The Sequence Dependence of Photoinduced Single Strand Break in 5

Subscriber access provided by UNIVERSITY OF THE SUNSHINE COAST

Article

The Sequence Dependence of Photoinduced Single Strand Break in 5-Bromo-2’-Deoxyuridine Labeled DNA Supports Electron Transfer to be Responsible for the Damage Kinga Westphal, Samanta Makurat, and Janusz Rak J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.7b07338 • Publication Date (Web): 11 Sep 2017 Downloaded from http://pubs.acs.org on September 15, 2017

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

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

Page 1 of 22

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 Sequence Dependence of Photoinduced Single Strand Break in 5-Bromo-2’-Deoxyuridine Labeled DNA Supports Electron Transfer to be Responsible for the Damage Kinga Westphal, Samanta Makurat, and Janusz Rak* Faculty of Chemistry, University of Gdańsk, Wita Stwosza 63, 80-308 Gdańsk, Poland

Corresponding Author *E-mail: [email protected]

1 ACS Paragon Plus Environment


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

Page 2 of 22

ABSTRACT The UVB irradiation of DNA labeled with 5-bromo-2’-deoxyuridine (BrdU) leads to single strand breaks (SSBs) as a major photochemical damage. Some time ago we demonstrated that SSB is a secondary damage forming due to thermal dissociation of 2’-deoxyribonolactone generated photochemically in DNA labeled with BrdU. For the first time, we study here the variation of the yield of UVB generated SSBs with the alteration of 3’-neighbour nucleobase of electron donor (2’-deoxyguanine (dG)) and acceptor (excited BrdU) in double stranded DNA. We showed that the experimental damage yields can be explained by the calculated ionization potentials of dG and electron affinities of excited BrdU via a kinetic scheme based on the Marcus model of electron transfer (ET). Hence, our studies on the sequence dependence of photochemical damage in DNA labeled with BrdU constitute a further argument that photochemically generated SSBs occur as a result of long-range ET.


Page 3 of 22

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


INTRODUCTION The main goal of anticancer therapy is degradation of cellular DNA in tumour cells.1,2 One of the modern methods of cancer treatment is photodynamic therapy (PDT).3 This modality requires the presence of a selected photosensitizing agent and exposure to light, usually form the visible region.4 The ideal photosensitizer (PS) should have a negligible dark toxicity while be characterized by a significant cytotoxicity in the presence of light. The latter is frequently related to high quantum yield of singlet oxygen formation.5 In fact, most of the commonly used photosensitizers are oxygendependent and exert their photo-toxic action (i) via reactive radicals which react with oxygen dissolved in the cell (type I)3,4 or by the formation of singlet oxygen (type II).3 In opposition to the above mentioned type of PS, the modified nucleosides, such as 5-bromo-2’-deoxyuridine (BrdU), do not need oxygen for DNA damage. This is especially important because solid tumor cells suffer from oxygen deficiency (hypoxia).6 Incorporation of such modified nucleosides into DNA strand effectively sensitizes cancer cell to UV photons with energies >290 nm in the lack of oxygen.7 It is believed that single strand break (SSB) is formed as a result of UVB-induced longrange electron transfer (ET) from a distant 2’-deoxyguanine (dG) to the photo-excited BrdU.8-15 In the next step, the reactive uridine-5-yl radical occurs as a consequence of dissociation of a short-living BrdU anion.16-18 The above mentioned uridine-5-yl radical may stabilize itself by the detachment of a hydrogen atom from the adjacent sugar residue, which ultimately leads to single strand break formation.19 SSB does not directly induce cell death and is easily repaired via the single strand break repair (SSBR) pathway.20 However, such a damage can lead to highly cytotoxic double strand breaks (DSBs) in two cases: (i) as a result of temporary nick formed in the vicinity of photodamage during base excision repair (BER) or (ii) as a consequence of SSB



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

Page 4 of 22

formation in the replication fork during S phase of cell cycle.21 Thus, SSBs may be precursors of highly cytotoxic DSBs and the concept of using a BrdU type photosensitizer incorporated into the cellular DNA can be regarded as a “Trojan horse” therapy.22 The sensitizing properties of modified nucleosides have been studied already for several decades.23,24 Nevertheless these derivatives are still not in use in PDT. One of the reasons accounting for this situation is probably the fact that the sensitizing action of the BrdU-type modified nucleosides strongly depends on the DNA sequence in their proximity. Indeed, in the photolytic experiments carried out under anaerobic conditions, Greenberg et al.25 observed that strand scission in the 5’-ABrU sequence was generated ca. eight-fold more efficiently than in the 5’-GBrU one. Recently, the 5’-GAABrU sequence rather than the 5’-ABrU one was determined as a “hot spot”, the presence of which in dsDNA guarantees biopolymer’s photosensitivity.12 A bridge of nucleobases between the electron donor (D; dG) and acceptor (A; the excited modified nucleobase) has significant influence on the efficiency of damage, that was confirmed in a number of photolytic experiments on model labeled oligonucleotides.12-26,27 A strong argument for an electron transfer nature of the photochemical process triggered in the labeled DNA are the experiments carried out by the Sugiyama group.11,26 By mutation of the electron donor in the hot spot sequence, they showed that the damage yield remains in a straightforward correlation with its ionization potential.12 Furthermore, the stability of charge separated state, also affects photoinduced damage. Indeed, the reversed hot spot sequence, 5’-BrUAAC-3’, is also photosensitive but its damage yield is lower than that of the hot spot one.7 Moreover, the same group demonstrated that the length of the bridge (the number of AT base pairs separating the photoexcited BrdU from dG) influences the damage yield


Page 5 of 22

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


exponentially,26 thus in a manner which accounts for the superexchange-mediated electron transfer mechanism.28 Thus, in order to generate SSBs photochemically the irradiated oligonucleotide should comprise a modified electrophilic nucleoside, which forms an unstable anion, and is separated from dG in the same or opposite strand by a bridge of 2-3 AT base pairs. It is, however, still unclear how the neighborhood of dG (electron donor) and the modified electrophilic nucleoside (BrdU; electron acceptor) affects the yield of photodamage. Here, for the first time, we study in a systematic manner the effect of nucleobases, adjacent to dG and BrdU, on the photochemical cleavage. In the following, we will demonstrate that change in the ionization potential of dG and in the electron affinity of BrdU, due to interactions with neighboring bases, accounts well for the observed variation in the yield of damage.

EXPERIMENTAL AND THEORETICAL METHODS DNA sequences. All seven sequences labeled with fixed motif 5'-XCAABrUY-3’ were chemically synthesized and purified with HPLC by Metabion, Germany. The four variants of the mentioned above motif, 5'-XCAABrUC-3’, arising from the variation of X led to four single stranded 25 base oligonucleotide, while three other sequences resulting from modification of Y (where Y= G, T, A) - see Scheme 1.



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 22

Scheme 1. DNA sequences labeled with 5’-XCAABrUY-3’ motif employed in the current study and fragments formed after UV irradiation

Annealing. All single stranded sequences were annealed with the appropriate complementary strand to form double stranded labeled 25bp DNA. Hybridization was carried out by mixing an equal amount of complementary oligonucleotides in K/PO4 and NaCl buffer (10mM, pH = 7.5), heating the solution to 90 oC and incubating it for 1 min, and then cooling slowly (for 10 min) to 50 oC. The sample containing the hybridized oligonucleotides was kept at 4 oC.

Irradiation conditions. The DNA solution was freshly prepared prior to irradiation. Photolysis was carried out in quartz capillaries (3 x 3 mm) filled with the DNA solution in concentration of 10 µM and a total volume of 50 µl with a 500 W high-pressure mercury lamp for 15 min at 320 nm.

Liquid

chromatography

and

mass

spectrometry

(LC-MS)

conditions.

Chromatographic separations of DNA were carried out using an Ultra High Performance Liquid Chromatography (UHPLC) system Nexera X2 with a binary solvent manager (Shimadzu, Japan). The analytes were separated at a flow rate of 0.1 mL/min on a 1.7 µm Acquity UPLC BEH C18 1.0 x 50 mm column (Waters, USA). The mobile phase A consisted of 400 mM HFIP and 50 mM TEA in deionized water, and mobile phase B consisted of the same concentration of HFIP and TEA in water:methanol (50:50, v/v). A 2 µl injection of sample was loaded onto the column and separated using the following gradient conditions [time (min), % mobile phase B]: (0,0) (10, 15) (40, 40). The column temperature was maintained at 85 oC. The effluent was diverted to waste for 2 min after injection. The UHPLC was coupled directly to a TripleTOF 5600+ mass spectrometer (AB 6 ACS Paragon Plus Environment

Page 7 of 22

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


SCIEX, USA) equipped with a duo-electrospray ion source operated in negative ionization mode. MS operation parameters were as follows: the spray voltage was 4.5 kV, the nebulizer gas (N2) pressure was 25 psi, the flow rate was 11 L min-1 and the source temperature was 300 oC. Spectrum was obtained by averaging 3 scans, and the time of each scan was 0.25 s. Models and computational methods. The geometries (see Figure 1) corresponding to the regular B-DNA form were obtained with the use of x3DNA software.29 Hydrogen atoms were added using Gromacs software.30 The phosphates were protonated manually, and the hydrogen positions were optimized with the employed DFT method. It was shown before that such protonated phosphate model is as successful as the model with explicit counterions but less computationally demanding.37 Single-stranded, short DNA sequences of 5’-GX-3’ for VIP calculations and 5’-BrUX-3’ for VEA calculations were used (5BrdU was constructed by replacing the 5-methyl group of thymidine by Br atom of optimized bond length). Further enlarging of the system would not change the results much, while making the calculations more demanding.31

Figure 1. Visualization of exemplary (A) 5’-GG-3’ and (B) 5’-BrUT-3’ structures as used for calculations. The other systems were prepared in the same manner, only the 3’ nucleobase (bold) was changed. 7 ACS Paragon Plus Environment


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

Page 8 of 22

Single point calculations were carried out using the Gaussian09 package.32 Density Functional Theory (DFT) employing the 6-31++G(d,p) basis set and B3LYPD3 functional, with the empirical correction for dispersion energy by Grimme33 for VIP and PBE0 for VEA in Time Dependent DFT (TD-DFT) approach were used as the methods of choice as suggested in the literature.34-37 The first excited state of BrdU was taken into account. Water environment was simulated with the PCM solvation model.38 Structure visualizations were carried out with the GaussView539 and VMD40 packages.

RESULTS AND DISCUSSION We designed seven double–stranded sequences resulting from the mutation of X and Y (X,Y = G, C, A or T) nucleotides within the model 5'-XCAABrUY-3’ motif. Two AT pairs in this sequence act as a bridge for a long-range inter-strand electron transfer from the distant 2’-deoxyguanine to the excited BrdU, while the 3’-flanking nucleobase complementary to the X residue affects the ionization potential of dG (located in the complementary strand). On the other hand, the Y residue adjacent to BrdU from the 3’-side has an impact on the electron affinity of the bromoderivative. The UV irradiation of an oligonucleotide solution leads to a well-defined SSB41 and the LC-MS separation of the photolyte results in four ssDNA fragments: the labeled and complementary strand as well as two oligonucleotides, 11 (ssP11) and 13 (ssP13) bases in length phosphorylated at the 3’ and 5’ end, respectively (Table 1), being the products of SSB (for the exemplar LC-MS chromatograms see Figure S1 in Supplementary Information (SI)).


Page 9 of 22

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


Table 1. Sequences of fragments formed after UV irradiation.

5’-XCAABrUC-3’ X A C T G Y T G A

ssP11

ss13P 5’-TCA TGC GAG CAC A-P-3’ 5’-TCA TGC GAG CCC A-P-3’ 5’-P-U CAT CAC GTG T-3’ 5’-TCA TGC GAG CTC A-P-3’ 5’-TCA TGC GAG CGC A-P-3’ 5’-CCAABrUY-3’ ssP11 ss13P 5’-P-U TAT CAC GTG T-3’ 5’-P-U GAT CAC GTG T-3’ 5’-TCA TGC GAG CCC A-P-3’ 5’-P-U AAT CAC GTG T-3’

The extracted ion chromatograms (XICs) corresponding to the ssP11 fragment recorded for the irradiated duplexes are shown in Figure 2a.

Figure 2. XIC of ssP11 (a) and ss13P (b) fragments formed after UV irradiation of dsDNA labeled with 5’-XCAABrUY-3’ motif. 9 ACS Paragon Plus Environment


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

Page 10 of 22

In order to determine the yield of SSB formation by high resolution LC-MS, we obtained the standard curve using the chemically synthesized ssP11 oligonucleotide (see Figure S2 in Supplementary Information (SI)). This calibration curve and the actual integrated XIC values measured for the particular photolytes were used to calculate the relative amount of ssP11 in the irradiated samples (see Scheme 1). The damage yields span a range from 12 to 20% (see Table 2) which demonstrates a significant influence of 3’-neigbour of guanine on the process under consideration.

Table 2 Percentage of damage (SSBs) in the samples irradiated with 320 nm photons. Relative standard deviations (in %) are given in parentheses. 5’-XCAABrUC-3’ A

G

C

T

% of

12.12

17.40

19.96

13.5

damage

(0.06)

(0.07)

(0.31)

(0.6)

5’-CCAABrUY-3’ % of

8.37

2.05

24.94 -

damage (0.005)

(0.02)

(0.35)

As indicated by figures gathered in Table 2 dsDNA containing 5’-CCAABrUC-3’ motif exhibits the highest photosensitivity among the four examined duplexes. Therefore, we employed this motif to study the influence of 3’ neighbor of BrdU on the photodamage. The irradiation and LC-MS analysis of 25bp dsDNA labeled with 5’-CCAABrUY-3’ (where Y=G, T or A) was carried out as described in the experimental section. XICs for all 3 duplexes are depicted in Figure 2b. In order to 10 ACS Paragon Plus Environment

Page 11 of 22

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


quantify the amount of SSB we used the standard curve (see Figure S2 in SI) obtained using the solutions of chemically synthesized ss13P oligonucleotide. Of the three analyzed duplexes, the irradiation of dsDNA labeled with 5’-CCAABrUT-3’ motif leads to the formation of the largest amount of SSBs. For Y=T almost 25% of the oligonucleotide undergoes damage under the employed experimental conditions (see Table 2 and Figure 2b), while only 2% forms SSBs when Y=G, which demonstrates that the photochemical process is more sensitive to the neighborhood of electron acceptor than to that of electron donor. In order to explain our experimental results at the molecular level, we employed a quantum chemistry approach. A number of experiments carried out so far suggest that long-range electron transfer is responsible for the studied process.42 On the other hand, the neighborhood of electron donor and acceptor modifies their ionization potential and electron affinity, respectively, which, in turn, changes the free energy of electron transfer, ∆GET that, according to the Marcus theory, is one of primary parameters influencing the rate of the process.43 This is why we calculated the vertical ionization potentials (VIPs) and electron affinities (VEAs) for dG and the excited BrdU, respectively, in all irradiated sequences. In order to account for the nearneighborhood effects we considered all possible dimers for dG and BrdU (see Figure 1 for the exemplar dimeric geometries). As a consequence, single-stranded sequences of 5’-GX-3’ for VIP calculations and 5’-BrUX-3’ for VEA calculations were used. For the considered 5’-GX pairs the VIP of guanine increases in the following order: GG(5.87 eV) < GC(5.97 eV) < GA(5.98 eV) < GT(6.02 eV), while the order of VEA for the excited BrdU in the corresponding 5’-BrUX dimers is as follows: BrUG(5.84 eV) < BrUA(6.19 eV) < BrUC(6.48eV) < BrUT(6.49 eV). It is worth noticing that the



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 22

VIP-VEA difference for a given electron donor-electron acceptor (D-A) pair estimates ∆GET – a driving force for electron transfer. One should emphasize that the employed methodology leads to reasonable results as demonstrated by comparison with the previous calculations, where the VIPs of two nucleobase pairs were calculated at the EOM-IP-CCSD/6-31+G(d,p) level of theory. Indeed, the B3LYP-D3/6-31++G(d,p) VIPs for our 5’-GX-3’ dimers compare very well to the EOM-IP-CCSD/6-31+G(d,p) values. Both methods predict the same order of the VIPs for the discussed dimers.44 Similarly, our VEAs for 5’-BrUX follows the trend predicted by the AEAs (Adiabatic Electron Affinities) of thymine and cytosine in 5’-XTY45 and 5’-XCY46 timers calculated at the MP2/CCSD(T) level of theory. Despite the fact the our VEAs concerns the vertical electron attachment to the excited BrdU while literature data describe the adiabatic formation of the T or C anions, both approaches predict that 3’-pyrimidines stabilize the occurring anion better than 3’-purines. If electron transfer does govern the studied damage process the damage yield should be described by the Marcus theory.43 Let us assume the following kinetic scheme:

→ ∗

absorption

(1)

∗ electron transfer

(2)

∗ internal conversion (3) where DA, DA*, D+A-, Ia, kET and kIC stand for the ground state of the labeled DNA molecule, the locally excited state (excitation localized to BrU), charge separated state, rate of photon absorption, rate constant of electron transfer, and rate constant of internal conversion, respectively. Thus, the rate of charge separation, vD+A-, is given by eq. (4):


Page 13 of 22

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


= ∗ (4) The stationary concentration of DA*, [DA*], can be estimated using the steady-state approximation: !

− ∗ − # ∗ = 0 (5)

∗ =

!

+ #

(6)

As mentioned above, the overall kinetics of damage formation should be determined by the rate of electron transfer in our kinetic model. Hence:

−

() = = (* + #

!

≅

#

! if

kIC ≫kET (7)

The rate of photon absorption, equal to the intensity of light absorbed by the molecules per unit volume, can be expressed as follows:

!

=

-.

/ ℎ12

=

. 4 (1 − 1067 87 9 ) = 4 (1 − : ;7 87 9 ) (8) / ℎ32 (

where Is, NA, h, ν, V, S, d, I0, and εA denote intensity of absorbed light, Avogadro’s constant, Planck’s constant, frequency of incident light, absorbing volume, irradiated area, optical length equal to V/S, incident light intensity, and molar absorption coefficient, respectively; κA=2,303εA and k=1/(NAhν). For κAcAd

The Sequence Dependence of Photoinduced Single Strand Break in 5

Recommend Documents