Article pubs.acs.org/JPCB
Thermal Isomerization Mechanism in Dronpa and Its Mutants Daryna Smyrnova,† Kirill Zinovjev,‡ Iñaki Tuñoń ,‡ and Arnout Ceulemans*,† †
Quantum Chemistry and Physical Chemistry Division, Department of Chemistry, KU Leuven, Celestijnenlaan 200F, 3001 Heverlee, Belgium ‡ Departament de Química Física, Universitat de València, 46100 Burjassot, Spain S Supporting Information *
ABSTRACT: The photoswitching speed of the reversibly switchable fluorescent proteins (RSFPs) from the family of green fluorescent proteins (GFPs) changes upon mutation which is of direct importance for various high-resolution techniques. Dronpa is one of the most used RSFPs. Its point mutants rsFastLime (Dronpa V157G) and rsKame (Dronpa V157L) exhibit a striking difference in their photoswitching speed. Here the QM/MM on-the-fly string method is used in order to explore the details of the thermal isomerization mechanism. The four principal ways in which isomerization may occur have been scrutinized for each of the three proteins. It has been shown that thermal isomerization occurs via a onebond-flip mechanism in all three proteins, although, in rsKame, where the chromophore is constrained more, the activation free energy difference between hula-twist and one-bond-flip is significantly smaller. Functional mode analysis has been applied to examine the motions of the amino acids during the isomerization. It clearly identifies the importance of Val/Leu 157 as well as the amino acids in the α-helix during the isomerization.
■
INTRODUCTION For already two decades, reversibly switchable fluorescent proteins (RSFPs) have been indispensable in high-resolution microscopy and biotechnology.1 They represent a broad range of mutants, while maintaining the same overall β-barrel structure.2−8 One of the prominent RSFP representatives is Dronpa, which exists in bright “on” and dark “off” thermostable states. The photoswitching speed changes upon mutation, which is beneficial for various high-resolution techniques.9−11 Here we are interested in single mutations of Dronpa which have a large effect on photoswitching: the Dronpa V157G mutant, named rsFastLime,12 has a 20-fold increase in the photoswitching speed. In contrast, in Dronpa V157L, named rsKame,13 mutation at the same position leads to a 2-fold decrease in photoswitching speed (see details in Table S1). It is evident that the effect of the protein environment can play a key role in the photophysics of the RSFPs. Recent experimental studies14−16 suggest that the photoswitching mechanism consists of two main steps: (1) the excited state chromophore isomerization, occurring on a picosecond time scale, and (2) ground state proton transfer which takes up to several microseconds. Nevertheless, experimental techniques fail to provide more insights into the process. A quantum mechanics/ molecular mechanics (QM/MM)-molecular dynamics (MD) study has already been successfully applied to study the thermal isomerization in asFPF95.17 However, to our knowledge, there are no such studies available for Dronpa and its mutants, nor for any other proteins with tunable photoswitching speed. Although the photoisomerization occurs in the excited state, a ground state study is indispensable to obtain a full depiction © XXXX American Chemical Society
of the process. Also, the thermal isomerization is very probably a bottleneck of the thermal relaxation of the protein. Although overall studies on the ground and excited state potential energy surface have been performed for chromophore isomerization in proteins, there is still a disagreement about the isomerization mechanism.18−21 The purpose of the present contribution is to perform a detailed analysis of the cis−trans isomerization on the ground state surface, comparing the two viable reaction modes: the so-called one-bond-flip (OBF) which corresponds to the rotation around the imidazolic double-bond (τ dihedral angle as seen in Figure 1) and the hula-twist mechanism, which is a concerted rotation of the adjacent central bonds of the chromophore (τ and ϕ dihedral angles in Figure 1). For both modes, forward and backward rotation will be compared. In this analysis, we are particularly interested in the effect of the protein environment, which requires a thorough conformational analysis. To evaluate the role of the protein environment, all simulations will be conducted for Dronpa and two of its mutants, rsFastLime (V157G) and rsKame (V157L), where mutated residue closely interacts with the chromophore site. There are numerous studies on p-hydroxybenzylidene-2,3dimethylimidazolinone (HBDI) photodynamics in a vacuum, solution, and condensed phase, including experimental and computational studies. The study of the chromophore in various environments by Wang et al.22 identifies a low dependence of the thermal isomerization on the viscosity of Received: October 28, 2016 Revised: November 22, 2016
A
DOI: 10.1021/acs.jpcb.6b10859 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
Figure 2. M06-2X/6-311G*//AM1/MM PMF for the lowest energy barrier isomerization of Dronpa, rsKame, and rsFastLime. λ is a collective variable (describing the rotation around the ϕ and τ dihedral angles) representing the extent of the reaction. Figure 1. QM region (shown in black) used in the simulation comprising the side chains of Arg 66, Arg 91, Glu 144, His 193, Glu 211, full Ser 142, and the entire chromophore. The most important amino acids from the MM region are shown in green. The water molecules present in the crystal structure were maintained and included in the MM region.
lowest for rsFastLime at 18 kcal/mol, followed by Dronpa at 30 kcal/mol, and rsKame having the highest barrier of 36 kcal/ mol. Also, for all three proteins, the trans-chromophore structure is destabilized in comparison to the cis state. For Dronpa and rsFastLime, the difference is around 4 kcal/mol, while, in rsKame, the trans state is 17 kcal/mol higher than the cis state. Another common feature for all three proteins is that the isomerization barrier is lowest for OBF rotation in the direction of His 193. The difference between OBF-Up and OBF-Down is more pronounced for rsFastLime (the difference being 19 kcal/mol), while, in rsKame, the OBF-Up mechanism is favored by only 1 kcal/mol (see Table S2). The reason why in these proteins the OBF mechanism results in a lower energy barrier is that it involves not only a plain change in τ angle but also a displacement of the chromophore backbone, thus reducing the cavity required for rotation. The transition and product state structures are shown in Figure S5 and Figure S6, respectively. It is clear that in comparison to the stable cis state the imidazoline ring goes out of plane, while the H-bonds between Arg 66 and the oxygen of the imidazoline ring get stronger, assisting the isomerization. The rsKame crowded amino acid environment in front of the phenyl moiety of the chromophore requires bigger out-of-plane movement adding to a higher energy barrier of the OBF isomerization mechanism. Our calculations of Dronpa and rsFastLime hula-like rotations converge to a consecutive rather than concerted rotation around τ and then ϕ dihedral angles (see Figure S3) with barriers larger than in the OBF path. For rsKame, the hulatwist isomerization mechanism results in a converged minimum free energy path (MFEP) that has a barrier comparable to OBF. We have studied it further and obtained the PMF at the M062X/6-311G* level (Figure 2). The resulting energy barrier for hula-twist isomerization is 52 and 49 kcal/mol for rotation in the direction of Arg 66 (Up) and Arg 91 (Down), respectively. The PMF for the downward hula-twist (see Figure 2) corresponds to a mechanism where rotation around ϕ slightly precedes the τ-rotation, resulting in an asynchronous hula-twist, as described in the recent work by Zhang et al.21 (see Figure S3). We attribute the decrease in free energy between the hulatwist and one-bond-flip mechanism in rsKame to a strong protein environment effect. In rsKame, the V157L mutation results in a smaller cavity around the chromophore as well as a more restrained protein matrix. The free energy difference between the cis and trans state after OBF-Up and downward
the solution, suggesting that chromophore isomerization does not involve big conformational change. This may lead to an assumption that the hula-twist mechanism should be prevalent, and the importance of both dihedral angles is confirmed by computational studies in the gas phase.23 However, computational studies in a polarizable environment suggest a lower barrier for rotation around the imidazolic double-bond.24 Martinez and co-workers in their studies show how the environment may determine the promoting coordinate leading to the isomerization25 and identify the rotation around the double-bond to be prominent for the chromophore in water solution. A molecular dynamics study of Dronpa by Moors et al.18 led to a hula-twist mechanism, based on the steric constraints of the crowded protein matrix around the chromophore. Also, a recent study of the GFP excited state dynamics by Zhang et al.21 suggests an asynchronous hula-twist mechanism to be feasible. Although in a vacuum a chromophore favors OBF, in a protein environment, hulatwist could be more preferable as less sterically hindered movement. In none of these studies was it realized that for both twists the sense of rotation is an important mechanistic issue in view of the influence of the protein environment. Therefore, in both cases, two opposite pathways are considered here: facing up (the phenole ring facing toward His 193) or facing down (the phenole ring facing toward Arg 91).
■
RESULTS Isomerization Free Energy. The results for all 12 calculations (4 kinds of rotations for each of 3 proteins) can be found in the Supporting Information (Figures S3 and S4 and Table S2), while here we present the comparison of the most energetically favorable potential of mean force (PMF) profiles in Figure 2 corresponding to the OBF-Up mechanism and the lowest lying hula-twist down mechanism. When comparing these profiles, two aspects have to be considered: the minimum free energy path and the concertedness of the mechanism. For all three proteins, the OBF-Up mechanism is more favorable with the energy barrier being the B
DOI: 10.1021/acs.jpcb.6b10859 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
Figure 3. PLS-FMA analysis on rsFastLime, Dronpa, and rsKame for the OBF-UP mechanism. Residues contributing the most to the isomerization are shown as a widening of the cartoon representation. The trans state chromophore in sticks representation is shown inside of the β-barrel.
and MECI structures presented in the Supporting Information of the aforementioned study. A superimposed image of all three structures can be seen in Figure S9. It is clear that our TS structure is closer to the I-twisted conformation obtained by Polyakov et al. The energy of this structure is approximately 32 kcal/mol (or 37 kcal/mol) which is close to an average thermal isomerization energy barrier obtained in our study. The energy at the conical intersection point is 69 kcal/mol, which is more than 30 kcal/mol higher than the transition state energies obtained in our work (18, 30, and 36 kcal/mol for rsFastLime, Dronpa, and rsKame, respectively). This suggests that no significant mixing between the ground and excited states occurs during the thermal isomerization. Although our study identifies the OBF-Up mechanism as the most viable for the thermal isomerization, we would like to clearly separate it from an excited state isomerization in Dronpa and its mutants. There is no clear agreement on the excited state isomerization mechanism: Morokuma and co-workers29,30 identify the OBF mechanism in Dronpa, while the Morozov and Groenhof study31 of a fast-switching Dronpa M159T points to the hula-twist mechanism. However, both studies determine the energy of the MECI point to be around 60−70 kcal/mol. This is at least 30 kcal/mol higher than the thermal isomerization barrier calculated in our study. While in the excited state hot species the hula-twist mechanism is possible, we think that rotation around the double-bond is energetically more preferable in the ground state. If we take into account the resonance structure of the anionic chromophore, hula-twist isomerization would require a partial transfer of the doublebond character from the I-bond (double-bond adjacent to the imidazoline ring) to the single P-bond (single bond adjacent to the phenol ring). This situation can be avoided during the OBF mechanism. The more rigid protein environment in front of the chromophore P-ring can enforce the hula-twist mechanism by significantly displacing the α-helix. That can be observed in the rsKame protein. However, even bigger sterical hindrance in front of the P-ring would probably distort the chromophore planarity, making it nonfluorescent as well. The importance of the protein environment flexibility has been shown in various studies.27,29,32 Our results also indicate direct dependence between the protein flexibility and the activation energy barrier. The hypothesis that rsFastLime should have a flatter energy surface due to highly similar promoting motions in the dark and bright states27 holds as well. Comparison of the off → on relaxation times (see Table S1) demonstrates a direct proportionality to the ground state isomerization barrier, which is also in accordance with the speed of the photo-
hula-twist rotation in rsKame is equal to 15 and 17 kcal/mol, respectively. Conformational Analysis. Previous studies by Mizuno et al. suggest that the β-barrel flexibility substantially differs between Dronpa dark and Dronpa bright. It has been hypothesized that such a difference in the flexibility may contribute to the relative stability of the bright over the dark protein state. Partial-least-squares functional mode analysis (PLS-FMA)26 indicates the amino acids which contribute the most to such a difference.27 Here we have applied PLS-FMA further in order to identify which amino acids hinder the isomerization. The result for the lowest energy mechanism for rsFastLime, Dronpa, and rsKame can be seen in Figure 3, and full analysis is available in Figures S7 and S8. It is clear that Val/Leu 157 plays a vital role in hindering the isomerization. The role of Ser 142 is well-known for stabilizing the chromophore in the cis state. We would like to pay attention to Asn 65, which is situated directly under the chromophore and in rsKame participates in the out of plane movement of the imidazoline ring in the course of the isomerization. Thr 58 establishes a hydrogen-bond with the Arg 91, which also interacts with the imidazoline ring. Comparison to the results of the PLS-FMA analysis for the OBF-Down and hula-twist mechanisms confirms that the OBF-Up mechanism results in lesser hindrance by the protein environment. In particular (as can be seen in Figures S7 and S8), isomerization via any other route has to include greater conformational changes around Met 159 and in the α-helix maintaining the chromophore. Conformational analysis of the product states yields the same conclusions: more pronounced displacement of the α-helix and consequent lower stabilization of Thr 58 in the products with higher energy, as well as conformational changes in Met 159. Detailed figures can be found in the Supporting Information (Figure S6).
■
DISCUSSION We have compared our results to already available data on the ground state and excited state potential energy surfaces.28 In particular, we were interested in how well the MO-62X method used here can reproduce the ground state energy barriers in comparison with multiconfigurational methods such as CASPT2. A study by Polyakov et al. identifies a so-called Itwisted conformation leading to the minimum energy conical intersection (MECI). Authors anticipate its significance on the way to the MECI structure. We have compared the transition state structure of the OBF-Up mechanism (as the most energetically favorable) obtained in our study to the I-twisted C
DOI: 10.1021/acs.jpcb.6b10859 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
QM region can be seen in Figure 1. For the rest of the system, the Amber 03 force field was used. The median structures from each run were chosen to start the potential energy surface (PES) scan. DFT calculations were made using fDynamo coupled to Gaussian 09.37 The resulting PES can be seen in the Supporting Information (Figure S3). The structures from the PES were taken as starting points for the following minimum free energy profile (MFEP) calculation with the on-the-fly string method38 using rotation around ϕ and τ dihedral angles as reaction coordinates. After MFEPs were obtained, a path collective variable39,40 was defined in order to obtain the potential of mean force (PMF) using umbrella sampling (US).41 A few extra reactant and product points were extrapolated to well describe reactant and product minima. PMFs were obtained at the AM1/MM level and then were corrected using interpolated single-point energies obtained at the M06-2X/6-311G*/MM level.42,43 This approach is justified by the similarities, at the qualitative level, of the low QM level and the high QM level PES (see Figure S2). In order to validate the DFT-level results, we also performed CASPT2-level calculations for the chromophore isomerization in a vacuum. Optimization along the isomerization coordinate was done at the M06-2X/6-311G* level using Gaussian 09.37 Consequent single-point energy calculation at the CASPT2/ ANO-L-VZTP level was performed using Molcas.44 Details of these calculations can be found in the Supporting Information.
switching. It is clear that the thermal isomerization should be taken into account along with the excited state process in order to access and explain the photoswitching mechanism of the RSFPs.
■
CONCLUSION
In this paper, we have investigated the ground state properties of the Dronpa, rsFastLime, and rsKame thermal isomerization mechanisms. These three proteins differ only in one singlepoint mutation and from rsFastLime to rsKame show a 40-fold increase in the speed of photoswitching. Our work connects this photophysical property to kinetic and thermodynamic properties. In total, four isomerization mechanisms were investigated for each protein. We have concluded that the one-bond-flip mechanism is more prevalent than the hula-twist one. The potential importance of the protein environment to aid the hula-twist rotation is demonstrated in the case of rsKame. Here the chromophore cavity has the smallest volume, and thus, the free energy barrier for the hula-twist mechanism becomes energetically closer to that of the one-bond-flip mechanism. Also, we have inspected the possibility for rotation proceeding in different directions. One-bond flip toward His 193 results in the lowest barrier. The thermal isomerization barrier for rsKame is the highest (36 kcal/mol) and the lowest for rsFastLime (18 kcal/mol). Hula-twist isomerization in rsKame has a barrier of 49 kcal/mol; therefore, it is still not probable to occur. We hypothesize that tighter packing of the protein may result in a lower difference between OBF and hulatwist and in some cases protein environment can enforce hulatwist mechanism. The manifestation of the protein environment effect has been analyzed using functional mode analysis and confirms the importance of the Val/Leu/Gly 157 residue. The importance of the α-helix stability has been revealed; in particular, the interaction of Thr 58 with Arg 91 and Asn 65 with Arg 66 can hinder the isomerization. We would like to emphasize consideration of both the hula-twist and OBF mechanisms for any method where the mechanism depends on the initial guess (such as nudged elastic band or string methods). Results of our study will also be used for further study of the excited state mechanism in Dronpa’s single-point mutants. Complementary consideration of both ground state and excited state mechanisms will help to explain the drastic change in their photoswitching speed.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcb.6b10859. Computational details, PES for Dronpa/rsFastLime/ rsKme at DFT and AM1 levels, FEPs and PLS-FMA analysis for all isomerization mechanisms, details of the transition and product state structures, comparison of DFT and CASPT2, and details of CASPT2 calculation (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Daryna Smyrnova: 0000-0002-4530-2470
■
Notes
COMPUTATIONAL METHODS Molecular Dynamics (MD) and Replica-Exchange Molecular Dynamics (REMD). Molecular dynamics (MD) and replica-exchange molecular dynamics (REMD) were run using the Gromacs 5.0.133 package and the Amber99sb-ILDN force field.34 All of the crystal waters remained, and protein was placed in a 80.5 × 80.5 × 80.5 Å3 TIP3P water box. REMD has proved to be a useful tool in exploration of GFP conformational space.35 Hence, it was used in order to incorporate a rather bulky Val157Leu mutation in rsKame. Then, cluster analysis was performed to pick the median structure. These structures were used as the input structures for the QM/MM-MD simulations. The QM/MM-MD calculations. The QM/MM-MD calculations were performed using fDynamo.36 The QM part, described by the AM1 semiempirical Hamiltonian, consisted of the full chromophore and the side chains of Arg 66, Arg 91, Ser 142, Glu 144, His 193, and Glu 211. The representation of the
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Financial support from the Flemish Government through the KU Leuven concerted action scheme is gratefully acknowledged. The computational resources and services used in this work were provided by the VSC (Flemish Supercomputer Center), funded by the Hercules Foundation and the Flemish Government. I.T. and K.Z. gratefully acknowledge a grant (Project No. CTQ2015-66223-C2-2-P) and a FPU fellowship from Ministerio de Economiá y Competitividad (Spain). D.S. would like to acknowledge the help of Ingrid Barcena Roig while using the VSC facilities.
■
REFERENCES
(1) Bourgeois, D.; Adam, V. Reversible Photoswitching in Fluorescent Proteins: A Mechanistic View. IUBMB Life 2012, 64, 482−491. D
DOI: 10.1021/acs.jpcb.6b10859 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B (2) Shaner, N. C.; Steinbach, P. A.; Tsien, R. Y. A Guide to Choosing Fluorescent Proteins. Nat. Methods 2005, 2, 905−909. (3) Heim, R.; Cubitt, A. B.; Tsien, R. Y. Improved Green Fluorescence. Nature 1995, 373, 663−664. (4) Cormack, B. P.; Valdivia, R. H.; Falkow, S. FACS-Optimized Mutants of the Green Fluorescent Protein (GFP). Gene 1996, 173, 33−38. (5) Yang, T.-T. Improved Fluorescence and Dual Color Detection with Enhanced Blue and Green Variants of the Green Fluorescent Protein. J. Biol. Chem. 1998, 273, 8212−8216. (6) Ormö, M.; Cubitt, A. B.; Kallio, K.; Gross, L. A.; Tsien, R. Y.; Remington, S. J. Crystal Structure of the Aequorea Victoria Green Fluorescent Protein. Science 1996, 273, 1392−1395. (7) Shaner, N. C.; Campbell, R. E.; Steinbach, P. A.; Giepmans, B. N. G.; Palmer, A. E.; Tsien, R. Y. Improved Monomeric Red, Orange and Yellow Fluorescent Proteins Derived from Discosoma Sp. Red Fluorescent Protein. Nat. Biotechnol. 2004, 22, 1567−1572. (8) Lelimousin, M.; Noirclerc-Savoye, M.; Lazareno-Saez, C.; Paetzold, B.; Le Vot, S.; Chazal, R.; Macheboeuf, P.; Field, M. J.; Bourgeois, D.; Royant, A. Intrinsic Dynamics in ECFP and Cerulean Control Fluorescence Quantum Yield. Biochemistry 2009, 48, 10038− 10046. (9) Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.; Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincott-Schwartz, J.; Hess, H. F. Imaging Intracellular Fluorescent Proteins at Nanometer Resolution. Science 2006, 313, 1642−1645. (10) Hell, S. W.; Wichmann, J. Breaking the Diffraction Resolution Limit by Stimulated Emission: Stimulated-Emission-Depletion Fluorescence Microscopy. Opt. Lett. 1994, 19, 780. (11) Moeyaert, B.; Dedecker, P. pcSOFI As a Smart Label-Based Superresolution Microscopy Technique. Methods Mol. Biol. 2014, 1148, 261−276. (12) Stiel, A. C.; Trowitzsch, S.; Weber, G.; Andresen, M.; Eggeling, C.; Hell, S. W.; Jakobs, S.; Wahl, M. C. 1.8 Å Bright-State Structure of the Reversibly Switchable Fluorescent Protein Dronpa Guides the Generation of Fast Switching Variants. Biochem. J. 2007, 402, 35. (13) Rosenbloom, A. B.; Lee, S.-H.; To, M.; Lee, A.; Shin, J. Y.; Bustamante, C. Optimized Two-Color Super Resolution Imaging of Drp1 During Mitochondrial Fission with a Slow-Switching Dronpa Variant. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 13093−13098. (14) Yadav, D.; Lacombat, F.; Dozova, N.; Rappaport, F.; Plaza, P.; Espagne, A. Real-Time Monitoring of Chromophore Isomerization and Deprotonation During the Photoactivation of the Fluorescent Protein Dronpa. J. Phys. Chem. B 2015, 119, 2404−2414. (15) 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. Nat. Commun. 2013, 4, 1461. (16) Colletier, J.-P.; Sliwa, M.; Gallat, F.-X.; Sugahara, M.; Guillon, V.; Schirò, G.; Coquelle, N.; Woodhouse, J.; Roux, L.; Gotthard, G.; et al. Serial Femtosecond Crystallography and Ultrafast Absorption Spectroscopy of the Photoswitchable Fluorescent Protein IrisFP. J. Phys. Chem. Lett. 2016, 7, 882−887. (17) Mironov, V. A.; Khrenova, M. G.; Grigorenko, B. L.; Savitsky, A. P.; Nemukhin, A. V. Thermal Isomerization of the Chromoprotein AsFP595 and Its Kindling Mutant A143G: QM/MM Molecular Dynamics Simulations. J. Phys. Chem. B 2013, 117, 13507−13514. (18) Moors, S. L. C.; Michielssens, S.; Flors, C.; Dedecker, P.; Hofkens, J.; Ceulemans, A. How Is Cis-Trans Isomerization Controlled in Dronpa Mutants? a Replica Exchange Molecular Dynamics Study. J. Chem. Theory Comput. 2008, 4, 1012−1020. (19) Jonasson, G.; Teuler, J.-M.; Vallverdu, G.; Mérola, F.; Ridard, J.; Lévy, B.; Demachy, I. Excited State Dynamics of the Green Fluorescent Protein on the Nanosecond Time Scale. J. Chem. Theory Comput. 2011, 7, 1990−1997. (20) Groenhof, G.; Bouxin-Cademartory, M.; Hess, B.; de Visser, S. P.; Berendsen, H. J. C.; Olivucci, M.; Mark, A. E.; Robb, M. A. Photoactivation of the Photoactive Yellow Protein: Why Photon
Absorption Triggers a Trans-To-Cis Isomerization of the Chromophore in the Protein. J. Am. Chem. Soc. 2004, 126, 4228−4233. (21) Zhang, Q.; Chen, X.; Cui, G.; Fang, W.-H.; Thiel, W. Concerted Asynchronous Hula-Twist Photoisomerization in the S65T/H148D Mutant of Green Fluorescent Protein. Angew. Chem., Int. Ed. 2014, 53, 8649−8653. (22) Wang, D.; Merz, T.; van Gunsteren, W. F. The Thermal Isomerization of the GFP Chromophore: A Computational Study. Phys. Chem. Chem. Phys. 2010, 12, 11051. (23) Martin, M. E.; Negri, F.; Olivucci, M. Origin, Nature, and Fate of the Fluorescent State of the Green Fluorescent Protein Chromophore at the CASPT2//CASSCF Resolution. J. Am. Chem. Soc. 2004, 126, 5452−5464. (24) Altoé, P.; Bernardi, F.; Garavelli, M.; Orlandi, G.; Negri, F. Solvent Effects on the Vibrational Activity and Photodynamics of the Green Fluorescent Protein Chromophore: a Quantum-Chemical Study. J. Am. Chem. Soc. 2005, 127, 3952−3963. (25) Toniolo, A.; Olsen, S.; Manohar, L.; Martínez, T. J. Conical Intersection Dynamics in Solution: The Chromophore of Green Fluorescent Protein. Faraday Discuss. 2004, 127, 149−163. (26) Krivobokova, T.; Briones, R.; Hub, J. S.; Munk, A.; de Groot, B. L. Partial Least-Squares Functional Mode Analysis: Application to the Membrane Proteins AQP1, Aqy1, and CLC-Ec1. Biophys. J. 2012, 103, 786−796. (27) Smyrnova, D.; Moeyaert, B.; Michielssens, S.; Hofkens, J.; Dedecker, P.; Ceulemans, A. Molecular Dynamic Indicators of the Photoswitching Properties of Green Fluorescent Proteins. J. Phys. Chem. B 2015, 119, 12007−12016. (28) Polyakov, I. V.; Grigorenko, B. L.; Epifanovsky, E. M.; Krylov, A. I.; Nemukhin, A. V. Potential Energy Landscape of the Electronic States of the GFP Chromophore in Different Protonation Forms: Electronic Transition Energies and Conical Intersections. J. Chem. Theory Comput. 2010, 6, 2377−2387. (29) Li, X.; Chung, L. W.; Mizuno, H.; Miyawaki, A.; Morokuma, K. A Theoretical Study on the Nature of On- and Off-States of Reversibly Photoswitching Fluorescent Protein Dronpa: Absorption, Emission, Protonation, and Raman. J. Phys. Chem. B 2010, 114, 1114−1126. (30) Li, X.; Chung, L. W.; Mizuno, H.; Miyawaki, A.; Morokuma, K. Primary Events of Photodynamics in Reversible Photoswitching Fluorescent Protein Dronpa. J. Phys. Chem. Lett. 2010, 1, 3328−3333. (31) Morozov, D.; Groenhof, G. Hydrogen Bond Fluctuations Control Photochromism in a Reversibly Photo-Switchable Fluorescent Protein. Angew. Chem., Int. Ed. 2016, 55, 576−578. (32) Mizuno, H.; Mal, T. K.; Walchli, M.; Kikuchi, A.; Fukano, T.; Ando, R.; Jeyakanthan, J.; Taka, J.; Shiro, Y.; Ikura, M.; et al. LightDependent Regulation of Structural Flexibility in a Photochromic Fluorescent Protein. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 9227− 9232. (33) Pronk, S.; Pall, S.; Schulz, R.; Larsson, P.; Bjelkmar, P.; Apostolov, R.; Shirts, M. R.; Smith, J. C.; Kasson, P. M.; van der Spoel, D.; et al. GROMACS 4.5: A High-Throughput and Highly Parallel Open Source Molecular Simulation Toolkit. Bioinformatics 2013, 29, 845−854. (34) Case, D. A.; Babin, V.; Berryman, J. T.; Betz, R. M.; Cai, Q.; Cerutti, D. S.; Cheatham, T. E., III; Darden, T. A.; Duke, R. E.; Gohlke, H.; et al. AMBER 14; University of California, San Francisco, 2014. (35) Mitchell, F. L.; Frank, F.; Marks, G. E.; Suzuki, M.; Douglas, K. T.; Bryce, R. A. Molecular Dynamics Study of Chemically Engineered Green Fluorescent Protein Mutants: Comparison of Intramolecular Fluorescence Resonance Energy Transfer Rate. Proteins: Struct., Funct., Genet. 2009, 75, 28−39. (36) Field, M. J.; Bash, P. A.; Karplus, M. A Combined Quantum Mechanical and Molecular Mechanical Potential for Molecular Dynamics Simulations. J. Comput. Chem. 1990, 11, 700−733. (37) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al.; Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009. E
DOI: 10.1021/acs.jpcb.6b10859 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B (38) Maragliano, L.; Vanden-Eijnden, E. On-The-Fly String Method for Minimum Free Energy Paths Calculation. Chem. Phys. Lett. 2007, 446, 182−190. (39) Zinovjev, K.; Martí, S.; Tuñoń , I. A Collective Coordinate to Obtain Free Energy Profiles for Complex Reactions in Condensed Phases. J. Chem. Theory Comput. 2012, 8, 1795−1801. (40) Zinovjev, K.; Ruiz-Pernía, J. J.; Tuñoń , I. Toward an Automatic Determination of Enzymatic Reaction Mechanisms and Their Activation Free Energies. J. Chem. Theory Comput. 2013, 9, 3740− 3749. (41) Torrie, G.; Valleau, J. Nonphysical Sampling Distributions in Monte Carlo Free-Energy Estimation: Umbrella Sampling. J. Comput. Phys. 1977, 23, 187−199. (42) Ruiz-Pernía, J. J.; Silla, E.; Tuñoń , I.; Martí, S.; Moliner, V. Hybrid QM/MM Potentials of Mean Force with Interpolated Corrections. J. Phys. Chem. B 2004, 108, 8427−8433. (43) Ruiz-Pernía, J. J.; Silla, E.; Tuñoń , I.; Martí, S. Hybrid Quantum Mechanics/Molecular Mechanics Simulations with Two-Dimensional Interpolated Corrections: Application to Enzymatic Processes. J. Phys. Chem. B 2006, 110, 17663−17670. (44) Aquilante, F.; Autschbach, J.; Carlson, R. K.; Chibotaru, L. F.; Delcey, M. G.; De Vico, L.; Fdez. Galván, I.; Ferré, N.; Frutos, L. M.; Gagliardi, L.; et al. Molcas 8: New Capabilities for Multiconfigurational Quantum Chemical Calculations Across the Periodic Table. J. Comput. Chem. 2016, 37, 506−541.
F
DOI: 10.1021/acs.jpcb.6b10859 J. Phys. Chem. B XXXX, XXX, XXX−XXX
