Kinetic Study of DNA Modification by Phthalocyanine Derivative of the Oligonucleotide

Design of chemically modified oligonucleotides for regulation of gene expression has attracted considerable attention over the last decades. One actively pursued approach involves antisense or antigene constructs carrying reactive groups, many of these based on transition metal complexes. The complexes of Co(II) with phthalocyanines are extremely good catalysts of oxidation of organic compounds with molecular oxygen and hydrogen peroxide. In this study, we have investigated the kinetics and thermodynamics of sequence-specific modification of DNA with deoxyribooligonucleotide linked to Co(II)-tetracarboxyphthalocyanine (PtcCo(II)) in the presence of H2O2.


INTRODUCTION
The principles underlining antisense and antigene strategies are conceptually very simple and straightforward. The use of a complementary sequence can inhibit the expression of a specific mRNA, breaking the transfer of genetic information from DNA to protein. The development of oligonucleotide derivatives that can bind sequence specifically to unique sites in mRNA or genomic DNA and modify the target to a great extent or even completely may have major implications for the treatment of hereditary diseases, cancers, and viral infections [1][2][3][4]. This approach, relying on sequence-specific targeting of reactive compounds, was initially called "complementary addressed modification of nucleic acids" [5]. Kinetic studies of these processes provide quantitative estimates of the selectivity and efficiency of the modification of nucleic acids.
Oligonucleotides derivatives carrying catalytic groups can achieve multiple turnover and convert many target molecules. Complexes of porphyrins and phthalocyanines with the transition metal ions (Fe, Co, etc.) are considered among the most efficient catalytic groups for this purpose [6].
Molecular oxygen is a four-electron oxidant; its reduction to H 2 O is presented in Scheme 1. As one-step two-or four-electron reduction of O 2 is forbidden by spin exclusion principle, direct oxidation of organic substrates with a triplet oxygen molecule does not occur. This difficulty can be overcome either by converting the oxygen molecule from its ground triplet state to the excited singlet state or by sequential one-electron reduction to H 2 O catalyzed by transition metal ions. Among reactive oxygen species formed in the latter pathway (Scheme 1), hydroxyl radical • OH is the strongest oxidant [7,8] capable of damaging various cell constituents including DNA [9].
It was recently shown that complexes of phthalocyanines with Co(II) and Fe(II) are very efficient catalysts of oxidation of various organic substrates with molecular oxygen and hydrogen peroxide [6]. If molecular oxygen is the oxidant, a reducing agent is also required to convert the metal ion to a low valent state.
The main goal of this paper was to determine the kinetic features of the interaction of the PtcCo(II)-oligonucleotide conjugate with single-stranded DNA. Earlier we have shown that the system of O 2 , a PtcCo(II) conjugate, and a reductant can oxidize DNA [10], but this reaction is very slow. Therefore, in the present study H 2 O 2 was taken as an oxidant instead of molecular oxygen. In this case, the first unfavorable step in Scheme 1 is left out and the target oxidation is accelerated significantly, allowing one to estimate the contribution of other stages to the oxidation process. In addition to being the source of • OH radicals, H 2 O 2 serves as the reducer of the oxidized form of catalyst. The structures of the metallophthalocyanine conjugate and the target DNA are presented in Figure 1.
To gain a deeper insight into the reactivity of PtcCo(II)group and the mechanism of the DNA target modification by the PtcCo(II)-oligonucleotide conjugate, we have studied separately the different stages of this process. The first step was duplex formation between the target and the conjugate (X) or a nonmodified oligonucleotide (N). This equilibrium was studied by stopped-flow kinetics and UV melting curve analysis. Degradation of the PtcCo(II) residue in the conjugate in the presence of H 2 O 2 was detected by changes in the absorption spectrum of this moiety during the reaction. The products of the catalytic oxidative modification of the target deoxyribooligonucleotide were registered by gel electrophoresis after treatment with piperidine (to reveal alkali-labile sites) or Escherichia coli Fpg protein (to reveal 8oxoguanine and abasic sites).

Oligonucleotides and the conjugate
The 20 nt and 10 nt deoxyribonucleotides d(AATGGGAA-GAGGGTCAGGTT), d(TCTTCCCATT), and pd(TCTT-CCCATT) were synthesized on an ASM-700 automated synthesizer (Biosset, Russia) from phosphoramidites purchased from Glen Research (USA) according to the manufacturer's protocol. The oligonucleotides were deprotected with ammonium hydroxide and purified by ion exchange HPLC on a Nucleosil 100-10 N(CH 3 ) 2 column followed by reversephase HPLC on a Nucleosil 100-10 C 18 column (both 4.6 × 250 mm, purchased from Macherey-Nagel, Germany). The purity of the oligonucleotides exceeded 98%, as estimated by electrophoresis in 20% denaturing polyacrylamide gel and staining with Stains-All dye (Sigma-Aldrich, USA). Concentrations of the oligonucleotides were determined from their absorbance at 260 nm [12].
The conjugate PtcCo(II)-NH-(CH 2 ) 6 -O-pd(TCTTCC-CATT) was synthesized using a previously reported solidphase method [13] with 40% yield. The formation of the conjugate as the main reaction product was confirmed by MALDI-TOF. The mass spectrum contained the peak with m/z = 3839.52 corresponding to the mass of the molecular ion [M + H] + (the calculated molecular mass of the conjugate is 3838.62 g/mol).

Stopped-flow experiments
Stopped-flow measurements with UV absorbance detection were carried out using a model SX.18MV stopped-flow spectrometer (Applied Photophysics, UK) fitted with a 150 W Xe arc lamp and a 1 cm path length cell. The optical density of the solution was recorded at 255 nm. Solution of the target oligonucleotide P in one syringe was rapidly mixed with a solution of the conjugate X or nonmodified oligonucleotide N in another syringe. The concentration of P, X, and N were varied between 1.0 and 7.5 μM. Concentrations of reactants reported are those in the reaction chamber after mixing. Typically, each trace shown is the average of four or more individual experiments. The dead time of the instrument was 1.4 ms.

UV melting experiments
Absorbance versus temperature profiles were recorded at 260, 270, 280, and 300 nm using the optical detector of a Milikhrom chromatograph (Russia) connected to a PC. Melting profiles were obtained by heating at 0.5-0.9 • C/min. The concentration of each strand was 5.0 × 10 −6 M and the cell volume was 2 μl. The data were analyzed taking into account the thermal expansion coefficient of water. The differential curves were obtained from the integral ones by calculating the increment of the optical density per 1 • C of temperature growth. Thermodynamic parameters (ΔH 0 , ΔS 0 ) were calculated according to [14].

Degradation of the phthalocyanine group attached to oligonucleotide
The change of the optical density of the solution at 682 nm where only the phthalocyanine group of the conjugate absorbs was followed using a Shimadzu UV2100 spectrophotometer. The concentrations of the conjugate and hydrogen peroxide were changed in the ranges 8.

Modification of the target oligonucleotide
Modification of the [ 32 P]-labeled P was carried out in the presence of hydrogen peroxide. The concentration of P in the reaction mixture was 1.0 × 10 −8 M, concentrations of the conjugate and hydrogen peroxide were changed in the range 0.4×10 −6 −1.0×10 −5 M and 1.0×10 −3 −1.0×10 −1 M, respectively. The reaction was initiated by adding H 2 O 2 . Aliquots were taken from the reaction mixture at different times and were immediately transferred into polypropylene tubes containing 200 μl of 2% LiClO 4 in acetone. The precipitate was pelleted by centrifugation, washed twice with 80% ethanol and once with acetone, and dried in vacuum. The samples were then treated with piperidine or Fpg. The products of the modification were separated by 20% PAGE in the presence of 7 M urea. After electrophoresis, the gel was exposed to CP-BU X-ray film (Agfa-Gevaert, Belgium) for 10-20 h at −10 • C. The autoradiograms were scanned and quantified using Gel-Pro Analyzer v4.0 software (Media Cybernetics, MD). The extent of modification was calculated as the ratio of the integral intensity of the spot corresponding to the modification product to the sum of the intensities of the spots corresponding to the targets and the products.

Piperidine treatment
The precipitates were dissolved in 100 μl of 1 M piperidine (pH 12) and incubated for 30 min at 95 • C [16]. After that the reaction mixtures were precipitated with 10 volumes of 2% LiClO 4 in acetone, washed with 80% ethanol, then with acetone, dried under vacuum, and dissolved in 2-4 μl of the gel-loading dye containing 0.1% bromophenol blue, 0.1% xylene cyanol FF, and 7 M urea.

Stopped-flow kinetics
The influence of PtcCo(II) residue on the duplex formation between target P and oligonucleotide part of the conjugate was first investigated using the conjugate X and a nonmodified 10 nt oligonucleotide N. Formation of PX and PN was studied by stopped-flow kinetics with optical density detection at 255 nm.
As can be seen from Figures 2(a) and 2(b), the kinetic curves for the interaction of P with X and N were different. In the case of PN formation, a sharp decrease in the absorption was observed between 0 and 1 s, followed by a plateau phase after 1 s. When PX was formed, a slow decrease in the optical density after 1 s was detected. It should be noted that   The kinetic curves of PN formation were fitted to Scheme 2(a) using DynaFit software. The theoretical curves correlated well with the experimental data. The calculated rate constants are presented in Table 1.
In the case of PX, Scheme 2(a) did not describe the slight decrease of the optical density after 1 s. These data were treated using Scheme 2(b) containing two equilibria. The calculated constants are presented in Table 1. Accordingly, the process of the complex formation between the conjugate and target oligonucleotide can be described by a twostage mechanism. The overall association constant for the formation of PX was in a close agreement with the overall constant for the PN formation, indicating that the phthalocyanine residue did not influence the complex stability. It was shown earlier [18] that the association constant for Scheme 2: The formation of the complexes between the target P and the oligonucleotide N (a) or the conjugate X (b).
the complex formed by a conjugate of a 8 nt oligonucleotide with Co(II)-phthalocyanine and a 12 nt oligonucleotide target is 15-fold higher than that for the complex formed with the nonmodified 8 nt oligonucleotide and the 12 nt target. However, the stabilization/destabilization effects may depend on the lengths and sequences of the oligonucleotides used. Complex In our case when a 20 nt oligonucleotide was used as the target, no stabilization of its duplex with the conjugate was observed.
The process of DNA duplex formation includes two steps: nucleation and zipping of the duplex. The first stage is ratelimiting and proceeds with the rate constant about 1 × 10 6 − 1 × 10 7 (M × s) −1 [19]. In the case of the conjugate containing two binding parts, their influence on the complex formation could be rather complicated. Free porphyrins and their analogs are known to interact with DNA through either intercalation or outside binding [20,21]. The binding mode depends both on the nucleotide sequence and substitutions in the porphyrin moiety. It was shown [18] that Co(II)Ptc interacts with DNA chain. Thus, both fragments of the oligonucleotide-phthalocyanine conjugate can interact with the target oligonucleotide. The following mechanism of PX formation is possible. As can be seen from Figure 2, the optical density decreased quickly until 1 s when both X and N were binding to P. This decrease is due to the hypochromic effect arising from the formation of Watson-Crick base pairs between the oligonucleotides. Thus, the slow second stage present in the case of PX is likely caused by the phthalocyanine residue. Table 1 shows that rate constants k 1 for PN and PX * are similar but the values of k −1 differ twofold, suggesting that PX * is less stable than PN. In this case the stability of the complex depends on the number of the Watson-Crick base pairs formed at the first moment. The smaller change in the optical density during PX formation corroborates this conclusion because the change in the absorbance is proportional to the number of Watson-Crick base pairs. One can suggest that at the first stage only a few base pairs are formed between the conjugate and the oligonucleotide target and the phthalocyanine moiety of the conjugate sterically hinders base pairing. This effect may be due to the interaction of the phthalocyanine residue with 3-4 heterocyclic bases in the conjugate itself, or with bases in the target. In any case, incompletely "zipped" duplex PX * is likely to be formed at the first step. The slow phthalocyanine displacement at the second step leads to a formation of the fully complementary PX complex.

Melting curves
To determine the thermodynamic parameters of duplex stability, the UV melting curves were examined. Single-stranded oligonucleotide P yielded no visible transition in the differential melting curves recorded at 26 nm, indicating that P has no stable secondary structure. The same result was obtained when the conjugate was heated.
The differential melting curves of the complexes PN and PX had a single well-defined transition with T m = 32.6 and 32.9 • C, respectively (Figure 2(c)). Registration of melting curves at 300 nm permitted us to detect the interaction of phthalocyanine moiety with the target. The shape of the differential melting curve at 300 nm and the location of its maximum were identical to the shapes of the curves recorded at 260, 270, and 280 nm. Consequently, the melting curves at 260-280 nm and 300 nm described the same process, showing that the "melting" of the phthalocyanine residue occurred simultaneously with the melting of the oligonucleotide duplex.
The thermodynamic parameters of dissociation of PN and PX (ΔS 0 , ΔH 0 , ΔG 0 298 ) were determined using Simplex software. The obtained data are listed in Table 2. The association constants were close to those obtained from the stoppedflow experiments. Taken together, the results obtained by thermal denaturation method were in agreement with the data obtained by the fast kinetic method: the phthalocyanine residue does not influence the stability of the complex between the conjugate and the target.

Degradation of the phthalocyanine residue in the presence of hydrogen peroxide
Treatment of the conjugate with hydrogen peroxide resulted in a destruction of the phthalocyanine residue. This process is accompanied by a decrease in the optical density at 650-750 nm and at 320 nm, corresponding to the Q-bands and the Soret band, respectively. We have studied the kinetics of the phthalocyanine destruction by recording the absorbance at 682 nm.
The kinetic curves corresponding to different concentrations of H 2 O 2 are presented on Figure 3. Total degradation of the PtcCo(II) group was observed for high concentration of hydrogen peroxide (1.0 × 10 −2 and 5.0 × 10 −3 M). The reaction did not reach its plateau by 24 h at low concentrations of oxidizer (1.0×10 −3 and 5.0×10 −4 M). The repeated addition of H 2 O 2 to the reaction mixture after 24 h led to a further decrease in the optical density.
The degradation of PtcCo(II) residue by H 2 O 2 was accompanied with catalytic decomposition of hydrogen peroxide. The experimental data were satisfactorily described assuming that the destruction of conjugate followed second-order kinetics (first-order with respect to both components). The catalytic decomposition of H 2 O 2 was described as the third-order reaction (first-order by the conjugate and second-order by hydrogen peroxide). The values of rate constants k d (conjugate destruction) and k h (catalytic  Kremer in his work has shown [22] that the catalytic decomposition of H 2 O 2 by hemin is also second-order with respect to hydrogen peroxide. The process included formation of a primary heme-H 2 O 2 complex following by coordination of the second H 2 O 2 molecule and the catalytic act of hydrogen peroxide decomposition. It is quite possible that the catalysis of H 2 O 2 decomposition by the phthalocyanine Co(II) proceeds similarly.
Comparing our results with those obtained for the modification of a target with a conjugate of a 8 nt oligonucleotide with a Fe(III)-protoporphyrin IX (hemin) group in the presence of H 2 O 2 [23], we conclude that the rate constant of the phthalocyanine residue destruction was ∼120-fold lower than that of the hemin group. The destruction of the porphyrin system was the major cause of the low extent of target modification, which did not exceed 33%.

Modification of the target oligonucleotide
Kinetics of DNA modification within PX was studied by following the time course of cleavage of the [ 32 P]-labeled target at different times. No direct cleavage of the target strand was observed. Alkali-labile modifications (abasic sites and oxidized deoxyribose) were revealed by piperidine treatment of the DNA target. To digest the alkali-resistant products oxidized at deoxyguanosine residues (eg, 8-oxoguanine), the samples were treated with Fpg protein. Typical autoradiograms are presented in Figures 4(a) and 4(b). The yields and positions of the modified bases in the target were determined (Figure 4(c)). The modification occurred preferentially at guanine residues in the region G 9 -G 13 , indicating that the guanine bases close to the source of • OH radicals are the most susceptible, and that the stretch of P forming the duplex with the oligonucleotide part of the conjugate is protected from • OH radicals by X. These observations suggest that the preferential modification of P at G 9 -G 13 is due to the attack by • OH radicals before their diffusion in solution. Similarly localized damage was observed for irradiated DNA-Cu 2+ molecules [24].
The total modification extent of 80% was achieved with piperidine treatment and 40% with Fpg treatment. Since some of the modification products revealed by Fpg could be also determined by piperidine, the total modification extent was at least 80% and likely between 80 and 100%. The time courses of modification are shown in Figure 5.
To describe the oxidative modification of the target by the conjugate in the presence of hydrogen peroxide, Scheme 3 was proposed. The experiments were carried out under the conditions where x 0 p 0 . The large excess of the conjugate over the target was used in order to obtain complete binding of P into PX and to achieve the maximal level of target modification. In addition, as PX X, the decomposition of H 2 O 2 by PX was disregarded.
The kinetic curves were satisfactorily described with Scheme 3. The values of rate constants k d = 2.2 × 10 −2 (M × s) −1 , k h = 2.5 × 10 3 (M 2 × s) −1 , and K = 3.0 × 10 −6 M −1 determined previously (see Tables 1 and 3) were taken for fitting procedure using Scientist software. The rate constants of the target modification (k Scheme 3: The oxidation modification of the target with conjugate in the presence of H 2 O 2 . (In this scheme, P is the target oligonucleotide, X is the conjugate, R is the oxidation product of the phthalocyanine moiety possessing the same affinity to the target as the conjugate, PZ is the modification product, PX, PR, and PZX are respective complexes, and K, k d , k 0 , and k h are respective association and rate constants.) on the type of product analysis (Table 3). This means that differentproducts of guanine oxidation identified by piperidine and Fpg treatments accumulated with different rates. The piperidine treatment gave a higher total modification extent and k 0 value than the Fpg treatment.
When the DNA target is oxidized in the presence of a catalytically active oligonucleotide derivative and H 2 O 2 , several competing processes take place. First of all, the target is modified within the duplex with conjugate. Second, the catalytic group is damaged in the side reaction with hydrogen peroxide. Third, H 2 O 2 is catalytically decomposed into O 2 and H 2 O. The simulation according to Scheme 3 has shown that the dependence of the product modification on the initial concentrations of H 2 O 2 should have a maximum (Figure 6(a)), and its existence was experimentally confirmed (Figure 6(b)). The simulation of the dependence of the modification depth on the initial concentration of the conjugate also resulted in the curve with a maximum at about 30-100 μM (Figure 6(c)). The decrease in the modification extent with the increase in the conjugate concentration was possibly connected to the catalytic H 2 O 2 decomposition being the predominant process at these concentrations of the catalyst. Experimentally we could not achieve such high concentrations of conjugate; when they were varied from 0 to 50 μM; a predicted hyperbolic curve was observed (Figure 6(d)).
Our results show that the PtcCo(II) group in the oligonucleotide conjugate is able to modify target DNA in the presence of H 2 O 2 as an oxidant. Since this group acts as a catalyst of oxidation, such conjugates may be considered as artificial enzymes, the synthetic analogs of peroxidases. Hydrogen peroxide can be formed inside the cell in some endogenous processes (respiratory burst in mitochondria, oxidative stress, inflammation, etc, [25,26]) and can be involved in oxidation of target nucleic acids.
Oxidative modification of DNA with the PtcCo(II)-oligonucleotide conjugate is accompanied by destruction of the phthalocyanine macrocycle by the oxidant and catalytic decomposition of the oxidant. The bell-shaped dependence of the modification efficiency on the H 2 O 2 concentration suggests that these three processes are in competition. In comparison with the previously studied Fe(II)-porphyrinoligonucleotide conjugate [23], the phthalocyanine group is more stable to degradation. This property is useful for the design of oligonucleotides-based drugs, which can be promising candidates for cancer therapy.

ACKNOWLEDGMENTS
The authors are grateful to Dr Dmitry Pyshnyi for his help in the synthesis of the oligonucleotides and conducting UV melting experiments and Dr Dmitry Zharkov for critical reading of the manuscript. This research was made possible in part by Grants from RFBR (05-04-48447) and Grants from SD RAS (51, 60, 78) and CRDF (Y1-B-08-16). Alexander A. Chernonosov is supported by SD RAS (114).