Redox-Based Photostabilizing Agents in Fluorescence Imaging: The

Aug 30, 2017 - Article Options. ACS ActiveView PDF. Hi-Res Print, Annotate, Reference QuickView. PDF (1548 KB) · PDF w/ Links (452 KB) · Full Text HTM...
0 downloads 9 Views 2MB Size
Article pubs.acs.org/JACS

Redox-Based Photostabilizing Agents in Fluorescence Imaging: The Hidden Role of Intersystem Crossing in Geminate Radical Ion Pairs Viktorija Glembockyte and Gonzalo Cosa* Department of Chemistry and Center for Self-Assembled Chemical Structures, McGill University, 801 Sherbrooke Street West, Montréal, Québec H3A 0B8, Canada S Supporting Information *

ABSTRACT: Here we report transient absorption studies on the ground-state recovery dynamics of the single-molecule fluorophore Cy3B in the presence of four different photostabilizing agents, namely β-mercaptoethanol (β-ME), Trolox (TX), n-propyl gallate (n-PG), and ascorbic acid (AA). These are triplet-state quenchers that operate via photoinduced electron transfer (PeT). While quantitative geminate recombination was recorded following PeT for β-ME (∼100%), for Trolox, n-propyl gallate, and ascorbic acid the extent of geminate recombination was >48%, >27%, and >13%, respectively. The results are rationalized in terms of the rates of intersystem crossing (ISC) in the newly formed geminate radical ion pairs (GRIPs). Rapid spin relaxation in the radicals formed accounts for quantitative geminate recombination with β-ME and efficient geminate recombination with TX. Our results illustrate how the interplay of PeT quenching efficiency and geminate recombination dynamics may lead to improved photostabilization strategies, critical for single-molecule fluorescence and superresolution imaging.



INTRODUCTION Photobleaching and transitions to dark transient states (blinking) are limiting factors in many advanced fluorescence imaging applications relying on small-molecule organic fluorophores.1,2 Strategies have thus been developed to optimize the performance of these fluorophores that rely on the quenching of dark and potentially reactive fluorophore intermediates.1−12 In this context, cocktails bearing reducing agents such as β-mercaptoethanol (β-ME), ascorbic acid (AA), Trolox (TX), glutathione (GSH), and n-propyl gallate (n-PG) have been utilized toward photostabilizing fluorophores via photoinduced electron transfer (PeT)-mediated quenching of the excited triplet state.1,5,6,10,13 Here, the reducing agent serves as a sacrificial electron donor to yield a one-electron-reduced fluorophore and an oxidized counterpart (Scheme 1). Due to spin conservation rules, PeT quenching of the excited triplet state generates a geminate radical ion pair (GRIP) in the triplet manifold. Back electron transfer (geminate recombination) in the triplet GRIPrestoring the ground states of the fluorophore and e-donorwhile highly feasible from an energetic standpoint, is a spin-forbidden process, and thus in non-viscous environments it is typically amply outcompeted by radical escape from the solvent cage.14 The net result is efficient escape of newly formed intermediates from the solvent cage as free radicals. To counter the detrimental effect of radical escape from a GRIP, reducing oxidizing systems (ROXS) were realized.9 ROXS utilize oxidizing agents such as methylviologen (MV) or Trolox quinone (TQ). The latter are necessary components to © 2017 American Chemical Society

Scheme 1. Photophysical and Photochemical Pathways That Are Exploited To Quench Triplet Excited States of Fluorophores (3*F) toward Enhanced Photostability and Improved Fluorescence Signal in SMF Imaginga

a

Energy-transfer processes (physical, kET) with energy acceptors (Q) and PeT (chemical, kPeT) with redox-active quenchers (Red and Ox) are illustrated. In the ROXS approach the fluorophore radical anions/ cations formed upon escape from the GRIP solvent cage are rescued by their counterpart redox agent. ISC favors direct recovery of the fluorophore ground state via geminate recombination (kISC, highlighted in red).

oxidize/rescue the non-emissive and highly reactive fluorophore radical anions formed upon escape from GRIP, thus Received: August 1, 2017 Published: August 30, 2017 13227

DOI: 10.1021/jacs.7b08134 J. Am. Chem. Soc. 2017, 139, 13227−13233

Article

Journal of the American Chemical Society

Scheme 2. (a) Formation of Cy3B in Its Triplet Excited State and (b) Subsequent Triplet-State Quenching by a Reducing Agent via PeTa

a

The process shown in part b results in a GRIP in the triplet manifold which following ISC may lead to geminate recombination and recovery of Cy3B ground state (II), or following the radical escape from GRIP solvent cage may lead to formation of Cy3B•− (I), which may next become oxidized in solution. The transient absorptions of each state were monitored at the wavelengths indicated in blue.

Figure 1. Transient absorption spectra (left panel) and temporal evolution of ΔOD traces recorded at 440, 560, and 640 nm (right panel). LFP studies were done on Ar-equilibrated aqueous solutions of 6 μM Cy3B containing 50 mM KI and 1x PBS buffer pH 7.4 in the absence of reducing agents (a,b) and in the presence of 1 mM AA (c,d), 0.1 mM n-PG (e,f), 2 mM TX/TQ (g,h), and 143 mM β-ME (i,j). Kinetic traces at longer time scales displaying the Cy3B•−decay and the full Cy3B ground-state recovery are provided in Figure S2. Note: the very fast recovery (within the first few ns following the laser pulse) in the ΔOD traces at 560 nm is associated with fluorescence process and PMT response and was not included in our kinetic analysis.

13228

DOI: 10.1021/jacs.7b08134 J. Am. Chem. Soc. 2017, 139, 13227−13233

Article

Journal of the American Chemical Society restoring the fluorophore to its ground state. In the presence of ROXS the fluorophore triplet state can also be quenched via PeT with the oxidizing agent, where the resulting fluorophore radical cation (F•+) can be next rescued by the reducing agent. Efficient intersystem crossing (ISC) in the triplet GRIP to yield a GRIP in the singlet manifold, on the other hand, would provide an efficient mechanism wherefrom back electron transfer, also referred to as geminate recombination, will readily take place (Scheme 1). Fluorophore photostabilization via geminate recombination not only eliminates the need for an oxidizing counteragent but also minimizes the residence time of the fluorophore as a radical species. Remarkably, the role of ISC within a GRIP and the investigation of factors that steer its efficiency, to our knowledge, have never been explicitly addressed or investigated in the context of fluorophore photostabilization. Typically, ISC in GRIPs is slow and only observed in confining environments, e.g., micellar or other supramolecular systems14−16 where GRIP residence time is lengthened. Interestingly, Tinnefeld and colleagues recently reported single-molecule fluorescence (SMF) studies where efficient (>99% yield) geminate recombination was recorded in a non-viscous environment following photoinduced reduction of a number of dyes in their triplet excited state, in the presence of thiolate ions of β-ME.17 This result sparked our interest in further exploring the mechanism of geminate recombination and optimizing the properties of redox-based triplet-state quenchers to ultimately yield optimized photostabilizing agents. Our results emphasize the importance of efficient PeT as well as efficient ISC in a GRIP to maximize the outcome of fluorophore direct recovery via geminate recombination.

2b, Figure 1c−j).5,9,10,17,25,26 With no additives the initially formed triplet state relaxes back to the ground state with a characteristic long lifetime of 70 μs (see also Table 1). Both the Table 1. Summary of Mechanistic Laser Flash Photolysis Studies on Quenching of Cy3B Triplet Excited State with Different Reducing Agents τ, μs reducing agent/ triplet-state quencher no reducing agent 2 mM AA 0.1 mM n-PG 2 mM TX/TQ 2 mM TX/TQ, 10% glycerolb 143 mM β-ME

estimated % of geminate recombinationa

*Cy3B

Cy3B•−

70 ± 1 9.1 ± 0.2 11.3 ± 0.2 11.2 ± 0.2 10.5 ± 0.1

n/a 612 ± 7 462 ± 5 79 ± 3 31 ± 2

n/a >13 >27 >48 >84

34.8 ± 0.7

86 ± 10

∼100

3

The extent of geminate recombination was determined from fitting the transient traces at 560 nm (recovery of Cy3B) to doubleexponential growth function (see Supporting Information for details). b Even though all samples were equilibrated with Ar for 30 min prior to LFP studies, based on the shorter Cy3B•− lifetime we suggest that the concentration of oxygen was slightly higher under these conditions. a

absorption band at 640 nm and the depletion band in the 480− 580 nm region follow the same kinetics. While only the triplet state is initially expected, closer inspection of the transient absorption spectrum at very short time scales (Figure 1a, blue) and kinetic traces acquired at 440 nm (Figure 1b) revealed a faint transient band at 440 nm formed within the lifetime of the laser pulse. This contribution to ΔOD possibly results from reduction of 1*Cy3B by KI and is considered in our quantitative analysis below. In the presence of reducing agents, 3*Cy3B was quenched via PeT to form a GRIP in the triplet manifold (Scheme 2b). The GRIP then either underwent ISC and geminate recombination leading to direct recovery of Cy3B ground state (pathway II in Scheme 2) or escape from the solvent cage leading to the formation of Cy3B•− (pathway I in Scheme 2). In the former case, 3*Cy3B decay recorded at 640 nm occurred simultaneously with ground-state recovery of Cy3B at 560 nm. Since geminate recombination of geminate radical ion pair in its singlet manifold is a very fast process, no additional intermediate is observed in transient absorption spectra. In contrast, for radical escape, 3*Cy3B decay (640 nm) occurred concomitant with the growth of a new band at 440 nm corresponding to Cy3B•− with no recovery of ground-state Cy3B. The latter is, however, observed to occur at a later stage simultaneously with the decay of the intermediate Cy3B•− species (Figure S2), presumably quenched by traces of O2 in solution (see also discussion in the Supporting Information). When ascorbic acid was used as a 3*Cy3B quencher, following laser excitation, we initially observed the groundstate depletion band with a minimum at 560 nm and a band at 640 nm, corresponding to 3*Cy3B formation (Figure 1c).23 Decay of the latter band occurred with formation of a transient band at 440 nm, which we assigned to Cy3B•− consistent with previous studies on related cyanines.24 The decay of Cy3B•− at 440 nm was observed at much longer times, occurring concomitant with the recovery of ground-state Cy3B at 560 nm (Figure S2). This observation is consistent with pathway I proposed in Scheme 2. Altogether, these results indicate that in the presence of ascorbic acid, PeT with 3*Cy3B is primarily



RESULTS Utilizing time-resolved transient absorption studies here we explore the extent of, and address the mechanism behind, ground-state recovery of the single-molecule fluorophore Cy3B18 (Scheme 2a) from its triplet excited state in the presence of common reducing agents/triplet-state quenchers utilized in SMF and super-resolution fluorescence imaging (specifically AA, n-PG, TX, and β-ME). The cyanine dye Cy3B (the locked form of the most frequently encountered single-molecule dye Cy3) was chosen as a model fluorophore for this study because it cannot undergo photoisomerization18−20 in the excited state. By preventing photoisomerization and the generation of an additional transient cis isomer species (which for non-rigid analogue Cy3 has been shown to absorb near where the triplet excited state absorbs),21,22 data analysis is simplified. To generate Cy3B in its triplet excited state (3*Cy3B), the chromophore was directly excited with the 532 nm output of a Nd:YAG laser in Ar-saturated aqueous solutions in the presence of 50 mM KI, which rapidly (within the laser pulse) catalyzes the ISC from 1 *Cy3B to 3*Cy3B (see Scheme 2a).14,23 Displayed in Figure 1 are typical transient spectra recorded at different times following the laser excitation. The 3*Cy3B transient absorption spectrum was observed in the 600−700 nm range.23 A bleach in the 480−580 nm region in turn corresponded to the depletion of ground-state Cy3B (Figure 1a). The Cy3B radical anion species (Cy3B•−), when generated, were observed as a band centered at ∼440 nm (Figure 1c,e,g).24 The temporal evolution of 3*Cy3B was monitored at 640 nm23 either in the absence of additives (Scheme 2a, Figure 1a,b) or with the different reducing agents/photostabilizers used at concentrations recommended in the literature (Scheme 13229

DOI: 10.1021/jacs.7b08134 J. Am. Chem. Soc. 2017, 139, 13227−13233

Article

Journal of the American Chemical Society

Figure 2. Temporal evolution of the transient absorbance (ΔOD) recorded at 640 nm (3*Cy3B, black) and 440 nm (Cy3B•−, red), adapted from Figure 1 and normalized to the 3*Cy3 absorption: (a) 2 mM AA, (b) 0.1 mM n-PG, (c) 2 mM TX/TQ, (d) 2 mM TX/TQ in a buffer containing 10% v/v glycerol, and (e) 143 mM β-ME. Note: All kinetic traces were first corrected by subtracting the initial ΔOD immediately after laser excitation (0.32 μs) to exclude the contribution from species generated from 1*Cy3B within the laser pulse. Kinetic traces acquired in the presence of TX (c,d) were also corrected to subtract the contribution of the chromanoxyl radical (TX•) that absorbs at 440 nm (see Supporting Information for additional details).

recombination pathway for this triplet-state quencher as previously reported.17 These observations are consistent with PeT quenching followed by ISC in the GRIP and geminate recombination. Importantly, while geminate recombination was fast and efficient, the preceding triplet-state quenching was in turn slow. Here, the lifetime of 3*Cy3B species in the presence of 143 mM β-ME was much longer than in the presence of AA, n-PG or TX/TQ (all at concentrations ≤2 mM, see Figure 1 and Table 1). The lower triplet quenching efficiency observed for β-ME compared to the three other redox agents is not surprising considering the small fraction of the thiolate form present at this pH and the lower redox potential of the thiolate anion compared to the other reducing agents used herein.29 To estimate the ratio of geminate recombination to radical escape (pathway II vs pathway I in Scheme 2, respectively) occurring with each reducing agent, we resorted to the kinetic analysis of the Cy3B ground-state recovery that is governed by the two competing processesthe fast geminate recombination of 3*Cy3B following ISC within the GRIP and the much slower decay of Cy3B•− following escape from the solvent cage of the GRIP. By finding the lifetimes of 3*Cy3B and Cy3B•− under each condition, and then fitting the Cy3B recovery kinetic traces recorded at 560 nm, we deconvoluted the contribution of both processes to the recovery of Cy3B (see also the Supporting Information, and Figures S4−S7). Following our analysis of kinetic traces, we estimated >13%, >27%, >48%, and ∼100% of geminate recombination occurring for AA, n-PG, TX, and β-ME respectively (see Table 1). Given that traces of Cy3B•− are formed immediately after laser excitation (vide supra), the estimated % of geminate recombination represents only a lower boundary of 3*Cy3B decay via geminate recombination. Importantly, increasing the solvent viscosity almost doubled the efficiency of geminate recombination for TX (from 48% to 84%), highlighting, once again, the competition between ISC within the solvent cage and escape from the solvent cage.

followed by radical escape from the solvent cage and formation of Cy3B•−, with very little extent of geminate recombination taking place. This result agrees with SMF studies with the analogue fluorophore Cy3 when attached to a DNA duplex, where a significant extent of redox blinking is observed in the presence of ascorbic acid (Figure S1).9,27 Transient absorption studies conducted with n-PG (Figure 1e,f) were comparable to those done in the presence of AA, where with quenching and decay of 3*Cy3B we observed the formation of significant amounts of Cy3B•− at 440 nm. However, as it is shown by the temporal evolution of the transient absorption traces at 560 nm (Figure 1f), a small (yet not negligible) extent of ground-state recovery was detectable in the presence of n-PG during the 3*Cy3B decay. Such a recovery was not apparent in the presence of AA. Kinetic traces acquired at 440 and 640 nm, normalized relative to the 3*Cy3B absorption at 640 nm, demonstrate a lower yield of Cy3B•− relative to 3*Cy3B with n-PG vs with AA (see Figure 2a,b), indicating that the geminate recombination pathway is more prevalent for n-PG than for AA. With the triplet quencher TX, a larger extent of geminate recombination was recorded compared to the results with AA and n-PG (Figure 1g,h). The decay of 3*Cy3B species was followed by a large extent of direct Cy3B ground-state recovery (Figure 1g,h), and a significantly smaller amount of Cy3B•− was formed in turn (compare relative absorbance at 640 and 440 nm in Figure 2c). It is also important to point out that the lifetime of Cy3B•− was significantly shorter (∼8 fold) in the presence of TX, compared with results with AA and n-PG (Figure S2, also Table 1), indicating that the fresh TX solution contained a fraction of its oxidized quinone form, TQ.9,28 Based on the reduction of the Cy3B•− lifetime we estimate the concentration of TQ at ∼1 μM (0.05% of TX oxidized) assuming diffusion controlled quenching of Cy3B•− by the latter. These results are in accordance with no redox blinking observed for the Cy3-DNA duplex in SMF studies, even when fresh solutions of TX (rather than purposely pre-oxidized solutions) were used (Figure S1).5,9,12 Quenching experiments with TX at increased viscosity (∼1.4-fold higher than PBS buffer) in a buffer containing 10% v/v glycerol (Figure 2d) showed a substantial reduction in the relative amount of Cy3B•− generated from 3*Cy3B. Given the higher viscosity, the rate of radical escape from the solvent cage (Scheme 2, pathway I) was reduced, favoring in turn the pathway II, i.e., fast ISC and regeneration of ground-state Cy3B via geminate recombination. Finally, results obtained with β-ME showed that the 3*Cy3B decay at 640 nm was primarily followed by direct recovery of Cy3B at 560 nm, indicating a quantitative geminate



DISCUSSION To rationalize the different degrees of geminate recombination observed in our studies, we compared the spin properties of the radicals formed following PeT between Cy3B and the four reducing agents studied herein. The rate of ISC in a GRIP is dictated by the quantum mechanical mixing of triplet and singlet sublevels that can be induced by either or all of the following mechanisms: (i) differences in radical g-factors;14,16,30−32 (ii) sums of hyperfine coupling (HFC) interactions within the two radical ions;14,16,30−32 and (iii) 13230

DOI: 10.1021/jacs.7b08134 J. Am. Chem. Soc. 2017, 139, 13227−13233

Article

Journal of the American Chemical Society

Scheme 3. Illustration of the Proposed Photostabilization Mechanism of Self-Healing Fluorophores, Highlighting the ISC Step

type of mechanism is difficult to prove experimentally, it was previously proposed to account for efficient ISC and groundstate recovery in triplet exciplexes for systems where triplet excited states are quenched by electron donors.33−35 The efficiency of geminate recombination is very sensitive to the local environment,14,42 as is illustrated by the transient absorption studies at increased viscosity (Figure 2). While our mechanistic studies were performed with a free fluorophore in solution, we note that in SMF studies the fluorophore is typically attached to a biomolecule of interest and immobilized on a crowded PEG or BSA surface, where a higher microviscosity prevails around the fluorophore. Under these conditions the extent of geminate recombination would be more pronounced. In that regard, it is worth emphasizing that for the photostabilization of fluorophores embedded within lipid membranes (or media where the local viscosity is much higher than that of an aqueous buffer), PeT with, e.g., αtocopherol12 as a photostabilizing agent may, in fact, mainly follow a geminate recombination pathway precluding the use of an oxidizing agent. The results of our transient absorption studies (Table 1) illustrate that a successful redox reagent for fluorophore photostabilization operating via PeT mechanism should satisfy two criteria: first, it should be an efficient electron donor (or acceptor) and second, it should provide an efficient mechanism to catalyze ISC from triplet to singlet GRIP. We propose that better photostabilization via geminate recombination will be achieved using more efficient reducing or oxidizing agents than β-ME, which after PeT can however form radicals with anistropic g tensors, like thiyl radicals. In that respect thiolcontaining antioxidants could be of potential interest. Tripletstate quenchers that after PeT transfer can form radicals with large numbers of hyperfine interactions would also be of potential interest in the situations where the GRIP between the fluorophore and the quencher is longer-lived (viscous environments). In those cases, one could even envision altering the fluorophore scaffold itself (e.g., by fluorination) to maximize the number of hyperfine interactions capable of assisting the ISC step in a GRIP or a biradical (vide inf ra). An ISC step is also involved (yet has not been explicitly addressed) in the photostabilization scheme involving redoxbased “self-healing” dyes.43−48 Here, the triplet excited state of the fluorophore is quenched by a covalently linked redox-active moiety resulting in the formation of triplet biradical intermediate (Scheme 3). An efficient back electron transfer, which requires fast ISC, is crucial for the performance of these constructs as it minimizes the time fluorophore spends in the reactive radical intermediate. The rate of ISC in the biradical intermediates has been shown to be sensitive not only to spin properties of the radicals formed but also to the distance and geometry between two radical centers (which in turn would be sensitive to the nature of the linker used).14,16,30,41,49−53 Further mechanistic studies and the optimization of ISC rates

spin−orbit coupling (SOC) in one or both of the radicals.14,33−35 We postulate that the quantitative geminate recombination observed when β-ME is used as a triplet quencher17 is related to the very anisotropic g-tensors of thiyl (RS•) radicals, which lead to its fast spin relaxation and the loss of triplet character in the GRIP.36 Bohne et al.37 showed that this type of mechanism is very efficient in micelles for GRIPs consisting of PhS• and Ph2CO•H radicals. In their work, the authors also pointed out that alkanethiyl and alkoxyl radicals (characterized by fast spin relaxation and thus “invisible” in EPR) would be capable of inducing a rapid loss of triplet character when partners in a triplet GRIP, thus reducing the efficiency of escape even in non-viscous solvents.36,37 This would indeed account for the quantitative geminate recombination pathway observed when β-ME is used as a triplet quencher, where a thiyl radical is formed upon PeT.17 The increasing extent of geminate recombination observed for ascorbyl, n-propyl gallyl, and TX chromanoxyl radicals generated after PeT between 3*Cy3B and AA, n-PG, and TX, respectively, may not be accounted for by the same mechanism as discussed for a thiyl radical above. These three radicals possess very similar g-factors indicative of isotropic g-tensor values (2.005, 2.0051, and 2.0046, respectively, Table S1 and Figure S8).38−40 The observed differences in geminate recombination efficiencies between the three radicals may in turn be related to a combination of factors that include different extents of HFC in the GRIP, SOC in the radicals formed and associated spin relaxation dynamics, and different escape rate constants from GRIPs (where differences would arise from the size and charge of the radicals). The EPR spectrum of the TX chromanoxyl radical, being significantly broadened compared to those of ascorbyl and npropyl gallyl radicals, is consistent with the presence of enhanced spin relaxation for the former, plausibly due to SOC.38−40 The enhanced spin relaxation dynamics are consistent with the higher extent of geminate recombination observed when TX chromanoxyl radical is part of the GRIP. The number and extent of HFC interactions are also highest for TX chromanoxyl radical, followed by n-propyl gallyl and ascorbyl radicals (Table S1, Figure S8).38−40 However, for HFC interactions to induce ISC in the GRIP, these interactions would have to be fast enough to compete with rapid radical escape from the solvent cage (typically 50−100 ps in nonviscous solvents).14 Since ISC induced by HFC interactions is typically much slower than that (lifetimes ≥ 10−9 s),16,41 it is difficult to foresee how HFC alone could account for the observed differences in geminate recombination efficiencies in a non-viscous environment. We also do not discard the possibility that the different efficiencies of geminate recombination observed for these radicals could be attributed to the formation of triplet exciplexe(s) between the fluorophore and the electron donors and different extent of SOC within the exciplexes. While this 13231

DOI: 10.1021/jacs.7b08134 J. Am. Chem. Soc. 2017, 139, 13227−13233

Article

Journal of the American Chemical Society in these constructs may lead to improved “self-healing” constructs for fluorescence single-molecule imaging.

(4) Juette, M. F.; Terry, D. S.; Wasserman, M. R.; Zhou, Z.; Altman, R. B.; Zheng, Q.; Blanchard, S. C. Curr. Opin. Chem. Biol. 2014, 20, 103. (5) Rasnik, I.; McKinney, S. A.; Ha, T. Nat. Methods 2006, 3, 891. (6) Dave, R.; Terry, D. S.; Munro, J. B.; Blanchard, S. C. Biophys. J. 2009, 96, 2371. (7) Zheng, Q.; Jockusch, S.; Zhou, Z.; Blanchard, S. C. Photochem. Photobiol. 2014, 90, 448. (8) Sauer, M.; Hofkens, J.; Enderlein, J. Handbook of Fluorescence Spectroscopy and Imaging; Wiley-VCH: Weinheim, 2011. (9) Vogelsang, J.; Kasper, R.; Steinhauer, C.; Person, B.; Heilemann, M.; Sauer, M.; Tinnefeld, P. Angew. Chem., Int. Ed. 2008, 47, 5465. (10) Widengren, J.; Chmyrov, A.; Eggeling, C.; Löfdahl, P.-Å.; Seidel, C. A. M. J. Phys. Chem. A 2007, 111, 429. (11) Chmyrov, A.; Sandén, T.; Widengren, J. J. Phys. Chem. B 2010, 114, 11282. (12) Glembockyte, V.; Lincoln, R.; Cosa, G. J. Am. Chem. Soc. 2015, 137, 1116. (13) Roy, R.; Hohng, S.; Ha, T. Nat. Methods 2008, 5, 507. (14) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Modern Molecular Photochemistry of Organic Molecules; University Science Books: Sausalito, CA, 2012. (15) Scaiano, J. C.; Abuin, E. B.; Stewart, L. C. J. Am. Chem. Soc. 1982, 104, 5673. (16) Turro, N. J. Proc. Natl. Acad. Sci. U. S. A. 1983, 80, 609. (17) Holzmeister, P.; Gietl, A.; Tinnefeld, P. Angew. Chem., Int. Ed. 2014, 53, 5685. (18) Cooper, M.; Ebner, A.; Briggs, M.; Burrows, M.; Gardner, N.; Richardson, R.; West, R. J. Fluoresc. 2004, 14, 145. (19) Levitus, M.; Ranjit, S. Q. Rev. Biophys. 2011, 44, 123. (20) Cao, J.; Wu, T.; Hu, C.; Liu, T.; Sun, W.; Fan, J.; Peng, X. Phys. Chem. Chem. Phys. 2012, 14, 13702. (21) Chibisov, A. K.; Zakharova, G. V.; Goerner, H.; Sogulyaev, Y. A.; Mushkalo, I. L.; Tolmachev, A. I. J. Phys. Chem. 1995, 99, 886. (22) Sanborn, M. E.; Connolly, B. K.; Gurunathan, K.; Levitus, M. J. Phys. Chem. B 2007, 111, 11064. (23) Ciuba, M. A.; Levitus, M. ChemPhysChem 2013, 14, 3495. (24) Chibisov, A. K. J. Photochem. 1976, 6, 199. (25) Giloh, H.; Sedat, J. Science 1982, 217, 1252. (26) Gaigalas, A. K.; Wang, L.; Cole, K. D.; Humphries, E. J. Phys. Chem. A 2004, 108, 4378. (27) Stein, I. H.; Capone, S.; Smit, J. H.; Baumann, F.; Cordes, T.; Tinnefeld, P. ChemPhysChem 2012, 13, 931. (28) Cordes, T.; Vogelsang, J.; Tinnefeld, P. J. Am. Chem. Soc. 2009, 131, 5018. (29) Buettner, G. R. Arch. Biochem. Biophys. 1993, 300, 535. (30) Levanon, H.; Norris, J. R. Chem. Rev. 1978, 78, 185. (31) Closs, G. L.; Forbes, M. D. E.; Norris, J. R. J. Phys. Chem. 1987, 91, 3592. (32) Buckley, C. D.; Hunter, D. A.; Hore, P. J.; McLauchlan, K. A. Chem. Phys. Lett. 1987, 135, 307. (33) Lambert, C. R.; Kochevar, I. E. Photochem. Photobiol. 1997, 66, 15. (34) Steiner, U.; Winter, G.; Kramer, H. E. A. J. Phys. Chem. 1977, 81, 1104. (35) Steiner, U.; Winter, G. Chem. Phys. Lett. 1978, 55, 364. (36) Symons, M. C. R. J. Chem. Soc., Perkin Trans. 2 1974, 1618. (37) Bohne, C.; Alnajjar, M. S.; Griller, D.; Scaiano, J. C. J. Am. Chem. Soc. 1991, 113, 1444. (38) González, P. M.; Aguiar, M. B.; Malanga, G.; Puntarulo, S. Comp. Biochem. Physiol., Part A: Mol. Integr. Physiol. 2013, 165, 439. (39) Reddan, J. R.; Giblin, F. J.; Sevilla, M.; Padgaonkar, V.; Dziedzic, D. C.; Leverenz, V. R.; Misra, I. C.; Chang, J. S.; Pena, J. T. Exp. Eye Res. 2003, 76, 49. (40) Giulivi, C.; Romero, F. J.; Cadenas, E. Arch. Biochem. Biophys. 1992, 299, 302. (41) Woodward, J. R. Prog. React. Kinet. Mech. 2002, 27, 165. (42) Barry, J. T.; Berg, D. J.; Tyler, D. R. J. Am. Chem. Soc. 2016, 138, 9389.



CONCLUSIONS In summary, we designed an LFP strategy to study the mechanism of triplet-state quenching of Cy3B and to estimate the extent of geminate recombination that occurs for four different reducing agents that are typically used as photostabilizers in SMF imaging. We demonstrate that geminate recombination of the fluorophore ground state is observed for all reducing agents to varying degrees. Our results show almost quantitative geminate recombination for β-ME (∼100%), a considerable extent of geminate recombination for TX (>48%), and less efficient geminate recombination for n-PG and AA (>27% and >13%, respectively), providing a more detailed picture of the photostabilization mechanism for these tripletstate quenchers. We demonstrate that the efficiency of geminate recombination is very sensitive to the spin properties of the radicals that are formed following PeT. These spin properties influence the rate of ISC process in a GRIP or biradicals (of interest for self-healing dyes). ISC followed by geminate recombination has been extensively studied in many different systems; however, the key parameters that control ISC in the triplet GRIP have not been considered in the light of fluorophore photostabilization. We suggest that better understanding and control of these geminate recombination processes can lead to improved photostabilization approaches.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b08134. Materials and methods, Table S1, and Figures S1−S8 (PDF)



AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Gonzalo Cosa: 0000-0003-0064-1345 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS G.C. is grateful to the Natural Science and Engineering Research Council of Canada (NSERC), the Fonds de Recherche du Quebec − Nature et Technologie (FQRNT), and the Canadian Foundation for Innovation (CFI) for funding. V.G. is thankful to the Drug Discovery and Training Program, Groupe de Recherche Axé sur la Structure des Protéines (GRASP), and NSERC Bionanomachines programs for postgraduate scholarships. The authors are also grateful to Prof. J. C. (Tito) Scaiano at the University of Ottawa for access to his laboratory’s laser flash photolysis set-up and Michel Grenier for his assistance with transient absorption studies.



REFERENCES

(1) Joo, C.; Ha, T. Cold Spring Harb. Protoc. 2012, 2012, 1109. (2) Stennett, E. M. S.; Ciuba, M. A.; Levitus, M. Chem. Soc. Rev. 2014, 43, 1057. (3) Zheng, Q.; Juette, M. F.; Jockusch, S.; Wasserman, M. R.; Zhou, Z.; Altman, R. B.; Blanchard, S. C. Chem. Soc. Rev. 2014, 43, 1044. 13232

DOI: 10.1021/jacs.7b08134 J. Am. Chem. Soc. 2017, 139, 13227−13233

Article

Journal of the American Chemical Society (43) Tinnefeld, P.; Cordes, T. Nat. Methods 2012, 9, 426. (44) van der Velde, J. H. M.; Oelerich, J.; Huang, J.; Smit, J. H.; Aminian Jazi, A.; Galiani, S.; Kolmakov, K.; Guoridis, G.; Eggeling, C.; Herrmann, A.; Roelfes, G.; Cordes, T. Nat. Commun. 2016, 7, 10144. (45) van der Velde, J. H. M.; Oelerich, J.; Huang, J.; Smit, J. H.; Hiermaier, M.; Ploetz, E.; Herrmann, A.; Roelfes, G.; Cordes, T. J. Phys. Chem. Lett. 2014, 5, 3792. (46) van der Velde, J. H. M.; Ploetz, E.; Hiermaier, M.; Oelerich, J.; de Vries, J. W.; Roelfes, G.; Cordes, T. ChemPhysChem 2013, 14, 4084. (47) van der Velde, J. H. M.; Uusitalo, J. J.; Ugen, L.-J.; Warszawik, E. M.; Herrmann, A.; Marrink, S. J.; Cordes, T. Faraday Discuss. 2015, 184, 221. (48) Altman, R. B.; Terry, D. S.; Zhou, Z.; Zheng, Q.; Geggier, P.; Kolster, R. A.; Zhao, Y.; Javitch, J. A.; Warren, J. D.; Blanchard, S. C. Nat. Methods 2012, 9, 68. (49) Colvin, M. T.; Ricks, A. B.; Scott, A. M.; Co, D. T.; Wasielewski, M. R. J. Phys. Chem. A 2012, 116, 1923. (50) Wang, J.; Doubleday, C.; Turro, N. J. J. Phys. Chem. 1989, 93, 4780. (51) Font-Sanchis, E.; Miranda, M. A.; Pérez-Prieto, J.; Scaiano, J. C. J. Org. Chem. 2002, 67, 6131. (52) Scaiano, J. C. Acc. Chem. Res. 1982, 15, 252. (53) Doubleday, C.; Turro, N. J.; Wang, J. F. Acc. Chem. Res. 1989, 22, 199.

13233

DOI: 10.1021/jacs.7b08134 J. Am. Chem. Soc. 2017, 139, 13227−13233