Electron Transfer Pathways of Cyclobutane Pyrimidine Dimer

Jun 11, 2018 - Article. Previous Article · Next Article · Table of Contents. Electron Transfer Pathways of Cyclobutane Pyrimidine Dimer Photolyase Rev...
0 downloads 0 Views 2MB Size
Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

B: Biophysics; Physical Chemistry of Biological Systems and Biomolecules

Electron Transfer Pathways of Cyclobutane Pyrimidine Dimer Photolyase Revisited Ryuma Sato, Hirotaka Kitoh-Nishioka, Koji Ando, and Takahisa Yamato J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.8b04333 • Publication Date (Web): 11 Jun 2018 Downloaded from http://pubs.acs.org on June 21, 2018

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 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 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.

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 32 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

Electron Transfer Pathways of Cyclobutane Pyrimidine Dimer Photolyase Revisited Ryuma Sato1,2,*, Hirotaka Kitoh-Nishioka3,4, Koji Ando5, Takahisa Yamato1,6 1. Department of Physics, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602, Japan 2. Current Address; Center for Biosystems Dynamics Research, RIKEN, 6-2-3 Furuedai, Suita, Osaka 565-0874, Japan 3. Center for Computational Sciences, University of Tsukuba 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8577, Japan 4. Current Address; JST-PRESTO, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan 5. Department of Information and Sciences, Tokyo Woman’s Christian University, 2-6-1 Zempukuji, Suginami-ku, Tokyo 167-8585, Japan 6. Institute of Genetics and Molecular and Cellular Biology, University of Strasbourg, 1 rue Laurent Fries Parc d'Innovation 67404 ILLKIRCH, Cedex, France * Correspondence: [email protected] ABSTRACT The photoinduced electron transfer (ET) reaction of cyclobutane pyrimidine dimer (CPD) photolyase plays an essential role in its DNA repair reaction, and the molecular mechanism of the ET reaction has attracted a large number of experimental and theoretical studies. We investigated the quantum mechanical nature of their ET reactions, characterized by multiple ET pathways of the CPD photolyase derived from A. nidulans. Using the generalized Mulliken-Hush (GMH) method and the bridge green function (GF) methods, we estimated the electronic coupling matrix element, TDA, to be 36±30 cm–1 from the donor (FADH–) to the acceptor (CPD). The estimated ET time was 386 ps, in good agreement with the experimental value (250 ps) in the literature. Furthermore, we performed the molecular dynamics (MD) simulations and ab initio molecular orbital (MO) calculations, and explored the electron tunneling pathway. We examined 20 different structures during the MD trajectory and quantitatively evaluated the electron tunneling currents for each of them. As a result, we demonstrated that the ET route via Asn349 was the dominant pathway among the five major 1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

routes

via

(Adenine/Asn349),

(Adenine/Glu283),

Page 2 of 32

(Adenine/Glu283/Asn349/Met353),

(Met353/Asn349), and (Asn349), indicating that Asn349 is an essential amino acid residue in the ET reaction. INTRODUCTION Photolyase/cryptochrome

blue

light

photoreceptors

are

highly

conserved

flavoproteins with the flavin adenine dinucleotide (FAD). Among them, photolyases absorb blue or near-ultraviolet (UV) light and repair pyrimidine dimers in UV damaged DNA using photo induced electron transfer (ET) reaction.1-4 The most frequently occurring UV-induced lesion is the cyclobutane pyrimidine dimer (CPD) formed by a [2+2] cycloaddion of two adjacent pyrimidine bases, and the CPD lesions are specifically repaired by the CPD photolyase, which binds to CPD-containing DNA. The repair reaction is initiated by the photoexcitation of the fully reduced FADH–, and then ET takes place from FADH–* to CPD.5-8 A large number of theoretical,9-19 and experimental6-8,20-26 studies have been involved in this reaction. In the pioneering work of Medvedev and Stuchebrukhov,9,10 they employed the concept of interatomic electron tunneling current, and analyzed the ET pathways of CPD photolyase. As a result, they pointed out the importance of the adenine moiety. In 2004, Mees et al. revealed the atomic details of the complex structure of CPD and CPD photolyase derived from A. nidulans by X-ray crystallography.27 In the crystal structure, the adenine moiety of FADH– bridges between the electron donating isoalloxazine ring and CPD via two hydrogen bonds, suggesting the presence of ET pathways via adenine. They also pointed out the possibility of direct ET because the C8-methyl group of FADH– is in direct contact with the C4-carbonyl group of CPD. In 2007, Prytkova et al. proposed direct ET mechanism from FADH– to CPD. They performed excited state calculations using the INDO/S method, and demonstrated that the transitions S0→S1, S0→S2 and S0→S3 are forbidden, allowed, and allowed, respectively. Thereby, they proposed a reaction scheme that begins with the photoactivated excitation from S0 to either S2 or S3 followed by the relaxation to the S1 state. Furthermore, they demonstrated that the methyl group in the proximal side of FADH– had a significant influence on the value of TDA, while neither adenine nor ribitol part of FADH– did, suggesting the possibility of direct ET mechanism. In 2016, Lee et al. performed the QM/MM calculation17 of the isoalloxazine and adenine moiety with/without ribose ring of FADH–, and reported that S1 and S3 were locally excited (LE) states of isoalloxazine, and S2 was the charge transfer (CT) state from isoalloxazine moiety to adenine moiety. Accordingly, they concluded that the CT to adenine occurs in the ET reaction. Moreover, they demonstrated eight pathways from FADH– to CPD. Among them, electron tunneling occurs 2 ACS Paragon Plus Environment

Page 3 of 32 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

in the six pathways, while sequential electron hopping do in the remaining two pathways. In 2018, Rousseau et al. investigated the temperature dependence of the ET pathways of A. nidulans, E. coli, T. thermophilus, and S. tokodaii photolyases,19 and indicated the occurrence of the adenine-mediated superexchange ET in the photolyases of A. nidulans, T. thermophilus, and S. tokodaii at their physiological temperature, whereas that is not the case in E. coli. In 2008, Miyazawa et al. examined the roles of the surrounding environment of the active site of the CPD photolyase. With the amino acid residues near the active site being explicitly included, they performed molecular orbital (MO) calculations.12 As a result, they observed that the electron tunneling current was mediated not only by the adenine moiety but also by the methionine residue (Met353) near the active site. Hereafter, we denote this methionine site as M-site. Furthermore, they performed bioinformatics analysis of the photolyases/blue light photoreceptor family. Consequently, they found that the methionine residue was perfectly conserved at the M-site of the class I CPD photolyase subfamily. In 2015, we performed in silico mutations of the active site structures of the class I CPD photolyase and cryptochrome-DASH,27 the former (latter) of which exhibits (lacks) CPD repair activity, although they demonstrate strikingly similar structures. It should be noted that the adjacent residue to the 3'- (5'-) side of CPD is methionine (glutamic acid) in the class I CPD photolyase. It has been widely accepted that the H-bond between CPD and this glutamic acid is important in the ET reaction. In the in silico mutations we demonstrated that the H-bond was kept intact when the M-site residue was methionine, which is replaced with glutamine in cryptochrome-DASH, indicating an pivotal role of the M-site methionine in the active site formation of CPD photolyase. Liu et al. constituted four different types of CPD, i.e., U< >T, U< >U, T< >U, and T< >T, and measured the ET time of the complex form of the CPD photolyase with each type of CPD.8,20 As a result, the ET time increased in the order of U< >T, U< >U, T< >U, and T< >T as 63, 73, 85 and 250 ps, respectively. From this observation, they concluded that the forward ET reaction from the donor (FADH–) to the acceptor (CPD) takes place via the adenine moiety through the 5' side of CPD. On the other hand, Kao et al. highlighted the possibility of direct ET from FADH– to CPD.6,7 In the theoretical studies in the literatures, Prytkova et al. emphasized direct ET,11 while Antony et al., Medvedev et al. and Miyazawa et al. did the superexchange mechanism via adenine moiety and/or the surrounding amino acid residues in the ET reaction from FADH– to CPD.9,10,12 As described thus far, there is still ongoing debate regarding the ET pathways of the CPD photolyase. To address the issue, we performed a theoretical study on the ET pathways of CPD photolyase based on the standard Marcus formula:28 3 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

ET =

 |DA|

ℏ  B 

exp −

 

B 



Page 4 of 32

(1)

where DA is the electron coupling matrix element, Δ is the reaction driving force,  is the reorganization energy,  is the absolute temperature, B is the Boltzmann constant, and ℏ is the Planck constant divided by 2. To calculate TDA, we employed the generalized Mulliken-Hush (GMH) method29-31 and the bridge green function (GF) method32 with the fragment molecular orbital (FMO) method.33-36 Furthermore, we estimated the electron tunneling pathways using the tunneling current theorem.9,10,32,37,38 METHODS Molecular dynamics simulation The atomic coordinates of the initial structure for molecular dynamics (MD) simulations was derived from the X-ray structure (PDB entry: 1TEZ).26 The structure of CPD was extracted from the X-ray coordinates (PDB entry: 1SNH),39 and the –O–P2–O– part was replaced with –O–CH2–O–. Then we protonated Glu283, whereas the other titratable amino acid residues were maintained at the standard states at physiological pH. This model was then immersed in an octahedral box of water molecules, where the simulation boxes were constructed with a margin of at least 15 Å from the protein surface to the box boundaries. Thus, the total number of atoms was 84,033. The ff99SB force field40 and TIP3P model41 were employed for the polypeptide chain and water molecules, respectively. We developed the force field parameters of FADH–, CPD-analog, and 8-hydroxy-5-deazafravin, hereafter demoted (8-HDF). Here we assumed that FADH–, the electron donor, (CPD-analog, the electron acceptor) stays at the crossing point between the initial and the final electronic states with –3 and –2 (–1 and –2) charges, respectively. In the final state, FADH– becomes FADH• and CPD-analog becomes CPD-analog anion radical. To obtain the force field parameters for FADH– and CPD-analog, we performed the electronic structure calculations for the X-ray structure of FADH– and the NMR structure of CPD-analog using the Gaussian03 program,42 and their initial and final states were evaluated at the RHF/6-31G(d) and UHF/6-31G(d) levels, respectively. For these in the initial and the final states of these cofactors, the partial atomic charge and the other force field parameters are respectively calculated by the RESP and the antechamber programs of the AMBER 12 program package.43 Then, we set the force field parameters of the donor and the acceptor to the average values of their initial and final states. To obtain the force field parameters of 8-HDF, the structure optimization was 4 ACS Paragon Plus Environment

Page 5 of 32 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

performed by the Gaussian03 program at the RHF/6-31G(d) level. Then the atomic partial charges and the other force field parameters were respectively calculated by the RESP and the antechamber programs of the AMBER 12 program package. The initial model was immersed in a box of water molecules and seven sodium ions were added to neutralize the system. In the first (second) half of the energy minimization was performed with (without) harmonic restraints of the force constant of 100 (kcal/mol/Å2) imposed on the heavy atoms except for the water molecules. In the per-equilibration stage, the temperature of the system was gradually elevated from 0 K to 300 K for 20 ps during the NVT MD simulation with the time step of 2 fs. The covalent bond lengths involving hydrogen atoms were kept fixed using the SHAKE algorithm.44 The thermal equilibration was achieved after the NVT MD simulation at T=300 K for 5 ns without any positional restraints. The production run was performed by the NPT MD simulation at T=300 K and P=1 bar with the time step of 1 fs for 5 ns, and we recorded atomic coordinates every 100 fs, thereby we obtained 50000 snapshots from the MD simulations. The long-range non-bonded interactions were evaluated by the particle-mesh Ewald method in the entire calculations. Electronic coupling matrix element using GMH To estimate the electron coupling matrix element, TDA, we first calculated the electronic structures of the locally excited (LE) and the charge transfer (CT) states of the active site, where we explicitly treated the flavin hydroquinone (HQ) moiety, CPD, adenine moiety, Glu283, Asn349, and Met353 (see Figure 1) as a quantum mechanical (QM) region. In addition, we placed the atomic partial charges of the surrounding amino acid residues, DNA, and water molecules around the QM region, and the molecular orbitals calculations of the QM region were performed under the electrostatic influence of the surrounding environment. We estimated TDA using the configuration interaction singlet (CIS)45 with the generalized Mulliken-Hush (GMH) method29-31 as follows:

DA =

| |

(2)

   | |

where Δ is the excited state energy difference between the CIS wavefunctions ψ1 and ψ2.

Δ



is the absolute value of the difference between the corresponding state dipole moment

vectors.



is the transition dipole moment from ψ1 to ψ2. The GMH method has an

advantage over the ordinary energy splitting method that it can be applied to any geometry without limitation other them to the transition state.29,30,46-49 We calculated the electronic 5 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 32

structure using GAMESS program.50 Fragment molecular orbitals method and FMO-LCMO We estimated TDA and the electron tunneling pathways using FMO method and their linier combinations (FMO-LCMO).32,51-55 In FMO method with fragment dimer correction (FMO2), the total electronic energy, Etotal, is approximated by following equation:

(

Etotal = ∑ EIJ − N − 2 I >J

)∑E

(3)

I

I

where N is the number of the fragment monomers and ! and !" are the electronic energy of the fragment monomer I and the fragment dimer IJ, respectively. To decrease the calculation cost in the FMO method, it usually employs the electrostatic dimer (ES-DIM) approximation,34 which avoids self-consistent field (SCF) calculation of far separated dimers. But we did not employ ES-DIM approximation to estimate all inter-fragment tunneling currents. In this study, we preformed the FMO calculation using GAMESS program.50 The FMO-LCMO method52-55 utilizes the canonical molecular orbitals (MOs) and orbital energies of all fragment monomers and dimers which obtained from the FMO calculation. Here, we denote the pth MO of fragment monomer I by #$%! &. The FMO-LCMO method approximately expresses the total one-electron Hamiltonian matrix in the fragment monomer MO representation by the following equations52-55: '!( ,"* = +$%! #,!" #$-" & for / ≠ 0

HI

p ,I q

(

(4)

)

= ∑ φ pI hIJ φqI − N − 2 φ pI hI φqI

(5)

J ≠I

where ,! and ,!" are Fock matrices of fragment monomer I and dimer IJ, respectively.

Namely, eq 4 represents the inter-fragment Hamiltonian matrix elements between #$%! & and #$-" &. On the other hand, eq 5 represents the intra-fragment Hamiltonian matrix elements

between #$%! & and #$-! &. The MOs and corresponding orbital energies of the whole system

are obtained by solving the generalized eigenvalues problem with the overlap matrix of monomer MOs. Electronic coupling matrix element using the bridge GF/FMO-LCMO 6 ACS Paragon Plus Environment

Page 7 of 32 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

By using the GF method,56-59 we estimated TDA with the FMO-LCMO method as follows:32,51,55 N

N

TDA = Hφdirect + ∑∑ ,φ D

(

×G

∑ ) (

J I p ≠φ D ,φ A J q ≠φ D ,φ A

I B



A

(E )

tun I ,J p q

(E

tun

SJ

q ,φ A

() E

− HJ

q ,φ A

tun



D ,I p

− Hφ

D ,I p

) (6)

)

where $D and $A are fragment monomer MOs of the donor and acceptor, respectively. The 3

bridge GF is given in the matrix form by GB tun  = 1tun SB − HB 2

in which HB is the

bridge part of the FMO-LCMO Hamiltonian matrix in eqs 4 and 5, SB is the overlap matrix among the nonorthogonal fragment monomer, and Etun is the tunneling energy parameter and direct set the average value of the orbital energies between the donor and acceptor. H φD ,φA is direct

donor-acceptor coupling. Electron tunneling current analysis To analyze the electron tunneling pathway, we adopted the tunneling current 60-62

method

within the one-electron picture.32,51 We denote the MOs of the initial and final

diabatic states, #45 & and #4 6 &, in terms of the fragment monomer MOs, #$%! &, as follows:

N

ψ i = CiD φD + ∑∑ CiI φ pI

(7)

p

I

Ip

N

ψ f = CfA φA + ∑∑ CfI φ pI

(8)

p

I

Ip

where $D and $A are the donor and acceptor MOs and #CiD # ≃ 1 and #CfA # ≃ 1 can be assumed in the weak coupling case. The superscripts (i and f) mean the initial and final state. The coefficients Ci9( and Cf:( are denoted by the bridge GF matrix as follows:37,56,62,63

N

(

CiL = −∑∑ Etun Sφ p

M Mq

D ,M q

− Hφ

)G ( E B

D ,M q

)

(9)

tun M ,L q p

7 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

N

( )

CfL = −∑∑G B Etun p

M Mq

L p ,M q

(E

tun

SM

q ,φ A

− HM

q ,φA

)

Page 8 of 32

(10)

The tunneling current 09( ,:* between the fragment MOs, $%9 and $-: , is given by the following equation:



09( ,:* = ;'9( ,:* − tun 9,: = ℏ?09,: @/DA

(14)

We estimated the TDA using GMH method (eq 2) and the bridge GF method (eq 13) and estimated the ET rate by the Marcus theorem. Furthermore, we indicated the electron tunneling pathways from FADH– to CPD using eq 14. RESULTS AND DISCUSSION Excited state electron structure calculations To analyze the UV-induced DNA lesion repair mechanism, we calculated the excited states of the active site (including the electron donor and acceptor). In the previous studies, Grimme, and Dreuw and Head-Gordon performed single-reference ab initio quantum 8 ACS Paragon Plus Environment

Page 9 of 32 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

mechanical (QM) calculation to obtain the excited states for large biological systems.64,65 Here, we also employed the configuration interaction singles (CIS) method, and performed ab initio QM calculation of the model system (Figure 1) for 80 randomly chosen MD snapshots. In our MD simulations, the average value of the root mean square deviation (RMSD) of all atoms from the X-ray crystal structure was 2.8±0.1 Å (Figure S1). As a result, the two lowest singlet excited states (S1 and S2) of FADH– were π → π* transition for MD snapshots, indicating that the charge population shifted from the distal side to the proximal side, in line with the preceding studies.11,66-68 The S0 → S1 transition was neglected because the oscillator strength was very small (f = (4 – 860) × 10–4), while the S0 → S2 transition was allowed because the oscillator strength was large enough ( f = (2 – 2.5) × 10–1 ). The third excited state, S3, was the charge transfer (CT) state (Table 1). Although the exact values of the oscillator strength and the excited state wavefunctions were different among the 80 snapshots, in most case, the S1 and S2 state were locally excited (LE) state of hydroquinone (HQ), and the S3 state was the CT state from HQ to CPD. By taking these observations into account, it is likely that the scheme of this ET reaction is described as follows: (1) FADH— absorbs blue/near-UV light and the system is excited from S0 to S2. (2) internal conversion occurs from S2 to S1 (Figure 2). (3) the ET reaction occurs when S1 and S3 energies match by accidental coincidence during thermal fluctuation. Our Electronic coupling matrix element and electron transfer times We evaluated the electron coupling matrix element, DA , with two different 

ave methods (1) GMH and (2) bridge GF (Table 2), and the average value, DA = N ∑NCD|DA |

was obtained for each method, where N was the number of randomly chosen snapshots. (1) We performed CIS calculations for N (=80) structures. The values of DA were calculated based on the GMH method. Here we selected the two excited states, S1 (LE state of FADH—) and S3 (CT state from FADH— to CPD), to obtain DA . (2) Next, we performed FMO calculations for N (=20) snapshots at the RHF/6-31G(d) level, and the values of DA were calculated by using FMO-linear combination molecular orbital (LCMO) method and the bridge GF method. The MO energy spectrum of full RHF reference was reconstructed between –10 and 10 eV by FMO-LCMO (Figure 3). The energy spectrum obtained by full RHF and FMO-LCMO were in good agreement to each other, and the orbital energies of (ψ1 (LUMO) and ψ8 (LUMO+8)) were (4.78522, 6.96754) and (4.77389, 6.94241) eV, respectively. In Figure 4, we illustrated typical donor and acceptor orbitals based on a snapshot derived from the MD trajectory. The atomic coordinates of the QM region of the 9 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

snapshot are shown in Table S1. As seen in Figure 4(a) and (b), ψ1 and ψ8 are mostly π-conjugated orbitals on CPD and HQ, respectively. The FMO calculations demonstrated that the LUMOs of HQ (ψD) and CPD (ψA) were π-conjugated orbitals on the HQ and CPD, respectively, indicating that the results of the CIS and the FMO calculations were consistent with each other. In Table 2, we show the values of DA calculated by (1) the GMH method and (2) the bridge GF method, where the tunneling energy, Etun, was set to the midpoint value ave between the energy of HQ (LUMO) and that of CPD (LUMO). The value of DA calculated

by the GMH (bridge GF) method was 36±30 cm–1 (35±23 cm–1). We observed that the protein molecule demonstrated no large conformational fluctuations (Figure. S2) during the MD simulation. Nevertheless, the value of TDA has shown significant fluctuations from approximately 2 (conformation 1) to 156 cm–1 (conformation 2), indicating that the electronic coupling is very sensitive to a subtle change of the molecular structure. If we take a closer look at the active site in the X-ray structure, conformations 1 and 2, a set of interatomic distances (N61(FADH–)–O4(CPD), N61(FADH–)–O4'(CPD)) were (3.12Å, 3.16Å), (3.00Å, 3.09Å), and (2.90Å, 3.12Å), respectively (Figure. S3). Interestingly, we recognized a tendency in the MD trajectory that the value of TDA became large when the adenine moiety approached to the 3' side of CPD, whereas it became small when it approached to the 5' side. It may be true, however, that the behavior of the adenine moiety is not the sole factor determining the value of TDA. The thermal fluctuations of the surrounding protein environment also affect the electronic coupling. Note that these values include the contributions from both direct (i.e., FADH–* → CPD) and superexchange (i.e., FADH–* → amino acid residues → CPD) mechanisms, and the calculated ratio of the former to the latter were approximately 8:2 and 3:7 by the GMH and the bridge GF methods, respectively. The ET rate, kET, and the ET time, 1/kET, were calculated by using eq 1, where the driving force, ave ∆G, is –0.44 eV and the reorganization energy, λ, is 1.21 eV.20 The values of ET and ave 1/ET estimated by the GMH method (the bridge GF method) at T = 300 K were 2.59 (2.48)

s–1 and 386 (403) ps, respectively, in rough agreement with the experimental value (170–250 ps) in the literature.6,7,20 We constructed five different models to evaluate the influence of each amino acid residue to DA (Figure 5). We show the average value of DA , kET, and 1/kET, for each system, respectively in Table 3. DA of the system without all of the bridge (b) decreased by 7.86 cm–1 from the full system (a). Furthermore, these systems without Adenine (c) or Met353 (d) or Asn349 (e) decreased by 1.50, 3.69, 7.86 cm–1 from full system (a), respectively. The ET time were 386 (a), 635 (b), 420 (c), 480 (d) and 635 ps (e), respectively. 10 ACS Paragon Plus Environment

Page 10 of 32

Page 11 of 32 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

The ET time without Asn349 increased approximately by 250 ps from the full system (a). Taking these observations into account, we concluded that Asn349 has the largest influence on TDA. Electron tunneling pathways To analyze the electron tunneling pathways via the bridge, we performed FMO– RHF/6-31G(d) calculations for the 20 snapshots. Furthermore, we calculated the FMO– LCMO Hamiltonian, and obtained the tunneling current, >L,M , using eq 14 (see Table 4). We obtained five pathways, Adenine/Asn349, Adenine/Glu283, Adenine/Glu283/Asn349/Met353, Asn349/Met353, and Asn349 (Figure 6). In previous theoretical studies, the pathways via adenine and Met353 were already reported.9-13,17,19 In current study, we observed not only the pathways via adenine and Met353 but also the pathways via Glu283 and Asn349. Note that Glu283 and Asn349 form hydrogen bonds with CPD, indicating that these residues are important in active site formation. The average >L,M for the three major pathways from the donor via Asn349, Met353 and Adenine, were respectively 1.445 ± 0.533, 0.431 ± 0.454 and –0.692 ± 0.812, indicating that the pathway via Asn349 was dominant. Next, we analyzed the pathways to the acceptor. As a result, we indicated four major pathways to CPD from Glu283, Asn349, Met353 and adenine, with >L,M value of –0.082 ± 0.381, 3.776 ± 2.165, –0.926 ± 0.963, –1.596 ± 1.437, respectively. Among them, the pathway from Asn349 to CPD was dominant. Asn349 played the major role in mediating ET from HQ to CPD (Figure 7). In the current study, we identified multiple electron tunneling pathways. Note that the amino acid residues (except the methionine residue) in the bridge part, are also found in cryptochrome-DASH. As such, one might expect similar ET reaction in cryptochrome-DASH as well. In reality, however, that is not the case in cryptochrome-DASH, indicating that the methionine residue plays an essential role in the formation of the active site suitable for ET reaction.27 Comparison with the previous results In 2011, Domratcheva68 performed the electronic excited-state calculations of a DNA photolyase model at (I-A) CIS, (I-B) CASSCF, and (I-C) XMCQDPT2 levels, whereas we performed our calculations at (II-A) CIS, (II-B) TD-B3LYP, and (II-C) TD-LC-BLYP levels. In this study, we added Met353 to the QM region of Domratcheva's model and also considered the electrostatic effect of the surrounding environment represented as a distribution of atomic partial charges. By comparing these different calculation methods (Table S3), we see that the MOs of S1 (S2) exhibited LE states of HQ by I-A, I-C, II-A, and 11 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 32

II-C (I-A, II-A and II-C). At a closer look at the MOs of the S1 and S2 states, they exhibit the displacement of electron population from the distal to the proximal side of HQ form that of the ground state with the calculation methods (I-B) and (II-A). The values of the S1 (second lowest LE) state energy were respectively 2.65 and 2.72 (3.32 and 3.22) eV by (I-C) and (II-C), indicating these two methods provide quantitatively similar energy levels (Table S3). Among these six different methods, (II-B) the TD-B3LYP method gave considerable different results from those by the other methods. It would be interesting to compare (I-A) and (II-A) because these two adopted identical methods with different models. The MOs (energy level) of the S3 state were the LE of HQ and the CT state of CPD (5.91 and 5.51 eV) by I-A and II-A, respectively. The existence of Met353 might affect the nature of the S3 state. We compared the MOs obtained by the different methods. The lowest two singlet excited states of HQ (S1 and S4) are of π → π* character, and they exhibited the displacement of electron amplitude from the distal side to the proximal side of HQ in the XMCQDPT2 calculation.68 At a lower energy S2 and S3, intense electron amplitudes were observed on the 3’ or 5’ side of CPD. In our study, Figure S2 shows the orbital which were calculated using CIS/6-31G(d), TD-B3LYP/6-31G(d) and TD-LC-BLYP/6-31G(d). As a result of the CIS/6-31G(d) calculations, S1 and S2 exhibit the displacement of electron amplitude from the distal side to the proximal side of HQ, and S3 does electron amplitude on the 3’ side of CPD. On the other hand, as a result of the TD-B3LYP/6-31G(d) calculations, S1 and S2 exhibit electron amplitude on the 3’ side or the 5’ side of CPD, and S3 does electron amplitude on the adenine

moiety.

The

MOs

and

the

excited

state

energies

obtained

by

the

TD-LC-BLYP/6-31G(d) method were similar to those obtained by the CIS and XMCQDPT2 method, respectively. In summary, the CIS method demonstrated the excited state that is consistent with the other high level methods in the literature. As for the ET mechanism, it has been proposed that the ET reaction takes place via the 5' side of CPD by the experimental study8 based on the comparison of the ET time between different substrates, i.e., T< >U (85 ps) and U< >T (63 ps). However, this logic does not rule out the possibility that the ET reaction takes place via the 3' side of CPD because the ET time of T< >U is faster than that of T< >T (250 ps). The variation of the ET times of these pyrimidine dimers partly arose from the difference of the driving forces, where the driving force of T< >T, T< >U, and U< >T are –0.44, –0.57, and –0.61 eV, respectively. According to Liu et al., the preferred electron transfer to the 5' side is expected rather than that to the 3' side, considering the 40mV difference of the driving forces between U< >T and T< >U. However, they assumed the same electronic coupling for each pyrimidine dimer. It has been widely accepted that the electronic coupling fluctuates significantly by the thermal 12 ACS Paragon Plus Environment

Page 13 of 32 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

fluctuation of the surrounding environment. Probably, the electron transfer to the 5' side is preferred more than that to the 3' side. However, it may be difficult to exclude the possibility of the electron transfer to the 3' side. Furthermore, our calculations also support the possibility of the ET via the 3' side of CPD. As we have seen in Results, we observed multiple

electron

tunneling

pathways

via

(Adenine/Asn349),

(Adenine/Glu283/Asn349/Met353), (Asn349/Met353), and (Asn349), and the ET reaction takes place to the 3' side of CPD in the Asn349/Met353, Asn349 pathways. Also, we observed the dominant role of Asn349 in the ET reaction. The calculated ET time for the full QM region (386 ps) became significantly slower by the elimination of Asn349 from the QM region. Taking these observations into account, it is likely that the ET reaction takes place via both 5' and 3' sides of CPD. The majority of the studies of the ET reaction of CPD photolyase in literatures have supported the direct route and/or the route via adenine for the ET pathways.6-11,13,15-22,24,25 Recently, the results that support the adenine-mediated superexchange mechanism19 and the sequential electron hopping mechanism17 were reported, respectively. However, Macfarlane and Stanley,69 and Kao et al.6 found that the ET does not occur from isoalloxazine ring to adenine moiety of FADH– without DNA substrate. Therefore the electron hopping mechanism was ruled out. In addition, each previous calculation system included FADH– and CPD, but the surrounding amino acid residues were not included.11,17,19 Our calculation system includes the surrounding amino acid residues, and our results support not only the superexchange mechanism of adenine moiety but also that of amino acid residues. On the other hand, Miyazawa et al. suggested the alternative route via Met353.12 Unfortunately, there has been no direct experimental evidence of the existence of the Met353 route for the ET reaction. In most studies on into subject, it has been proposed that the major mechanism is the direct ET from FADH– to CPD and/or the bridge-mediated ET via adenine moiety. 6-11,13,15-22,24,25

Miyazawa et al. pointed out the other possibilities for the ET mechanism,12

although there has been no direct experimental evidence for their hypothesis reported so far. Interestingly,

however,

Yamada

et

al.

demonstrated

that

the

H365N/H369M/L366R of (6-4) photolyase attained CPD repair activity,

triple 70

mutant,

indicating that

Asn349 and Met353 may play essential roles in the ET reaction. In addition, Sato et al. performed MD simulations for CPD photolyase (PDB entry: 1TEZ) for 150 ns, and observed stable hydrogen bonding between Asn349 and CPD (Figure S5).71 In this study, we also observed stable hydrogen bonds between them, and the hydrogen bond lengths, d1 and d2, were 3.0±0.3 Å and 3.1±0.3 Å, respectively. Therefore, it is likely that Asn349 stably exists near CPD, and mediates ET reaction. 13 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

CONCLUSIONS We studied the electron transfer (ET) mechanism of the cylcobutane pyrimidine dimer (CPD) photolyase derived from A. nidulans by theoretical/computational techniques. The QM region of our model was constructed with HQ, CPD, adenine, Glu283, Asn349, and Met353. The ab initio calculations of the CIS/6-31G(d) level indicated that S1 and S2 were locally excited (LE) states of HQ and S3 was the charge transfer (CT) state from HQ to CPD. We evaluated the tunneling matrix element, TDA, (ET time), by using GMH and the bridge GF ave methods, and the average DA (ET time) were 35.58 and 34.81 cm–1 (386 ps and 403 ps),

respectively. This result is in good agreement with the experimental ET time (250 ps). Furthermore, we evaluated the ET times for models without either adenine, Asn349, or Met353, and the calculated values were 420 ps, 635 ps, or 480 ps. In addition, the theoretical analysis of the electron tunneling pathways also indicated that Asn349 plays a dominate role in the ET reaction. Importantly, we observed significant fluctuations of electron tunneling pathways during thermal fluctuations, and it is indicated that the ET reaction takes place via both 5' side and 3'side of CPD. Associated content Supporting Information The root mean square deviation of all atoms in the protein during the MD simulation (Figure S1), The fluctuations of the electronic coupling matrix elements (Figure S2), The distances between the N61 atom of FADH– and the O4/O4' atom of CPD (Figure S3), The Molecular orbitals of each excited sate for CIS/6-31G(d), TD-B3LYP/6-31G(d), and TD-LC-BLYP/6-31G(d) (Figure S4), The hydrogen bonds between Asn349 and CPD (Figure S5), The atomic coordinates of the QM region of a MD snapshot which gave rise to a typical molecular orbital and the electronic coupling matrix element (Table S1), Excited state calculations by CIS, TD DFT and XMCQDPT2. (Table S2). Author Information Corresponding Author *

E-mail: [email protected]

ORCID Ryuma Sato: 0000-0003-2523-8966 Hirotaka Kitoh-Nishioka: 0000-0002-6210-3641 14 ACS Paragon Plus Environment

Page 14 of 32

Page 15 of 32 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

Notes The authors declare no competing financial interest.

Acknowledgements The work was supported by a MEXT Grant-in-Aid for Scientific Research on Innovative Areas (#22104009) to T.Y. The computations were performed at the computer centers of Nagoya University, and the Okazaki Research Center for Computational Sciences. H.K.-N. also acknowledges support from Collaborative Research Program for Young Scientists of ACCMS and IIMC, Kyoto University. References (1) Kim, S. T.; Sancar, A. Photochemistry, Photophysics, and Mechanism of Pyrimidine Dimer Repair by DNA Photolyase. Photochem. Photobiol. 1993, 57, 895-904. (2) Sancar, A. Structure and Function of DNA Photolyase. Biochemistry. 1994, 33, 2-9. (3) Friedberg, E. C.; Walker, G. C.; Siede, W.; Wood, R. D. DNA Repair and Mutagenesis (ASM Press, Washington DC, 1995). (4) Heelis, P. F.; Hartman, R. F.; Rose, S. D. Photoenzymic Repair of UV-Damaged DNA: a Chemist's Perspective. Chem. Soc. Rev. 1995, 24, 289-297. (5) Heelis, P. F.; Rosemarie, F.; Hartman, F.; Rose, S. D. Energy and Election Transfer Processes in Flavoprotein-Mediated DNA Repair. J. Photochem. Photobiol. 1996, 95, 89-98. (6) Kao, Y.-T.; Saxena, C.; Wang, L.; Sancar, A.; Zhong, D. Direct Observation of Thymine Dimer Repair in DNA by Photolyase. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 16128-16132. (7) Kao, Y.-T.; Saxena, C.; Wang, L.; Sancar, A.; Zhong, D. Femtochemistry in Enzyme Catalysis: DNA Photolyase. Cell. Biochem. Biophys. 2007, 48, 32-44. (8) Liu, Z.; Tan, C.; Guo, X.; Kao, Y.-T.; Li, J.; Wang, L.; Sancar, A.; Zhong, D. Dynamics and Mechanism of Cyclobutane Pyrimidine Dimer Repair by DNA Photolyase. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 14831-14836. (9) Antony, J.; Medvedev, D. M.; Stuchebrukhov, A. A. Theoretical Study of Electron Transfer between the Photolyase Catalytic Cofactor FADH– and DNA Thymine Dimer. J. Am. Chem. Soc. 2000, 122, 1057-1065. (10) Medvedev, D.; Stuchebrukhov, A. A. DNA Repair Mechanism by Photolyase: Electron 15 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Transfer Path from the Photolyase Catalytic Cofactor FADH(-) to DNA Thymine Dimer. J. Theor. Biol. 2001, 210, 237-48. (11) Prytkova, T. R.; Beratan, D. N.; Skourtis, S. S. Photoselected Electron Transfer Pathways in DNA Photolyase. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 802-807. (12) Miyazawa, Y.; Nishioka, H.; Yura, K.; Yamato, T. Discrimination of Class I Cyclobutane Pyrimidine Dimer Photolyase from Blue Light Photoreceptors by Single Methionine Residue. Biophys. J. 2008, 94, 2194-2203. (13) Acocella, A.; Jones, G. A.; Zerbetto, F. What is Adenine doing in Photolyase? J. Phys. Chem. B. 2010, 114, 4101-4106. (14) Kholmurodov, K.; Dushanov, E.; Yasuoka, K. Molecular Dynamics Dimulations of a DNA Photolyase Protein: High-mobility and Conformational Changes of the FAD Molecule at Low Temperatures. Adv. Biosci. Biotechnol. 2012, 3, 169-180. (15) Harbach, P. H. P.; Schneider, M.; Faraji, S.; Dreuw, A. Intermolecular Coulombic Decay in Biology: The Initial Electron Detachment from FADH– in DNA Photolyases. J. Phys. Chem. Lett. 2013, 4, 943-949. (16) Wang, H.; Chen, X.; Fang, W. Excited-State Proton Coupled Electron Transfer between Photolyase and the Damaged DNA through Water Wire: a Photo-Repair Mechanism. Phys. Chem. Chem. Phys. 2014, 16, 25432-25441. (17) Lee, W.; Kodail, G.; Stanley, R. J.; Matsika, S. Coexistence of Different Electron-Transfer Mechanism in the DNA Repair Process by Photolyase. Chem. Eur. J. 2016, 22, 11371-11381. (18) Narayanan, M.; Singh, V. R.; Kodail, G.; Moravcevic, K.; Sranley, R. J. An Ethenoadenine FAD Analog Accelerates UV Dimer Repair by DNA Photolyases. Photochem. Photobiol. 2017, 93, 343-354. (19) Rousseau, B. J. G.; Shafei, S.; Migliore, A.; Stanley, R. J.; Beratan, D. N. Determinants of Photolyase's DNA Repair Mechanism in Mesophiles and Extremophiles. J. Am. Chem. Soc. 2018, 140, 2853-2861. (20) Liu, Z.; Guo, X.; Tan, C.; Li, J.; Kao, Y.-T.; Wang, L.; Sancar, A.; Zhong, D. Electron Tunneling Pathways and Role of Adenine in Repair of Cyclobutane Pyrimidine Dimer by DNA Photolyase. J. Am. Chem. Soc. 2012, 134, 8104-8114. (21) Liu, Z.; Wang, L.; Zhong, D. Dyamics and Mechanisms of DNA Repair by Photolyase. Phys. Chem. Chem. Phys. 2015, 17, 11933-11949. (22) Zhang, M.; Wang, L.; Shu, S.; Sancar, A.; Zhong, D. Bifurcating Electron-transfer Pathways in DNA Photolyases Determine the Repair Quantum Yield. Science, 2016, 354, 209-213. 16 ACS Paragon Plus Environment

Page 16 of 32

Page 17 of 32 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

(23) Wijaya, I. M. M.; Domratcheva, T.; Iwata, T.; Getzoff, E. D.; Kandori, H. Single Hydorogen Bond Donation from Flavin N5 to Proximal Asparagine Ensures FAD Reduction in DNA Photolyase. J. Am. Chem. Soc. 2016, 138, 4368-4376. (24) Zhang, M.; Wang, L.; Zhong, D. Photolyase: Dynamics and Mechanisms of Repair of Sun-Induced DNA Damage. Photochem. Photobiol. 2017, 93, 78-92. (25) Zhang, M.; Wang, L.; Zhong, D. Photolyase: Dynamics and Electron-Transfer Mechanism of DNA Repair. Arch. Biochem. Biophys. 2017, 632, 158-174. (26) Mees, A.; Klar, T.; Gnau, P.; Hennecke, U.; Eker, A. P.; Carell, T.; Essen, L. O. Crystal Structure of a Photolyase Bound to a CPD-like DNA Lesion after In Situ Repair. Science. 2004, 306, 1789-93. (27) Sato, R.; Kitoh-Nishioka, H.; Ando, K.; Yamato, T. Computational Study on the Roles of Amino Acid Residues in the Active Site Formation Mechanism of Blue-Light Photoreceptors. Chem. Phys. Lett. 2015, 633, 247-251. (28) Marcus, R. A.; Sutin, N. Electron Transfers in Chemistry and Biology. Biochimica et Biophysica Acta. 1985, 811, 265-322. (29) Cave, R. J.; Newton, M. D. Generalization of the Mulliken-Hush Treatment for the Calculation of Electron Transfer Matrix Elements. Chem. Phys. Lett. 1996, 249, 15-19. (30) Cave, R. J.; Newton, M. D. Calculation of Electronic Coupling Matrix Elements for Ground and Excited State Electron Transfer Reactions: Comparison of the Generalized Mulliken-Hush and Block Diagonalization Methods. J. Chem. Phys. 1997, 106, 9213-9226. (31) Chen, H.-C.; Hsu, C.-P. Ab Initio Characterization of Electron Transfer Coupling in Photoinduced Systems: Generalized Mulliken-Hush with Configuration-Interaction Singles. J. Phys. Chem. A. 2005, 109, 11989-11995. (32) Kitoh-Nishioka, H.; Ando, K. Fragment Molecular Orbital Study on Electron Tunneling Mechanism in Bacterial Photosynthetic Reaction Center. J. Phys. Chem. B. 2012, 116, 12933-12945. (33) Kitaura, K.; Ikeo, E.; Asada, T.; Nakano, T.; Uebayasi, M. Fragment Molecular Orbital Method: an Approximate Computational Method for Large Molecules. Chem. Phys. Lett. 1999, 313, 701-706. (34) Nakano, T.; Kaminuma, T.; Sato, T.; Fukuzawa, K.; Akiyama, Y.; Uebayashi, M.; Kitaura, K. Fragment Molecular Orbital Method: Use of Approximate Electrostatic Potential. Chem. Phys. Lett. 2002, 351, 475-480. (35) Fedorov, D. G.; Kitaura, K. Extending the Power of Quantum Chemistry to Large Systems with the Fragment Molecular Orbital Method. J. Phys. Chem. A. 2007, 111, 17 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

6904-6914. (36) Tanaka, S.; Mochizuki, Y.; Komeiji, Y.; Okiyama, Y.; Fukuzawa, K. Electron-Correlated Fragment-Molecular-Orbital Calculations for Biomolecular and Nano Systems. Phys. Chem. Chem. Phys. 2014, 16, 10310-10344. (37) Kawatsu, T.; Kakitani, T.; Yamato, T. A Novel Method for Determining the Electron Tunneling Pathway in Protein. Inorg. Chem. Acta. 2000, 300-302, 862-868. (38) Nishioka, H.; Kimura, A.; Yamato, T. Interference, Fluctuation, and Alternation of Electron Tunneling in Protein Media. 1. Two Tunneling Routes in Photosynthetic Reaction Center Alternate due to Thermal Fluctuation of Protein Conformation. J. Phys. Chem. B. 2005, 109, 1978-1987. (39) Lee, J.-H.; Park, C.-J.; Shin, J.-S.; Ikegami, T.; Akutsu, H.; Choi, B.-S. NMR Structure of the DNA Decamer Duplex Containing Double T•G Mismatches of Cis-Syn Cyclobutane Pyrimidine Dimer: Implications for DNA Damage Recognition by the XPC-hHR23B Complex. Nucleic. Acids. Res. 2004, 32, 2474-2481. (40) Landorff-Larsen, K.; Piana, S.; Palmo, K.; Maragakis, P.; Klepeis, J. L.; Dror, R. O.; Shaw, D. E. Improved Side-Chain Torsion Potentials for the Amber ff99SB Protein Force Field. Proteins. 2010, 78, 1950-8. (41) Jorgensen, W.; Chandrasekhar, J.; Nadura, J. D. Comparison of Simple Potential Functions for Simulating Liquid Water. J. Chem. Phys. 1983, 79, 926-935. (42) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C. et al. Gaussian03, revision D.01; Gaussian, Inc.: Wallingford, CT, 2004. (43) Case, D. A.; Darden, T. A.; Cheatham, T. E., III; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.; Walker, R. C.; Zhang, W.; Merz, K. M. et al. Amber 12; University of California: San Francisco, CA, 2012. (44) Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. Numerical Integration of the Cartesian Equations of Motion of a System with Constraints: Molecular Dynamics of n-Alkanes. J. Compt. Phys. 1977, 23, 327-341. (45) Szabo, A.; Ostlund, N. S. Modern Quantum Chemistry: Introduction to Advanced Electronic Structure Theory, 2012 (New York: Dover). (46) Voityuk, A. A.; Rosch, N. Fragment Charge Difference Method for Estimating Donor-Acceptor Electronic Coupling: Application to DNA π-Stacks. J. Chem. Phys. 2002, 117, 5607-5616. (47) Wallrapp, F.; Voityuk, A.; Guallar, V. Solvent Effects on Donor-Acceptor Couplings in Peptides. A Combined QM and MD Study. J. Chem. Theor. Comput. 2009, 5, 18 ACS Paragon Plus Environment

Page 18 of 32

Page 19 of 32 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

3312-3320. (48) Voityuk, A. A. Electron Transfer between [4Fe-4S] Clusters. Chem. Phys. Lett. 2010, 495, 131-134. (49) Stuchebrkhov, A. A. Tunneling Currents in Long-Distance Electron Transfer Reactions. V. Effective One Electron Approximation. J. Chem. Phys. 2003, 118, 7898-7906. (50) Schmidt, M. W.; Baldrige, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S. et al. General Atomic and Molecular Electronic Structure System. J. Compt. Chem. 1993, 14, 1347-1363. (51) Nishioka, H.; Ando, K. Electronic Coupling Calculation and Pathway Analysis of Electron Transfer Reaction using Ab Initio Fragment-Based Method. I. FMO-LCMO Approach. J. Chem. Phys. 2011, 134, 204109-12. (52) Tsuneyuki, S.; Kobori, T.; Akagi, K.; Sodeyama, K.; Terakura, K.; Fukuyama, H. Molecular Orbital Calculation of Biomolecules with Fragment Molecular Orbitals. Chem. Phys. Lett. 2009, 476, 104-108. (53) Kobori, T.; Sodeyama, K.; Otsuka, T.; Tateyama, Y.; Tsuneyuki, S. Trimer Effects in Fragment Molecular Orbital-Linear Combination of Molecular Orbitals Calculation of One-Electron Orbitals for Biomolecules. J. Chem. Phys. 2013, 139, 094113-11. (54) Fedorov, D. G.; Kitaura, K. Many-body Expansion of the Fock Matrix in the Fragment Molecular Orbital Method. J. Chem. Phys. 2007, 147, 104106. (55) Kitoh-Nishioka, H.; Ando, K. Charge-Transfer Matrix Elements by FMO-LCMO Approach: Hole Transfer in DNA with Parameter Tuned Range-Separated DFT. Chem. Phys. Lett. 2015, 621, 96-101. (56) Stuchebrukhov, A. A. Long-Distance Electron Tunneling in Proteins. Theor. Chem. Acc. 2003, 110, 291-306. (57) Skourtis, S. S.; Beratan, D. N. Theories of Structure-Function Relationships for Bridge-Mediated Electron Transfer Reactions. Adv. Chem. Phys. Electron Transf. Isol. Mol. to Biomol. Part1, 1999, 106, 377-452. (58) Regan, J. J.; Onuchic, J. N. Electron-Transfer Tubes. Adv. Chem. Phys. Electron Transf. Isol. Mol. to Biomol. Part2, 1999, 107, 497-553. (59) Kawatsu, T.; Kakitani, T.; Yamato, T. Destructive Interference in the Electron Tunneling thorough Protein Media. J. Phys. Chem. B. 2002, 106, 11356-11366. (60) Curtiss, L. A.; Naleway, C. A.; Miller, J. R. Theoretical Study of Long-Distance Electronic Coupling in H2C(CH2)n-2CH2 Chains, n=3-16. J. Phys. Chem. 1993, 97, 4050-4058. (61) Plato, M.; Michel-Beyerle, M. E.; Bixon, M.; Jortner, J. On the Role of Tryptophan as a 19 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Superexchange Mediator for Quinone Reduction in Photosynthetic Reaction Centers. FEBS. Lett. 1989, 249, 70-74. (62) Stuchebrukhov, A. A. Tunneling Currents in Electron Transfer Reaction in Proteins. II. Calculation of Electronic Superexchnage Matrix Element and Tunneling Currents using Nonorthogonal Basis Set. J. Chem. Phys. 1996, 105, 10819-10829. (63) Kawatsu, T.; Kakitani, T.; Yamato, T. Worm Model for Electron Tunneling in Proteins: Consolidation of the Pathway Model and the Dutton Plot. J. Phys. Chem. B. 2001, 105, 4424-4435. (64) Grimme, S. Calculation of the Electronic Spectra of Large Molecules. Rev. Compt. Chem. 2004, 20,153-218. (65) Dreuw, A.; Head-Gordon, M. Single-Reference Ab Initio Methods for the Calculation of Excited States of Large Molecules. Chem. Rev. 2005, 105, 4009-4037. (66) Skourtis, S. S.; Prytkova, T.; Beratan, D. N. Flavin Charge Transfer Transitions Assist DNA Photolyase Electron Transfer. AIP. Conf. Proc. 2007, 963, 674-677. (67) Zheng, X.; Ly, N. M.; Stuchebrukhov, A. A. Photoactivated Excited States of DNA Repair Photolyase: Dynamical and Semiempirical Identification. Int. J. Quant. Chem. 2007, 107, 3126-3131. (68) Domratcheva, T. Neutral Histidine and Photoinduced Electron Transfer in DNA Photolyase. J. Am. Chem. Soc. 2011, 133, 18172-18182. (69) Macfarlane, A. W., IV; Stanley, R. J.; Cis-syn Thymidine Dimer Repair by DNA Photolyase in Real Time. Biochemistry. 2003, 42, 8558-8568. (70) Yamada, D.; Dokainish, H. M.; Iwata, T.; Yamamoto, J.; Ishikawa, T.; Todo, T.; Iwai, S.; Getzoff, E. D.; Kitao, A.; Kandori, H. Functional Conversion of CPD and (6-4) Photolyases by Mutation. Biochemistry. 2016, 55, 4173-4183. (71) Sato, R.; Harada, R.; Shigeta, Y. The Binding Structure and Affinity of Photodamaged Duplex DNA with Members of the Photolyase/Cryptochrome Family: A Computational Study. Biophys. Physicobiol. 2018, 15, 18-27.

20 ACS Paragon Plus Environment

Page 20 of 32

Page 21 of 32 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

Table 1. The lowest three excited states of the QM region for a MD snapshot. State

CI vector (Top two coefficients are shown)

Oscillator Strength

S1

–0.814 (HOMO → LUMO+9),

– 0.358 (HOMO → LUMO+11)

0.0284

S2

0.757 (HOMO → LUMO+11),

– 0.382 (HOMO → LUMO+9)

0.2152

S3

0.996 (HOMO → LUMO),

– 0.059 (HOMO–1 → LUMO)

0.0004

Table 2. Average values of the electron coupling matrix element, TDA, and the electron transfer rate constant, kET. Methods

ave DA [cm–1]

ave ET × 10N [s–1]

ave 1/ET [ps]

GMH

35.58

2.59

386

bridge GF

34.81

2.48

403 21

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

Expt.[8,17]

170 – 250

Table 3. Model-dependence of TDA and kET. (a) full system (see Figure 1). Within the QM region, (b) all the atoms in the bridge region, (c) adenine, (d) Met353, and (e) Asn349 were replaced by atomic partial charges. Here, the driving force and the reorganization energy are –0.44 eV and 1.21 eV, respectively. Systems

ave DA [cm–1]

ave ET × 10N [s– 1

ave [ps] 1/ET

]

(a)

35.58

2.59

386

(b)

27.72

1.57

635

(c)

34.08

2.38

420

(d)

31.89

2.08

480

(e)

27.71

1.57

635

Table 4. Tunneling current KL,M obtained by the bridge GF/FMO-LCMO method. Route

Tunneling current

HQ → Asn349

1.445 ± 0.533

HQ → Met353

0.431 ± 0.454

HQ → Adenine

–0.692 ± 0.812

Asn349 → CPD

3.776 ± 2.165

Met353 → CPD

–0.926 ± 0.963

Adenine → CPD

–1.526 ± 1.437

Glu283 → CPD

–0.082 ± 0.381

Adenine → Glu283

0.172 ± 0.383

Asn349 → Met353

–1.364 ± 1.330

Asn349 → Adenine

–0.730 ± 0.766

22 ACS Paragon Plus Environment

Page 22 of 32

Page 23 of 32 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

23 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 1. Model structure. Here Glu283, Asn349, Met353, Adenine, HQ and CPD are explicitly included in the quantum mechanics (QM) region, while the rest of the system are represented as a distribution of atomic partial charges.

24 ACS Paragon Plus Environment

Page 24 of 32

Page 25 of 32 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

Figure 2. Characterization of excited-state MOs and reaction scheme. The distal (proximal) side of flavin is mostly populated in the S0 (S1 and S2) state, while the S3 state is the charge transfer state from flavin to CPD. After photoexcitation from S0 to S2, then the reaction proceeds from S2 to S1 by internal conversion, and then it does from S1 to S3 by charge transfer.

25 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 3. Orbital energy spectra. Energy spectrum of RHF and FMO-LCMO are very similar to each other. Here, LUMO of HQ and CPD of RHF (FMO-LCMO) are 4.78522 (4.77389) and 6.96754 (6.94241) eV, respectively.

26 ACS Paragon Plus Environment

Page 26 of 32

Page 27 of 32 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

Figure 4. Frontier MOs. (a) LUMO and (b) LUMO+8 obtained by CIS/6-31G(d) calculations. LUMOs of (c) HQ and (d) CPD obtained by FMO2-RHF/6-31G(d) calculations. These two different calculation methods provided similar results.

27 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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. Model structures. (a) full system of QM region. All the atoms in (b) the bridge region, (c) Adenine, (d) Met353, and (e) Asn349 were replaced by the atomic partial charges.

28 ACS Paragon Plus Environment

Page 28 of 32

Page 29 of 32 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

Figure 6. The electron tunneling pathways. Pathway from HQ to CPD via (a) Adenine/Asn349,

(b)

Adenine/Glu283,

(c)

Adenine/Glu283/Asn349/Met353,

Asn349/Met353, and (e) Asn349.

29 ACS Paragon Plus Environment

(d)

The Journal of Physical Chemistry 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 7. The averaged electron tunneling pathways.

30 ACS Paragon Plus Environment

Page 30 of 32

Page 31 of 32 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

TOC graphic

31 ACS Paragon Plus Environment

The Journal of Physical Chemistry 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

ACS Paragon Plus Environment

Page 32 of 32