of the form

⇌

. Furthermore, at pH > 8

a slight change of the T- band shape is observed. To model both effects, again, two protonating groups X and Y of the protein environment have to be introduced resulting in a scheme similar to Fig. 1c, where C has to be replaced by T. Note, that the X and Y groups are not necessarily the same for cis and trans (and if so they can still differ in orientation to the chromophore and in their acidity). For the global analysis of the trans isomer spectra only three species spectra are needed which we attribute to T-X-Y- and the non-distinguishable THXHYH/THXHY-/THX-YH/THXY- and T-XHY-/T-X-YH/T-XHYH, respectively. The agreement between the measured absorption spectra of Fig. 1b and the model fit (full lines) is excellent. The underlying pH dependent


Page 11 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


fractional concentrations are plotted in Fig. 1e, whereas species spectra (Fig. S1b) and corresponding fitting parameters (Table S1) are presented in the Supporting Information. Fig. 1e shows that THXHYH at low and T-X-Y- at high pH can be investigated as pure species. In between the T- species appears as a mixture consisting, however, mainly of T-XHY-. On the basis of Figs. 1d and 1e we have chosen to perform time-resolved experiments for cis and trans-Padron0.9 at pH-values of 4, 7 and 10. Stimulated Emission Spectrum of C–. After excitation by light, Padron0.9 shows green fluorescence. The intensity strongly depends on pH, chromophore configuration, and excitation wavelength. The shape of the emitted spectrum, however, does not change significantly with these parameters. Maximal fluorescence occurs after irradiating the cis-configuration in high pH buffer solution at 500 nm. For later analysis of the transient absorption spectra the stimulated emission spectrum of Padron0.9 was calculated from the stationary fluorescence spectrum23. It can be described by the sum of three Gaussian functions (see Fig. S1a and Table S1 in the Supporting Information) and is further referred to as 3.2

∗

.

Excited States Characterization

Deprotonated states C– and T–. Excited state dynamics of C- and T- was investigated with pump-probe spectroscopy at pH 10 and 7 using 505 nm excitation. Fig. 1d and e show that at these pH-values different deprotonated species exist. The analysis of the pH dependent absorption spectra indicates spectral differences between C-X-Y- and C-X-YH/C-XHY- as well as T-X-Y- and T-X-YH/T-XHY-. However, a comparison of our time-resolved signals at pH7 and 10 for both cis and trans shows no significant differences. For the cis isomer this result is not too surprising, as the fractional concentration of C-X-Y- even at pH10 is low. In the case of trans Padron0.9, however, the pH change from 7 to 10 involves a complete transition from mainly T-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 36

XHY- to pure T-X-Y-. As we find no difference in the transient signals for these samples we conclude that in kinetic terms these species behave identically.

Figure 2. Transient absorption spectra of Padron0.9 at pH 10 generated by 505 nm pump pulse excitation of the deprotonated chromophore with continuous LED illumination at 495 nm to obtain the cis-isomer,

, (a) and subsequently at 400 nm to obtain the trans-isomer,

, (b).

Open circles are measured spectra, solid lines were generated from the kinetic models and species spectra of Fig. 3 and 6 (see text). The yellow spectrum in (b) is the scaled 190 ps transient of (a). For each wavelength, data can be fitted by bi and tri-exponential decays, respectively. The corresponding amplitude spectra and time constants for (a) and (b) are shown in (c) and (d), respectively.


Page 13 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In the following we present experimental results obtained at pH 10. The measurements were performed for one and the same solution by switching first to the cis (Fig. 2a) and after that to the trans isomer (Fig. 2b) using LED illumination first at 495 and subsequently at 400 nm. The transient spectra of both isomers show regions which are dominated by excited state absorption (350-450 nm), bleaching of the ground state (450-520 nm), and stimulated emission (>520 nm). As expected for the bright state, excitation of C- leads to a long lived (>1 ns) excited state species C-* whereas for the dark state the corresponding T-* decay is much faster. However, Fig. 2b also shows a long-lived transient with identical spectral features as the cis-transient of Fig. 2a, which is illustrated by superimposing the scaled 193 ps spectrum of Fig. 2a on Fig. 2b (yellow line). For each wavelength the time evolution of the transients in Fig. 2a can be fitted by a biexponential decay with time constants

5.3

0.5

and

signal of the trans species (Fig. 2b) requires three time constants 6.5

0.7

, and

1.2

0.6

1.2

0.3 1.1

. Fitting the 0.2

,

. The wavelength dependent amplitudes are presented in

Fig. 2c and d. We attribute the dominating decays to the spectral evolution of excited C- and T-, respectively. The spectrum of the 1.2 ns component in Figs.2c appears as a minor contribution in Fig. 2d. Hence, we assign it to a contaminations of the undesired cis-isomer due to incomplete photo switching by LED-irradiation. On the other hand the 6.5 ps component dominating in Fig. 2d appears to be very similar to the 5.3 ps component in Fig. 2c. This again can be due to incomplete switching. However, for the cis-signal one also has to take into account the ground state protonation equilibrium

⇄

, i.e. once the

excitation pulse it is repopulated by the reaction

→

ground state is bleached by the . This would result in a similar spectral

evolution as represented by the 5.3 ps component in Fig. 2c. We will show below that both a



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 36

trans-contamination and the ground state equilibration are responsible for the fast spectral evolution in Fig. 2a. The decay of the photo-excited deprotonated cis isomer appears mono-exponential with a time constant of

∗

1.2

0.3

. Fron et al.

20

investigated the fluorescence emission of

Padron0.9 with time-correlated single photon counting and found three decay times of 3.6 ns, 500 ps, and 39 ps, where the 3.6 ns component was attributed to the lifetime of the deprotonated cis state C-* and the two faster time constants to relaxation processes within the excited state of the chomophore. Our transients do not show a 39 ps component. Since our observation time window is limited to 200 ps the 1.2 ns component might be composed of the 500 ps and 3.6 ns components found in Ref. [20]. In contrast to the spectrum of C- in Fig. 2a the T- stimulated emission in Fig. 2b is broad and featureless. Furthermore, the stimulated emission can be described almost entirely by the slow 6.5 ps component whereas for the bleach and excited state region also the fast 1.1 ps component is required. This result is consistent with Fron et al.20 who also observed a fast 0.7 ps and a slower 5.2 ps component. Since the 0.7 ps component did not appear in the fluorescence upconversion experiment they suggested the formation of a hot ground state or another excited state. Note, that in Fig. 2d the largest amplitude of the 1.1 ps component appears at the red edge of the T– absorption spectrum at 520 nm which would be consistent with fast recovery of the hot ground state Th by internal conversion. As stimulated emission and exited state absorption last considerably longer and the latter decays bi-exponentially it is likely that the initially excited trans-species T-* rapidly relaxes to a state Tr-* from which internal conversion to the hot ground state occurs at a slower rate as compared to T-*. The reaction scheme summarizing these considerations is presented in Fig. 3a. Assuming that in the reaction cascade from the relaxed


Page 15 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


excited trans state Tr-* towards the equilibrated ground state T- the depopulation of Tr-* is rate limiting, the states Tr,h- and Tr- will not be populated significantly. Hence only the coloured states and kinetic processes highlighted by thick arrows in Fig. 3a have to be considered. Note that the representation of vibrational cooling by just two states is a gross simplification which however significantly reduces the number of adjustable parameters in the model and leads to satisfactory results. Since the exited state relaxation process is accompanied by only minor spectral changes and the ground state T- is repopulated completely after 20 ps major structural modifications of the chromophore including ESPT can be excluded. Rather we attribute the relaxation to a combination of chromophore motion and reorganization of the local protein environment, stabilizing the chromophore and thus slowing down the rate of internal conversion to the ground state. At the same time, we do not observe any formation of the cis-isomer after excitation of T-. This observation is in contrast to the reaction scheme proposed in Ref. [20] which suggests a fast and quantitative trans to cis isomerization when T- is excited at 500 nm, but completely agrees with the low trans  cis isomerization yield of Padron0.9 suggested by Brakemann’s switching experiments12. Φ

→

500

In

separate

experiments

0.0003

we

determined

the

quantum

yield

to

be

0.0001. Due to this low yield we could not observe the

switching process directly. In a global analysis we fitted the transient spectra of

in Fig. 2b according to the model

presented in Fig. 3a with a small contamination of the cis isomer decaying with a time constant of 1.2 ns. In the procedure we used the previously determined pH 10 ground state spectrum of . The exited state spectra

∗

and

∗

were each described by three Gaussians. The fitted rate

constants are included in Fig. 3a, species spectra are presented in Fig. 3b. Parameters for



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 36

modelling these spectra by sums of Gaussian functions are summarized in Table S2 in the Supporting Information.

Figure 3. (a) Kinetic model describing the dynamics of photo-excited

at pH 10. Spectra of

appearing species are shown in (b).


Page 17 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Protonated State CH at high pH: Excited and ground state proton transfer. The excited state dynamics of CH at pH 10 was investigated using an excitation wavelength of 387 nm. Under these conditions only one species, namely CHX-Y- is excited (see Figs. 1a and d). Since at pH 10 the trans product is completely deprotonated it cannot contribute to a transient signal even if back switching with the 495 nm LED was incomplete.

Figure 4. (a) Transient absorption spectra of Padron0.9 at pH 10 generated by 387 nm pump pulse excitation of the protonated cis-chromophore

with continuous LED illumination at 495 nm.

Open circles are measured spectra, solid lines were generated from the kinetic model and species spectra of Fig. 6 (see text). For each wavelength, transients can be fitted by tri-exponential decays. The corresponding amplitude spectra and time constants are shown in (b). Corresponding results for the deuterated species

are presented in (c) and (d).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 36

Fig 4a shows that 0.5 ps after excitation of CH at pH 10 the transient spectrum consists of ground state bleach at 400 nm superimposed by excited state absorption with a maximum at 445 nm. At wavelengths > 510 nm the stimulated emission evolves within picoseconds from a broad, featureless to a narrow spectrum with one pronounced peak at 515 nm and a shoulder at 550 nm. As shown in Fig. 5 the stationary fluorescence spectra of CH and C- at pH10 following 395 nm and 503 nm excitation, respectively, are identical with a distinctly lower fluorescence yield for CH. A comparison of Figs. 4a and 2a also indicates striking similarities of the stimulated emission between the transient spectra at pump-probe delays >100 ps which is clarified by presenting 126 ps transients of both data sets in Fig. 5 as well. The two spectra are scaled to ∗

match the red edge of the stimulated emission spectrum

which is shown for comparison.

These observations reveal that the emitting species in both cases is the same and can safely be ∗

assigned to the deprotonated excited state

. Hence, we attribute the overall spectral evolution ∗

in Fig. 4a to an exited state proton transfer

→

exponential decays reveals three time constants:

∗

. Fitting of the transients in Fig. 4a by tri1.60.3 ps ,

265 ps, and

1.2 ns (the latter was fixed and corresponds to the excited state lifetime of amplitude spectra for these decays are shown in Fig. 4b.

and

evolution of the stimulated emission suggesting an intermediate state

∗

). The

appear in the spectral ∗

→

∗

→

∗

. The

1.6 ps component is also prominent in the decay of the ground state bleach. Hence, we attribute the depopulation of the initially excited

∗

state to two competing pathways, namely internal

conversion to the ground state and ESPT. A 1 ps excited state proton transfer to an intermediate deprotonated cis-state competing with internal conversion was also suggested in Ref.[20].


Page 19 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Transient absorption spectra of cis-Padron0.9 measured 126 ps after

(green) and

excitation (brown), respectively; both spectra are scaled to the stimulated emission spectrum of

∗

(orange). Fluorescence spectra generated with 503 nm (dashed line) and 395 nm (dotted

line) excitation are shown for comparison. Blue circles represent the difference between the and

transient spectrum. The red curve is the scaled difference of the

and

ground state

absorption spectra.

Fig. 5 shows that the strong negative signal at 500 nm of both 126 ps transients cannot be entirely explained by the stimulated emission of

∗

. When

by ground state bleach of the parent species, however, for

is excited this is easily explained excitation this is unexpected as the

corresponding bleach appears at 397 nm (see Fig. 1a). This discrepancy could be explained by a broader emission spectrum of the ESPT product as compared to the similarity of the stationary Hence, the depopulation of

and

∗

state. However, the

fluorescence spectra (see Fig. 5) precludes this conjecture.

transient must also have a

bleach component, which we attribute to a

in a ground state proton transfer reaction re-establishing the equilibrium



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

⇄

Page 20 of 36

after being disturbed by the pump pulse. The data of Fig. 4a suggest that this ground

state relaxation kinetics happens on a similar timescale as ESPT. Moreover, it must also be operative when

is excited, i.e. the 5.3 ps contribution shown in Fig. 2c can partially be

attributed to the reaction

→

reducing the initial bleach of

at 500 nm as well as

absorption at 400 nm. is known to induce photo-switching from the cis to trans configuration12. In

Excitation of

separate experiments we determined the quantum yield to be Φ

→

400

0.02

0.01. Due to this low value the formation of trans will be hardly detectable in the transients even if this product is formed within the time window of our experiment. Hence, assume for a while and

that the only photochemical reactions induced by exciting conversion and/or ground state equilibration

⇄

are ESPT, internal

, respectively. Then after these obviously

fast processes are over, i.e. at pump-probe delays >100 ps the system reaches a quasi-stationary state in which during the lifetime of

∗

the relative populations of

,

and

∗

stay

constant (for details see Supporting Information). Moreover, in this quasi-stationary phase it does not matter whether

or

is initially excited, the transient signals are identical apart from a

scaling factor. As shown in Fig. 5 for the two 126 ps transients this is actually not the case. When is excited there is less bleach (negative absorption) at 430-530 nm as compared to excitation, at lower wavelength the reverse is true. Plotting the difference of both transient spectra (i.e.

-transient minus

-transient) turns out to be very instructive. As shown by the

blue points in Fig. 5 the resulting spectrum almost coincides with the properly scaled difference of the ground state absorption spectra of

and

(red line). This is indicative of a more

complex ground state protonation equilibrium between

and

in which only a sub-

population equilibrates rapidly whereas another fraction on the timescale of our experiment does


Page 21 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


not. An inhomogeneous ground state distribution with two varieties of the deprotonated cis and

species

as shown in Fig. 6a can explain the observed behavior (for details see and

Supporting Information) if one equilibration, e.g. between 100 ps whereas for

and

is fast, i.e. settles within

it is slow, such that changes in the concentration of

induced

by the pump pulse remain unchanged for a long time. The already identified excited states and

∗

∗

are included in Fig. 6a as well. In accord with the two ground states we have to assume ∗

corresponding excited states

and

∗

.

To verify the assumption of ESPT, measurements were repeated with deuterated Padron0.9 using buffered D2O solutions. The results presented in Fig. 4c show obvious differences to Fig. 4a caused by an H/D isotope effect. A closer analysis reveals an almost identical ratio of ground state bleach (400 nm) to exited state absorption (445 nm) shortly after excitation. Fitting the spectral evolution exponentially with time constants of

∗

1.2 ns fixed to the

3.00.5 ps and

lifetime yields two additional

103 ps with amplitudes plotted in Fig. 4d.

These decay spectra indicate that the build-up of stimulated emission and refill of ground state bleach happen concertedly. However,

is twice as long as for the undeuterated sample. These

observations support our interpretation of two reaction channels, internal conversion and ESPT, depopulating the excited

∗

state. Since upon deuteration ESPT is significantly slowed down

internal conversion is favored and the yield of stimulated emission from

∗

is reduced.

Using a global analysis we simultaneously fitted the spectral changes after excitation of Fig. 2a and excitation of

in

in Figs 4a and c within the kinetic model of Fig. 6a. As discussed

we used two deprotonated ground state species

and

with corresponding excited states

which we assumed to be spectrally equivalent, respectively. The already identified spectra of and

(Fig. 1a and S1), the fluorescence lifetime 1⁄

1.2

, and the overall protonation



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 36

⁄

equilibrium constant at pH 10,

0.7 (Fig. 1d), were

used as fixed input. The rate constants defining the ground state equilibrium described by ⁄

were set to zero, whereas

was an adjustable parameter. Also, we assumed that a

fraction of the signal in Fig. 2a originates from contamination by the trans-isomer described by the established model and spectra of Fig. 3. The result of this fitting presented in Figs. 2a, 4a and 4c by solid lines is in excellent agreement with the experimental data. Species spectra of ∗

and

∗

∗

,

obtained from the fitting procedure are shown in Fig.6b. Parameters for modelling

these spectra are summarized in Table S3 in the Supporting Information. The deprotonated state ∗

shows stimulated emission which is less structured and broader than the emission of

∗

suggesting that the intermediate state after fast ESPT is further stabilized by rearrangements of the protein environment. The rate constants obtained from the fitting are included in Fig. 6a. The ESPT and the ground state equilibration happen on a ps timescale and show strong H/D isotope effects of 5.2 for ESPT and of

to

the two

,

⁄

⁄

3.5 for the ground state proton transfer. The concentration ration

,

⁄

at equilibrium is determined to be

0.94. One could argue that

species of the model correspond to C-XHY- and C-X-Y- of Fig. 1. However, the

kinetics we measured at pH7 was similar to pH10 although the species C-X-Y- does not appear at pH < 9. This indicates that the system apart from the complex protonation equilibrium of Fig. 1c is characterized by additional inhomogeneity in Establishing this inhomogeneity for

one could also assume it for

dynamics of Fig. 6a by a sum of processes cascade

∗

→

∗

→

∗

.

∗

→

∗

and

∗

→

∗

and

∗

instead of just one

1.60.3 ps and

. Then the two time constants

265 ps could be associated with the lifetimes of

to explain the ESPT

∗

, respectively. However,


Page 23 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


analyzing the amplitude spectra in Fig. 4b the spectra of different. Accordingly the ground state spectra of

and

∗

and

∗

would have to be very

can be expected to be different,

too. The fluorescence excitation spectrum of cis-Padron0.9 shown by Brakemann et al.12 for the protonated state shows the same shape as the corresponding absorption spectrum. If one assumes this band to consist of two distinct absorption spectra this observation can be only explained if the ESPT quantum yield for both between

and

∗

and

∗

are the same. In view of the large difference

this is unlikely because ESPT is competing with efficient internal conversion.

Figure 6 (a) Kinetic model describing the primary reaction steps of photo-excited

/

at pH 10.

Spectra of appearing species are shown in (b).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 36

Protonated States at pH 4: The excited state dynamics of CH and TH at pH 4 were investigated using an excitation wavelength of 387 nm. Fig. 1d and e show that at pH 4 the prevailing species are CHXHYH and THXHYH, respectively. Interestingly, the time-resolved signals for both appear very similar. They only differ in amplitude (the signal for the cis is larger than for the trans chromophore by a factor of 1.3) but not notably in shape. Note, that both experiments were performed directly one after another by switching from 490 nm to 400 nm LED illumination and without any variation of the pump-probe laser setup. Fig. 7 shows the results for the cis chromophore.

Figure 7. (a) Transient absorption spectra of Padron0.9 at pH 4 generated by 387 nm pump pulse excitation of the protonated cis-chromophore

with continuous LED illumination at 495 nm.

For each wavelength, data can be fitted by tri-exponential decays. The corresponding amplitude spectra and time constants are shown in (b).

At first, the spectrum consists of a ground state bleach at < 400 nm superimposed by excited state absorption with a maximum at 440 nm and broad, featureless stimulated emission at wavelengths > 480 nm. For each wavelength the time evolution of the transients can be fitted by


Page 25 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a tri-exponential decay with time constants of 240

and

50

0.9

0.1

8.7

,

1.5

. Their corresponding amplitudes are shown in Fig. 7b. The spectra

indicate that all the population relaxes back to the ground state. Also there is no hint of an excited state proton transfer which would be evidenced by the narrow stimulated emission band of

∗

at 515 nm like in Fig. 4a. Therefore, we attribute the observed dynamics to a cascade of

minor structural changes of the chromophore/protein environment which finally ends in the CH ground state. Alternatively, the multi-exponential dynamics could be explained by an inhomogeneous ground state distribution which upon excitation leads to a superposition of exited states with different lifetimes.

4

CONCLUSION In this work we present extensive measurements on the light-induced primary reaction steps of

the chromophore and the surrounding protein of Padron0.9. As verified by a strong H/D isotope effect excitation of the neutral cis chromophore

is followed by an ESPT competing with

internal conversion to the ground state. The initially populated ESPT state fluorescent state

∗

which is directly accessible by excitation of

∗

relaxes into the

. Excitation of

and

in

both cases is followed by a ground state proton transfer reaction partially re-establishing the perturbed ground state equilibrium or

⇄

. The level of equilibration depends on whether

is excited and points to an inhomogeneous distribution of

species which equilibrate on

different timescales. The isomerization to the trans form is too inefficient (2%) to be detectable with our method. The reverse trans-cis switching process is even less efficient as excitation of the deprotonated trans species is dominated by internal conversion to a hot ground state

on a



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ps timescale. This competes with excited state relaxation (

∗

→

Page 26 of 36

∗

), tentatively assigned to a

combination of chromophore motion and reorganization of the local protein environment.

ASSOCIATED CONTENT Supporting Information. Protonation equilibria, absorption and emission spectra, fitting parameters for species spectra, and the results of time-resolved pump-dump-probe measurements are presented. This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Notes The authors declare no competing financial interest.

ACKNOWLEDGEMENT We thank Stefan Hell for helpful discussions and continuous support of this work. We also acknowledge financial support by the Deutsche Forschungsgemeinschaft Research Center Nanoscale Microscopy and Molecular Physiology of the Brain.


Page 27 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


REFERENCES

(1) Bourgeois, D.; Adam, V.: Reversible Photoswitching in Fluorescent Proteins: A Mechanistic View. IUBMB Life 2012, 64, 482-491. (2) Andresen, M.; Stiel, A. C.; Fölling, J.; Wenzel, D.; Schönle, A.; Egner, A.; Eggeling, C.; Hell, S. W.; Jakobs, S.: Photoswitchable Fluorescent Proteins Enable Monochromatic Multilabel Imaging and Dual Color Fluorescence Nanoscopy. Nat. Biotechnol. 2008, 26, 1035-1040. (3) Chudakov, D. M.; Chepurnykh, T. V.; Belousov, V. V.; Lukyanov, S.; Lukyanov, K. A.: Fast and Precise Protein Tracking Using Repeated Reversible Photoactivation. Traffic 2006, 7, 1304-1310. (4) Adam, V.; Mizuno, H.; Grichine, A.; Hotta, J.; Yamagata, Y.; Moeyaert, B.; Nienhaus, G. U.; Miyawaki, A.; Bourgeois, D.; Hofkens, J.: Data Storage Based on Photochromic and Photoconvertible Fluorescent Proteins. J. Biotechnol. 2010, 149, 289-298. (5) Grotjohann, T.; Testa, I.; Leutenegger, M.; Bock, H.; Urban, N. T.; LavoieCardinal, F.; Willig, K. I.; Eggeling, C.; Jakobs, S.; Hell, S. W.: Diffraction-unlimited All-optical Imaging and Writing with a Photochromic GFP. Nature 2011, 478, 204-208. (6) Brakemann, T.; Stiel, A. C.; Weber, G.; Andresen, M.; Testa, I.; Grotjohann, T.; Leutenegger, M.; Plessmann, U.; Urlaub, H.; Eggeling, C.; et al.: A Reversibly Photoswitchable GFP-like Protein with Fluorescence Excitation Decoupled from Switching. Nat. Biotechnol. 2011, 29, 942-947. (7) Hofmann, M.; Eggeling, C.; Jakobs, S.; Hell, S. W.: Breaking the Diffraction Barrier in Fluorescence Microscopy at Low Light Intensities by Using Reversibly Photoswitchable Proteins. Proc. Natl. Acad. Sci. USA 2005, 102, 17565-17569. (8) Andresen, M.; Wahl, M. C.; Stiel, A.; Gräter, F.; Schäfer, L. V.; Trowitzsch, S.; Weber, G.; Eggeling, C.; Grubmüller, H.; Hell, S. W.; et al.: Structure and Mechanism of the Reversible Photoswitch of a Fluorescent Protein. Proc. Natl. Acad. Sci. USA 2005, 102, 1307013074. (9) Andresen, M.; Stiel, A. C.; Trowitzsch, S.; Weber, G.; Eggeling, C.; Wahl, M. C.; Hell, S. W.; Jakobs, S.: Structural Basis for Reversible Photoswitching in Dronpa. Proc. Natl. Acad. Sci. USA 2007, 104, 13005-13009. (10) Pletneva, N. V.; Pletnev, V. Z.; Shemiakina, I. I.; Chudakov, D. M.; Artemyev, I.; Wlodawer, A.; Dauter, Z.; Pletnev, S.: Crystallographic Study of Red Fluorescent Protein eqFP578 and its Far-red Variant Katushka Reveals Opposite pH-induced Isomerization of Chromophore. Protein Sci. 2011, 20, 1265-1274. (11) Mizuno, H.; Mal, T. K.; Wälchli, M.; Kikuchi, A.; Fukano, T.; Ando, R.; Jeyakanthan, J.; Taka, J.; Shiro, Y.; Mitsuhiko, I.; et al.: Light-dependent Regulation of Structural Flexibility in a Photochromic Fluorescent Protein. Proc. Natl. Acad. Sci. USA 2008, 105, 9227-9232. (12) Brakemann, T.; Weber, G.; Andresen, M.; Groenhof, G.; Stiel, A. C.; Trowitzsch, S.; Eggeling, C.; Grubmuller, H.; Hell, S. W.; Wahl, M. C.; et al.: Molecular Basis of the Light-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 36

driven Switching of the Photochromic Fluorescent Protein Padron. J. Biol. Chem. 2010, 285, 14603-14609. (13) Mizuno, H.; Mal, T. K.; Wälchli, M.; Fukano, T.; Ikura, M.; Miyawaki, A.: Molecular Basis of Photochromism of a Fluorescent Protein Revealed by Direct 13C Detection Under Laser Illumination. J. Biomol. NMR 2010, 48, 237-246. (14) Pletnev, S.; Shcherbo, D.; Chudakov, D. M.; Pletneva, N.; Merzlyak, E. M.; Wlodawer, A.; Dauter, Z.; Pletnev, V.: A Crystallographic Study of Bright Far-Red Fluorescent Protein mKate Reveals pH-induced cis-trans Isomerization of the Chromophore. J. Biol. Chem. 2008, 283, 28980-28987. (15) Chattoraj, M.; King, B. A.; Bublitz, G. U.; Boxer, S. G.: Ultra-fast Excited State Dynamics in Green Fluorescent Protein: Multiple States and Proton Transfer. Proc. Natl. Acad. Sci. USA 1996, 93, 8362-8367. (16) Ando, R.: Regulated Fast Nucleocytoplasmic Shuttling Observed by Reversible Protein Highlighting. Science 2004, 306, 1370-1373. (17) Fron, E.; Flors, C.; Schweitzer, G.; Habuchi, S.; Mizuno, H.; Ando, R.; De Schryver, F. C.; Miyawaki, A.; Hofkens, J.: Ultrafast Excited-State Dynamics of the Photoswitchable Protein Dronpa. J. Am. Chem. Soc. 2007, 129, 4870-4871. (18) Warren, M. M.; Kaucikas, M.; Fitzpatrick, A.; Champion, P.; Timothy Sage, J.; van Thor, J. J.: Ground-state Proton Transfer in the Photoswitching Reactions of the Fluorescent Protein Dronpa. Nature Communications 2013, 4, 1461. (19) Faro, A. R.; Carpentier, P.; Jonasson, G.; Pompidor, G.; Arcizet, D.; Demachy, I.; Bourgeois, D.: Low-Temperature Chromophore Isomerization Reveals the Photoswitching Mechanism of the Fluorescent Protein Padron. J. Am. Chem. Soc. 2011, 133, 16362-16365. (20) Fron, E.; Van der Auweraer, M.; Hofkens, J.; Dedecker, P.: Excited State Dynamics of Photoswitchable Fluorescent Protein Padron. J. Phys. Chem. B 2013, 117, 1642216427. (21) Habuchi, S.; Ando, R.; Dedecker, P.; Verheijen, W.; Mizuno, H.; Miyawaki, A.; Hofkens, J.: Reversible Single-molecule Photoswitching in the GFP-like Fluorescent Protein Dronpa. Proc. Natl. Acad. Sci. USA 2005, 102, 9511-9516. (22) Gayda, S.; Nienhaus, K.; Nienhaus, G. U.: Mechanistic Insights into Reversible Photoactivation in Proteins of the GFP Family. Biophys. J. 2012, 103, 2521-2531. (23) Schäfer, F. P.: Dye Lasers; 2nd ed.; Springer-Verlag: Berlin Heidelberg, 1977; Vol. 1.


Page 29 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents-Graphic



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

pH dependent absorption spectra of cis- (a) and trans- (b) Padron0.9. Open circles are measured spectra, solid lines are fits with a model describing the protonation equilibrium of the Padron0.9 chromophore interacting with two protonating amino acids in the protein environment. For cis-Padron0.9 this model is shown in (c) with fitted pK values (corresponding results for trans-Padron0.9 are shown in parenthesis). Species highlighted with equal color have identical absorption spectra. The fractional populations derived from the modelling for cis- and trans-Padron0.9 are shown in (d) and (e), respectively. Solid lines represent the sum of species with identical absorption spectra. For details see the text and Supporting Information. 288x201mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 30 of 36

Page 31 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Transient absorption spectra of Padron0.9 at pH 10 generated by 505 nm pump pulse excitation of the deprotonated chromophore with continuous LED illumination at 495 nm to obtain the cis-isomer, C-, (a) and subsequently at 400 nm to obtain the trans-isomer, T-, (b). Open circles are measured spectra, solid lines were generated from the kinetic models and species spectra of Fig. 3 and 6 (see text). The yellow spectrum in (b) is the scaled 190 ps transient of (a). For each wavelength, data can be fitted by bi and tri-exponential decays, respectively. The corresponding amplitude spectra and time constants for (a) and (b) are shown in (c) and (d), respectively. 288x201mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(a) Kinetic model describing the dynamics of photo-excited T- at pH 10. Spectra of appearing species are shown in (b). 139x201mm (300 x 300 DPI)


Page 32 of 36

Page 33 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a) Transient absorption spectra of Padron0.9 at pH 10 generated by 387 nm pump pulse excitation of the protonated cis-chromophore CH with continuous LED illumination at 495 nm. Open circles are measured spectra, solid lines were generated from the kinetic model and species spectra of Fig. 6 (see text). For each wavelength, transients can be fitted by tri-exponential decays. The corresponding amplitude spectra and time constants are shown in (b). Corresponding results for the deuterated species CD are presented in (c) and (d). 284x199mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Transient absorption spectra of cis-Padron0.9 measured 126 ps after C- (green) and CH excitation (brown), respectively; both spectra are scaled to the stimulated emission spectrum of C-* (orange). Fluorescence spectra generated with 503 nm (dashed line) and 395 nm (dotted line) excitation are shown for comparison. Blue circles represent the difference between the CH and C- transient spectrum. The red curve is the scaled difference of the C- and CH ground state absorption spectra. 287x201mm (300 x 300 DPI)


Page 34 of 36

Page 35 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a) Kinetic model describing the primary reaction steps of photo-excited CH/D at pH 10. Spectra of appearing species are shown in (b). 149x199mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47


Page 36 of 36

Primary Light-Induced Reaction Steps of Reversibly Photoswitchable

Recommend Documents