The reduction of ruthenium(III) complexes with triazolopyrimidine ligands by ascorbic acid and mechanistic insight into their action in anticancer therapy

Inorganica Chimica Acta 484 (2019) 305–310

Contents lists available at ScienceDirect

Inorganica Chimica Acta journal homepage: www.elsevier.com/locate/ica

Research paper

The reduction of ruthenium(III) complexes with triazolopyrimidine ligands by ascorbic acid and mechanistic insight into their action in anticancer therapy Joanna Wiśniewska , Marzena Fandzloch, Iwona Łakomska ⁎

T

⁎

Faculty of Chemistry, Nicolaus Copernicus University in Toruń, Gagarina 7, 87-100 Toruń, Poland

ABSTRACT

Kinetic studies of the reduction of two ruthenium(III) complexes, mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) and mer,cis-[RuCl3(dbtp)2(dmso)] (2) (where dmso – dimethylsulfoxide, tmtp – 5,6,7-trimethyl-1,2,4-triazolo[1,5-a] pyrimidine and dbtp – 5,7-ditertbutyl-1,2,4-triazolo[1,5-a]pyrimidine), by ascorbic acid were performed as a function of antioxidant concentration in acetate buffer within the pH range of 2.9–5. The rapid reduction of the ruthenium(III) complexes (1) and (2) resulted in the formation of the mer-[RuCl3(dmso)(H2O)(tmtp)]− and mer,cis-[RuCl3(dbtp)2(dmso)]− ions and was followed by successive dissociation of the chloride ligands. The second-order rate constant (k1) for the reduction of the mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O complex and the first-order rate constant for the hydrolysis of its reduced form were found to be 134 ± 2 M−1 s−1 and (3.8 ± 0.9) × 10−2 s−1 at 25 °C and pH = 2.9, respectively. Similarly, the fast process assigned to the reduction of the mer,cis-[RuCl3(dbtp)2(dmso)] complex and the subsequent, slower process attributed to the hydrolysis of the cis-[RuCl3(dbtp)2(dmso)]− ion were characterized by rate constants of 145.5 ± 0.8 M−1 s−1 and (9 ± 2) × 10−3 s−1 at 25 °C and pH = 2.9, respectively. Obtained data indicated that the reduction of the ruthenium(III) complexes strongly depends on pH and accelerates with increasing pH. The kinetic data indicates that the redox process followed an inner-sphere electron-transfer mechanism at pH higher than 3.

1. Introduction Compounds of almost all metals of the periodic table have been investigated for in vitro anticancer activity against cancer cell lines [1–11]. Among them, the ruthenium compounds are the potent chemotherapeutics that are the most promising to be competitive against platinum-based inorganic drugs [12–22]. A possible explanation for these effects of ruthenium compounds may be derived from their kinetic properties. Similar to platinum, ruthenium is a relatively inert metal, and its ligand exchange kinetics are typically within the same timescale as cellular division processes [20,23]. Ruthenium compounds exhibit a wide spectrum of redox properties and can exist at various oxidation states under biologically relevant conditions. The crucial points are to check whether ruthenium is transformed reductively and to discover the possible pathways of these transformations in vivo.

Moreover, ruthenium complexes have shown significant ability to bind many biological molecules, including serum proteins (e.g., transferrin and albumin). In the human body, ruthenium compounds circulate in the bloodstream and are then transported to the cell through the cellular membrane, mostly by transferrin-mediated uptake. It is important to determine whether ruthenium(III) is reduced to ruthenium(II) in the cytoplasm or extracellular fluids by endogenic reducing agents, which essentially modifies its reactivity with biomolecules or cell components and modulates its transformation into the active form of drugs. It is necessary to check whether the levels of antioxidant concentrations (vitamin C and glutathione) and the rate of redox processes are relevant and effective for such a process to actually occur in the cell environment. The answer to this question and those above will allow us to better understand the mechanism of action of potential ruthenium pharmaceuticals and to better plan the synthesis of new compounds

Abbreviations: ks, rate constant for slow process; kf, rate constant for fast process; k1obs, observed first-order rate constant for process 1 derived from two-exponential function; k2obs, observed first-order rate constant for process 2 derived from two-exponential function; k1, second-order rate constant for reduction by H2Asc; k-1, second-order rate constant for reverse reaction; k2, second-order rate constant for fast reduction by ascorbyl radical (not directly observed); k1aq, first-order rate constant for aquation of ruthenium(III); k2aq, first-order rate constant for aquation of ruthenium(II) in the second step of reaction; k1b, second-order rate constant for reduction by HAsc− ⁎ Corresponding author. E-mail addresses: [email protected] (J. Wiśniewska), [email protected] (I. Łakomska). https://doi.org/10.1016/j.ica.2018.09.051 Received 19 June 2018; Received in revised form 27 August 2018; Accepted 20 September 2018 Available online 22 September 2018 0020-1693/ © 2018 Elsevier B.V. All rights reserved.

Inorganica Chimica Acta 484 (2019) 305–310

J. Wiśniewska et al.

was prepared in water due to its instability at high pH [32] and stored for no longer than one day. Ascorbic acid solutions of desired concentrations (0.002, 0.0025, 0.005, 0.0075, 0.01, 0.0125 and 0.02 M) were prepared freshly by dilution of 1, 1.25, 2.5, 3.75, 5, 6.25 and 10 mL of ascorbic acid stock solution and 5 mL of 0.5 M acetate buffer in 25 mL volumetric flasks. Ultrapure water was obtained from a Milli-Q system (Millipore/Waters, Milford, MA, USA) and used to prepare all the solutions. 2.2. Instrumentation Cyclic voltammetry measurements were performed using an Eco Chemie Autolab PGSTAT30 Potentiostat under an inert atmosphere (Ar) in a three-electrode cell at 25 °C. The working electrode was a Pt-wire, the auxiliary electrode was a platinum–carbon electrode, and the reference electrode was the saturated Ag/AgCl electrode. The potentials were measured in a 0.1 M [nBu4N][PF6]/acetonitrile (ACN) versus a Ag/AgCl electrode. All measurements were performed at laboratory temperature and at 100 V s−1 scan rate.

Scheme 1. Structural formulas of the mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) and mer,cis-[RuCl3(dbtp)2(dmso)] (2) complexes.

with potential anticancer activity in future. Recently, the hypothesis that the antitumor activity of ruthenium (III) complexes involves activation by reduction in vivo prior to metal coordination to nucleic acids or other biological molecules has been proposed [19,24–26]. According to this hypothesis, ruthenium complexes should possess biologically accessible reduction potential (in the range from −0.4 V to +0.8 V vs NHE) in their in vivo activity to act as a prodrugs [27,28]. Nevertheless, the biological target and mechanism of action of ruthenium compounds are largely unknown. Particularly, it is still unknown whether ruthenium(II) binds to nucleic acid or other biological molecules in a cancer cell more effectively than ruthenium (III) as a result of much higher lability of ruthenium(II) in relevance to ruthenium(III). Recently, we have reported the synthesis, structures and spectroscopic properties of new ruthenium(III) complexes with triazolopyrimidine ligands, mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) [29], mer,cis[RuCl3(dbtp)2(dmso)] (2) [30] (Scheme 1), trans-[RuCl3(dbtp)2(H2O)] and mer-[RuCl3(dbtp)3] [31]. The last two compounds were determined to be significant relative to cisplatin, as they are inhibitors of tumour cell proliferation with IC50 values at micromolar concentrations against 4 cancer cell lines: a human lung adenocarcinoma epithelial cell line (A549), a cisplatin-resistant human ductal breast epithelial tumour cell line (T47D), a human breast adenocarcinoma and grade III-causing bone metastases line (MDA-MB-231), and a human breast adenocarcinoma line (MCF-7) [31]. Although complexes (1) and (2) were less toxic against the cancer cell lines mentioned above, they were also determined to be less toxic against normal murine embryonic fibroblast cells (BALB/3T3) and a non-tumourigenic human epithelial cell line (MCF-10A) than cisplatin [29,30]. More attention should be given to kinetic and mechanistic studies of any transformations of these potential ruthenium(III) pharmaceutics in aqueous solution, which functions as an adequate model of the cell environment, with a special emphasis on reduction processes. This aspect has now been studied in additional detail for these newly synthesized complexes. This report is the first example of mechanistic insight into the reduction processes of these ruthenium(III) complexes with triazolopyrimidine derivatives. Introduction of spectator chelates, i.e., innocent, non-innocent and labilizing ligands, may have a drastic effect on the reaction rate and the underlying reaction mechanism.

2.3. Kinetic measurements Time-resolved spectra and kinetic measurements were recorded on an SX 18 MV Applied Photophysics apparatus equipped with a J&M TIDAS diode-array detector. The data were acquired and analysed with Applied Photophysics software. In the experiments, the concentration of ruthenium(III) was fixed at (1–2) × 10−4 M. The concentration of ascorbic acid used was in excess and varied over the range of (1–10) × 10−3 M for complexes (1) and (2). The other experimental conditions were as follows: pH = 4.5, 0.1 M acetate buffer, I ≠ constant (H+, Na+, CH3COO−), T = 298 K, l = 1 cm. In another series of experiments, the rate was measured at a lower pH value, where ascorbic acid (H2A) is present in its fully protonated form. The other experimental conditions were as follows: pH = 2.9, 0.1 M acetic acid, I = constant (H+, CH3COO−), T = 298 K, l = 1 cm. The rate was also analysed at different H+ concentrations: pH = 2.9–5.0, 0.1 M acetate buffer, I ≠ constant (H+, Na+, CH3COO−), T = 298 K, l = 1 cm. The progress of the reaction was followed by monitoring the absorbance at 370 nm (electronic transitions in the ruthenium(III) complex region). Reactions were studied under pseudo-first-order conditions, i.e., ascorbic acid was in excess over the ruthenium(III) complex. In the reactions of (1) and (2), two steps were observed that could not be separated kinetically. Therefore, the data were analysed by a GaussNewton nonlinear least-squares fit of a two-exponential function of absorbance vs time. The reported rate constants are the mean values of at least four determinations. In the reaction of (2), due to some fluctuations in absorbance after mixing the reactant solutions, the analysis was performed in an adequate time scale, with the first 0.1 s of the overall process omitted. The relative standard errors of the pseudo-firstorder rate constants for a single kinetic trace were ca. 0.1–0.2%, and relative standard errors of the mean value were usually ca. 0.1–0.5%. 3. Results and discussion The stopped-flow rapid-scan spectra revealed characteristic spectral changes in the visible range attributed to rapid reduction of the ruthenium(III) complexes to ruthenium(II) complexes and the subsequent solvolysis. The former process is accompanied by a subsequent decrease in absorbance from the electron-transition band at 366 nm assigned to complex (1) ([RuIII] = 2 × 10−4 M, [H2Asc] = 1 × 10−3 M, pH = 2.9, 0.1 M acetic acid, T = 298 K), which appears simultaneously with a decrease in the absorbance of its low-energy transition band at 430 nm. The spectral changes characteristic for the degradation of the mer[RuCl3(dmso)(H2O)(tmtp)]− ion are consistent with a further decrease in the intensity of the electron-transition bands at 366 and 430 nm and with a blueshift of the high-energy transition band from 366 nm to

2. Experimental 2.1. Materials Ascorbic acid (Sigma-Aldrich), acetic acid (70%, Sigma-Aldrich), sodium acetate (Sigma-Aldrich) and all other chemicals were analytical grade reagents. The mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1), mer,cis[RuCl3(dbtp)2(dmso)] (2) complexes were prepared as described elsewhere [29,30]. Solutions of (1) and (2) were prepared freshly in methanol just before being mixed with aqueous acetate buffer (1:9) and used immediately in measurements. The 0.05 M ascorbic acid solution 306

Inorganica Chimica Acta 484 (2019) 305–310

J. Wiśniewska et al.

Table 1 Pseudo-first-order rate constants for the reduction of the mer-[RuCl3(dmso) (H2O)(tmtp)]·2H2O (1) complex (k1obs) and for the aquation of the reduced form of the ruthenium(III) complex (k2obs). Experimental conditions: [RuIII] = 1 × 10−4 M, T = 298 K, λ = 370 nm, l = 1 cm.

Fig. 1. Spectral changes observed during the reduction of the mer-[RuCl3(dmso) (H2O)(tmtp)]·2H2O (1) complex by ascorbic acid. Experimental conditions: [RuIII] = 2 × 10−4 M, [H2A] = 0.001 M, pH = 2.9, 0.1 M acetic acid, T = 298 K, l = 1 cm, t = 100 s, Δt = 0.5 s.

330 nm (Fig. 1). Finally, hydrolysis leads to gradual changes in both maxima, and after 100 s, only the broad band at 330 nm is observed. Reduction of complex (2) is consistent with the decreases in absorbance at 366 and 423 nm, and the solvolysis is accompanied by a further decrease in absorbance at 366, a significant blueshift of the high-energy transition band from 366 to 320 nm, and a slow decrease in the low-energy electron-transition band at 423 nm. The spectral changes occurring during the reductive degradation of the mer,cis-[RuCl3(dbtp)2(dmso)] complex are similar to those of the mer-[RuCl3(dmso)(H2O)(tmtp)] complex. These observations and conclusions are supported by values of εB for an intermediate, derived from the following equation [33]: Case 1 B

=

B

= (

A

af /[Ru]T )(1

ks / k f ) +

C ks /

kf

=

B

=(

C A

+ af /[Ru]T (1

k f / ks )

as /[Ru]T )(1

k f / ks ) +

(2)

ks

0.001 0.00125 0.0025 0.00375 0.005 0.00625 0.01

2.9

0.1715 0.1987 0.3798 0.5638 0.6911 0.8759 1.3848

0.001 0.00125 0.0025 0.00375 0.005 0.00625 0.01

4.5

9.16 ± 0.05 10.32 ± 0.04 17.92 ± 0.08 25.0 ± 0.1 32.4 ± 0.1 38.8 ± 0.1 59.4 ± 0.5

1.657 1.711 2.459 3.049 3.521 4.006 5.599

0.001

5.0 4.5 4.0 3.5 3.1 2.9

10.91 ± 0.05 9.16 ± 0.05 3.85 ± 0.07 0.5944 ± 0.0005 0.2748 ± 0.0004 0.1715 ± 0.0003

0.697 ± 0.005 1.657 ± 0.003 0.983 ± 0.006 0.0094 ± 0.0001 0.0105 ± 0.0001 0.0261 ± 0.0002

± ± ± ± ± ± ±

k2obs [s−1]a 0.0003 0.0003 0.0004 0.0006 0.0005 0.0006 0.0010

0.0261 0.0294 0.0236 0.0264 0.0258 0.0299 0.0295

± ± ± ± ± ± ±

± ± ± ± ± ± ±

0.0002 0.0003 0.0003 0.0004 0.0006 0.0007 0.0004

0.003 0.002 0.005 0.002 0.005 0.005 0.008

Refer to k2aq in Eq. (7).

Table 2 Pseudo-first-order rate constants for the reduction of the mer,cis[RuCl3(dbtp)2(dmso)] (2) complex (k1obs) and for the aquation of the reduced form of the ruthenium(III) complex (k2obs). Experimental conditions: [RuIII] = 1 × 10−4 M, T = 298 K, λ = 370 nm, l = 1 cm.

(3) C kf /

k1obs [s−1]

The kinetics of the overall process were studied in 0.1 M CH3COOH (pH = 2.9). The kinetic traces were collected at 370 nm and satisfactorily obey a two-exponential dependence of absorbance on time [33]. Pseudo-first-order rate constants (k1obs and k2obs) were analysed by digital simulations and are collected in Tables 1 and 2. Plots of the observed pseudo-first-order rate constant k1obs versus the concentration of H2A were found to be linear with meaningful intercepts (Fig. 2A and 2B). In contrast, the pseudo-first-order rate constant k2obs only slightly depended on the total H2A concentration at pH = 2.9. The rate constant k1obs was attributed to the reduction process, and k2obs was attributed to the subsequent aquation of the reduced form of the ruthenium(III) complex. Considering all these data, the ruthenium(III) complexes undergo the reaction with ascorbic acid according to the following reaction scheme:

where εA, εB and εC are the molar absorptivities of the reactant, an intermediate and the final product (εA = 4440 M−1 cm−1 for mer,cis[RuCl3(dbtp)2(dmso)], ε C = 100 M−1 cm−1 at 370 nm), respectively, kf is a rate constant assigned to the first, fast process and ks is a rate constant assigned to the second, slow process. Case 2 B

pH

a

(1)

+ as /[Ru]T (1 ks/ kf )

C

[H2A] [M]

(4)

where ks is a rate constant assigned to the first, slow process and kf is a rate constant assigned to the second, fast process. The results for case 1 are preferred for the following reasons. The value of εB was constant for kinetic traces of different experiments, and the molar coefficient of the intermediate lay between the values of εA and εC, in accordance with spectral changes observed during the reaction course. εB exhibited significantly larger discrepancies and higher values than εA in case 2. Preliminary studies revealed that (2) is less soluble in water in the presence of sodium chloride than in water without sodium chloride and slowly precipitated in the former medium. The same behaviour was observed in the presence of sodium chloride in methanol:water solution (1:9). Therefore, the kinetic measurements were performed in the absence of added salts apart from sodium acetate, which was used when the reactions were studied in acetate buffer solutions. The ruthenium (II) complexes are quite stable in aqueous solutions and are not reoxidized by atmospheric oxygen, therefore the kinetic measurements were performed under air atmosphere.

a

[H2A] [M]

pH

k1obs [s−1]

0.001 0.00125 0.0025 0.00375 0.005

2.9

0.1552 0.1931 0.3731 0.5528 0.7392

0.001 0.00125 0.0025 0.00375 0.005 0.00625 0.01

4.5

4.410 ± 0.008 5.28 ± 0.01 9.77 ± 0.02 12.79 ± 0.03 18.17 ± 0.05 21.58 ± 0.07 34.06 ± 0.08

0.501 ± 0.007 0.667 ± 0.009 0.693 ± 0.007 1.00 ± 0.01 1.21 ± 0.01 1.91 ± 0.02 2.56 ± 0.02

0.001

5.0 4.5 4.0 3.5 3.1 2.9

16.62 ± 0.05 5.442 ± 0.008 2.181 ± 0.002 0.2138 ± 0.0002 0.1862 ± 0.0005 0.1354 ± 0.0004

0.418 ± 0.002 0.0727 ± 0.0008 0.0199 ± 0.0004 0.0080 ± 0.0001 0.0081 ± 0.0001 0.0079 ± 0.0004

Refer to k2aq in Eq. (7).

307

± ± ± ± ±

k2obs [s−1]a 0.0002 0.0002 0.0004 0.0005 0.0007

0.0078 0.0075 0.0068 0.0061 0.0057

± ± ± ± ±

0.0006 0.0004 0.0003 0.0006 0.0004

Inorganica Chimica Acta 484 (2019) 305–310

J. Wiśniewska et al.

Fig. 2. Plots of k1obs versus [H2A] for the electron-transfer reaction between the mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) (A) and mer,cis-[RuCl3(dbtp)2(dmso)] (2) (B) complexes and ascorbic acid. Experimental conditions: [RuIII] = 1 × 10−4 M, [H2A] = 0.001–0.01 M, pH = 2.9, 0.1 M acetic acid, T = 298 K, l = 1 cm. Plots of k1obs (C) and k2obs (D) versus [H2A] for the electron-transfer reaction between the mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) complex and ascorbic acid. Experimental conditions: [RuIII] = 1 × 10−4 M, [H2A] = 0.001–0.01 M, pH = 4.5, 0.1 M acetate buffer, T = 298 K, l = 1 cm.

[RuIIICl3L 3] + H2 A [RuIICl3L3] + A ·

k2

k1 k 1

[RuIICl3L3] + A ·+ 2H+

[RuIICl3L 3] +A

[RuIICl3L3] + H2 O

k2aq

[RuIICl2 (H2 O)L 3] + Cl

Table 3 Linear regression data for the plots of kobs1 versus [H2A] for the reaction of the mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) and mer,cis-[RuCl3(dbtp)2(dmso)] (2) complexes with ascorbic acid. Experimental conditions: [RuIII] = 1 × 10−4 M, T = 298 K, λ = 370 nm, l = 1 cm.

(5) (6) (7)

The reduction of (1) and (2) by H2A affects the formation of the ruthenium(II) complexes and the ascorbate radical ion (A−%), which reduces the second molecule of the complex to form dehydroascorbate (A) as the final oxidation product of ascorbic acid. Thus, the overall rate law can be expressed by Eq. (8):

d[Ru(III)] = (k1aq + 2k1 [H2 A]T )[Ru(III)] dt

(8)

Complex

pH

2 k1 (M−1 s−1) Slope

k1aq (s−1) Intercept

mer-[RuCl3(dmso)(H2O) (tmtp)]·2H2O (1)

2.9

134 ± 2

(3.8 ± 0.9) × 10−2

4.5

5604 ± 53

3.8 ± 0.3

mer,cis-[RuCl3(dbtp)2(dmso)] (2)

2.9 4.5

145.5 ± 0.8 3289 ± 106

(9 ± 2) × 10−3 1.17 ± 0.4

potentials, 0.392 V and 0.370 V [30] (versus NHE and measured in 0.1 M [nBu4N][PF6]/ACN using CV method). The first-order rate constant k1aq refers to the intercepts in the plots (Fig. 2A and 2B) and was found to be (3.8 ± 0.9) × 10−2 s−1 and (9 ± 2) × 10−3 s−1, for (1) and (2), respectively. The values of k1aq are of the same order of magnitude as the values of the observed rate constants (k2obs, see Tables 1 and 2) for the two ruthenium(II) complexes. Moreover, the rate constants (k2obs) for mer-[RuCl3(dmso)(H2O)(tmtp)]– are also one order of magnitude higher than those of the mer,cis-[RuCl3(dbtp)2(dmso)]– complex. On account of greater electron donor properties of the dbtp derivative and higher stabilization energy of the t2g orbital, the dbtp ruthenium(II) complex is expected to be thermodynamically more

The reduction is followed by solvolytic processes, and the final products are generated as aqua derivatives of mer-[RuCl3(dmso)(H2O) (tmtp)]− and mer,cis-[RuCl3(dbtp)2(dmso)]−. In the cases of the mer[RuCl3(H2O)(dmso)(tmtp)]− and mer,cis-[RuCl3(dbtp)2(dmso)]− ions, hydrolysis of only one chloride is observed (Eq. (7)), and the solvolysis of ruthenium(II) is significantly slower than the reduction process (Eqs. (5) and (6)). The second-order rate constant (2 k1) for the reduction at 25 °C was 134 ± 2 M−1 s−1 for the mer-[RuCl3(dmso)(H2O)(tmtp)] complex, which is very close to the value of 2 k1 for the mer,cis[RuCl3(dbtp)2(dmso)] complex, viz. 145.5 ± 0.8 M−1 s−1 (Table 3). The similar reduction efficiency for the two structurally different complexes (1) and (2) is in accordance with their similar reduction 308

Inorganica Chimica Acta 484 (2019) 305–310

J. Wiśniewska et al.

stable and kinetically less labile than the tmtp ruthenium(II) complex. In acetic acid solution at pH = 2.9, ascorbic acid exists mainly in its fully protonated form H2A. At higher pH values, two conjugated bases are formed in solution, viz. HA− (pKa1 = 4.1) and A2− (pKa2 = 11.3) [34–36]. Due to the stronger reducing properties of the dehydroascorbate anion (HA−), the rate of reduction of the Ru(III) complexes should depend on acidity over the pH range of 2.9–5. Therefore, the kinetics of the overall process were also studied in acetate buffer solutions at pH = 4.5, under which conditions the HA− ion, expected to be a stronger reducing agent than ascorbic acid (H2A), is the major reactive species. The plot of the pseudo-first-order rate constants (k1obs) versus the total concentration of H2A was characterized by the presence of a significantly larger intercept (Fig. 2C) at pH 4.5 than at pH = 2.9. The second-order rate constant (2 k1) for the reduction of the mer[RuCl3(dmso)(H2O)(tmtp)] complex at pH = 4.5 and T = 25 °C (Table 3), (5.60 ± 0.05) × 103 M−1 s−1, is much higher than the value of 24.1 ± 0.6 M−1 s−1 reported for NAMI-A ((ImH)[trans-RuCl4(dmso) (Im)], Im = imidazole) at pH = 5.0 [37]. As shown in Tables 1, 2 and Fig. 2D, the observed pseudo-first-order rate constant (k2obs) for the solvolytic process at pH = 4.5 slightly increased with increasing total concentration of ascorbic acid, illustrating an inverse trend relative to the data obtained at lower pH = 2.9. The first-order rate constant (k1aq), representing the rate of degradation of the ruthenium(III) complex, equals 3.7 ± 0.3 s−1. As mentioned above, this value is higher than the k1aq term of (3.8 ± 0.9) × 10−2 s−1 obtained for the hydrolysis of (1) at pH = 2.9. These results are affected by the competing reaction path related to the displacement of the labile water molecule in (1) by the dehydroascorbate (HA−) anion. Thus, the rate of the reaction should increase when the concentration of HA− increases from 0.0007 M ([H2A]T = 0.001 M) to 0.007 M ([H2A]T = 0.01 M) at pH = 4.5 because the observed rate constants should be related to the more reactive species, HA−, that is present in solution. The increase in the observed rate constants (k2obs) resulted from the increase in the concentration of the HA− at pH > 3. These observations can be supported by the fact that the coordinated water molecule is known to be more labile than coordinated chloride in the coordination sphere of ruthenium(III) complexes [38] and can be easily substituted by HA−, which enters the complex before reduction. This parallel reaction path attributed to the substitution of HA− into the ruthenium(III) complex should also be involved in the mechanism. The formation of the Ru (III)–ascorbate complex has been suggested for NAMI-A and other aqua ruthenium(III) complexes [37,39]. The hydrolysis of (1) and (2), which mostly affects the k1aq term (Eq. (8)) (Fig. 2A and 2B), should be considered in the mechanism at higher pH values. The hydrolysed complexes, having water molecules as labile sites, should reduce more quickly than other complexes, thus providing an additional parallel reaction path that proceeds according to an inner-sphere electrontransfer mechanism in the overall process. As has been shown earlier [29,30], (1) and (2) are quite stable at pH = 3, and the rate of hydrolysis increased with increasing pH. The hydrolysis of (1) and (2) is characterized by zero-order kinetics and zero-order rate constants of (1.519 ± 0.007) × 10−8 M s−1 [29] and (0.539 ± 0.007) × 10−8 M s−1 (0.1 M acetate buffer, pH = 5.0), respectively [30]. This reaction may provide a parallel reaction path and affect the k1aq term but only when at least one process from the multi-step mechanism of hydrolysis is fast. The influence of the H+ ion on the reduction of (1) results from the deprotonation of H2A and changes with the percentage of HA− ions in the total concentration of ascorbic acid. The observed pseudo-firstorder rate constant k1obs significantly increased with decreasing H+ concentration (Fig. 3). On the assumption that the HA− ion is the main reactive species and under the selected conditions (pH = 2.9–5), the rate law can be expressed by Eq. (9):

d [Ru(III)] 2k K = (k1aq + + 1b 1 [H2 A]T )[Ru(III)] dt [H ] + K1

Fig. 3. Plot of k1obs versus [H+] for the electron-transfer reaction between the mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) complex and ascorbic acid. Experimental conditions: [RuIII] = 1 × 10−4 M, [H2A] = 0.001 M, pH = 2.9–5, 0.1 M acetate buffer, T = 298 K, l = 1 cm.

where 2k1b and K1 refer to the second-order rate constant for the reduction of the ruthenium(III) complex by the conjugate base HA− and the first acid dissociation constant of H2A, respectively. The observed first-order rate constant kobs can be expressed as in Eq. (10):

kobs = k1aq +

2k1b K1 [H2 A]T [H+] + K1

(10)

The parameters obtained from Eq. (10) lead to the following values: k1aq + 2k1b[H2A]T = 9 ± 2 s−1, K1 = (7 ± 2) × 10−5 M and pK1 = 4.15. The first dissociation constant obtained from the kinetic data is very close to the literature value of pKa1 = 4.1 [34]. The newly synthesized ruthenium(III) complexes with triazolopyrimidine derivatives should exhibit stronger cytotoxic activity upon reduction by ascorbic acid. In accordance with the activation by reduction theory [18,19,24,40], reduced NAMI-A, obtained by reduction of NAMI-A with ascorbic acid prior to administration, was found to be even more efficient than unreduced NAMI-A against metastasis growth [17,41]. This prominent reductant is present in the reducing sites within the tumour cells and is postulated to transform ruthenium(III) into the active and more labile ruthenium(II) species, which binds to DNA in the tumour cell as a result of ruthenium(II) being much more labile than ruthenium(III) and is responsible for the anti-metastatic activity in antitumour therapy [41]. At present, the only spectroscopic insight into the degradation of the Hdmtp[trans-RuCl4(dmso)(dmtp)], Na[trans-RuCl4(dmso)(dmtp)] and mer-[RuCl3(H2O)(dmso)(dmtp)] complexes, where dmtp – 5,7-dimethyl-1,2,4-triazolo[1,5-a]pyrimidine, has been published by Velders et al. [42]. Our studies provide the first example of a taking a mechanistic approach to research into the chemical reactivity of this class of ruthenium(III) compounds, which may be reductively transformed into ruthenium(II) complexes. 4. Conclusions These ruthenium(III) complexes exhibit a strong ability to be reduced by L-ascorbic acid, which is a known biomolecule present in blood serum. Both the mer-[RuCl3(dmso)(H2O)(tmtp)]·2H2O (1) and mer,cis-[RuCl3(dbtp)2(dmso)] (2) complexes are reduced by L-ascorbic acid in milliseconds, indicating that the reactive form of the complexes in the presence of endogenous reducers is ruthenium(II) rather than ruthenium(III); the postulated mechanism for potential platinum(IV) pharmaceutics states that are reduced to platinum(II) and then bind DNA in tumour cells as a result of platinum(II) being much more labile than platinum(IV). The similar rates of reduction for structurally

(9) 309

Inorganica Chimica Acta 484 (2019) 305–310

J. Wiśniewska et al.

different complexes mostly results from their similar reduction potentials. Due to the π-acceptor properties of S-bonded dmso, the reduction potential values, Ered, of such complexes are shifted to higher values than those of complexes without the coordinated dmso molecule [43]. The obtained value of Ered equals 0.392 V for mer-[RuCl3(dmso)(H2O) (tmtp)] is only slightly higher than the Ered for mer,cis[RuCl3(dbtp)2(dmso)], which equals 0.370 V [30] versus NHE and shows how the nature of N-donor ligands in the coordination sphere of the central atom influences redox properties. According to the hypothesis “activation by reduction”, the choice of ligands with suitable electron donor properties such as triazolopyrimidine derivatives and presence of dmso can preferably tune the redox potential and enable tuning the redox properties to obtain complexes, which are redox active in the biological environment.

[14] W. Peti, T. Pieper, M. Sommer, B.K. Keppler, G. Giester, Eur. J. Inorg. Chem. (1999) 1551–1555. [15] C.G. Hartinger, S. Zorbas-Seifried, M.A. Jakupec, B. Kynast, H. Zorbas, B.K. Keppler, J. Inorg. Biochem. 100 (2006) 891–904. [16] M.R. Berger, F.T. Garzon, B.K. Keppler, D. Schmähl, Anticancer Res. 9 (1989) 761–765. [17] G. Süss-Fink, Dalton Trans. 39 (2010) 1673–1688. [18] M.J. Clarke, Coord. Chem. Rev. 236 (2003) 209–233. [19] M.J. Clarke, F. Zhu, D.R. Frasca, Chem. Rev. 99 (1999) 2511–2534. [20] I. Bratsos, T. Gianferrara, E. Alessio, C.G. Hartinger, M.A. Jakupec, Bioinorganic Medicinal Chemistry, Wiley-VCH, Weinheim, 2011, pp. 151–174. [21] M. Melchart, P.J. Sadler, G. Jaouen (Ed.), Bioorganometallics, Wiley-VCH, Weinheim, 2006, p. 39. [22] L. Ronconi, P.J. Sadler, Coord. Chem. Rev. 251 (2007) 1633–1648. [23] L. Helm, A.E. Merbach, Chem. Rev. 105 (2005) 1923–1960. [24] M.J. Clarke, Prog. Clin. Biochem. Med. 10 (1989) 25–39. [25] M. Bacac, A.C. Hotze, K. van der Schilden, J.G. Haasnoot, S. Pacor, E. Alessio, G. Sava, J. Reedijk, J. Inorg. Biochem. 98 (2004) 402–412. [26] M.A. Jakupec, V.B. Arion, S. Kapitza, E. Reisner, A. Eichinger, M. Pongratz, B. Marian, N. von Graf Keyserlingk, B.K. Keppler, J. Int, Clin. Pharmacol. Ther. 43 (2005) 595–596. [27] F.Q. Schafer, G.R. Buettner, Free Radic. Biol. Med. 30 (2001) 1191–1212. [28] C.S. Sevier, C.A. Kaiser, Nat. Rev. Mol. Cell Biol. 3 (2002) 836–847. [29] M. Fandzloch, J.M. Méndez Arriaga, M. Sánchez-Moreno, A. Wojtczak, J. Jezierska, J. Sitkowski, J. Wiśniewska, J.M. Salas, I. Łakomska, J. Inorg. Biochem. 176 (2017) 144–155. [30] M. Fandzloch, L. Dobrzańska, J. Jezierska, B. Filip-Psurska, J. Wiśniewska, J. Wietrzyk, J.M. Salas, I. Łakomska, Polyhedron 141 (2018) 239–246. [31] I. Łakomska, M. Fandzloch, T. Muzioł, T. Lis, J. Jezierska, Dalton Trans. 42 (2013) 6219–6226. [32] H. Borsook, H.W. Davenport, C.E.P. Jeffreys, R.C. Warner, J. Biol. Chem. 117 (1936) 237–279. [33] R.G. Wilkins, Kinetics and Mechanism of Reactions of Transition Metal Complexes, second ed., VCH, Weinheim, 1991, p. 20. [34] D.H. Macartney, N. Sutin, Inorg. Chim. Acta 74 (1983) 221–228. [35] J.J. Warren, T.A. Tronic, J.M. Mayer, Chem. Rev. 110 (2010) 6961–7001. [36] J.J. Warren, J.M. Mayer, J. Am. Chem. Soc. 132 (2010) 7784–7793. [37] M. Brindell, D. Piotrowska, A.A. Shoukry, G. Stochel, R. van Eldik, J. Biol. Inorg. Chem. 12 (2007) 809–818. [38] B. Cebrián-Losantos, E. Reisner, Ch.R. Kowol, A. Roller, S. Shova, V.B. Arion, B.K. Keppler, Inorg. Chem. 47 (2008) 6513–6523. [39] M. Brindell, I. Stawoska, J. Supel, A. Skoczowski, G. Stochel, R. van Eldik, J. Biol. Inorg. Chem. 13 (2008) 909–918. [40] E. Alessio, G. Balducci, A. Lutman, G. Mestroni, M. Calligaris, W.M. Attia, Inorg. Chim. Acta 203 (1993) 205–217. [41] G. Sava, A. Bergamo, S. Zorzet, B. Gava, C. Casrsa, M. Cocchietto, A. Furlani, V. Scarcia, B. Serli, E. Iengo, E. Alessio, G. Mestroni, Eur. J. Cancer 38 (2002) 427–435. [42] A.H. Velders, A. Bergamo, E. Alessio, E. Zangrando, J.G. Haasnoot, C. Casarsa, M. Cocchietto, S. Zorzet, G. Sava, J. Med. Chem. 47 (2004) 1110–1121. [43] E. Reisner, V.B. Arion, B.K. Keppler, A.J.L. Pombeiro, Inorg. Chim. Acta 361 (2008) 1569–1583.

Acknowledgements The authors thank Prof. G. Stochel research group for the opportunity to perform all stopped-flow experiments at Jagiellonian University in Kraków, Poland. References [1] C.J. Jones, J.R. Thornback, Medicinal Application of Coordination Chemistry, RSCPublishing, Cambridge, 2007, pp. 324–339. [2] B. Lippert Cisplatin, Chemistry and Biochemistry of a Leading Anticancer Drug, Wiley-VCH, Weinheim, 1999. [3] J. Reedijk, Platinum Met. Rev. 52 (2008) 2–11. [4] J. Reedijk, Eur. J. Inorg. Chem. 10 (2009) 1303–1312. [5] C.V. Christodoulou, A.G. Eliopoulos, L.S. Young, L. Hodgkins, D.R. Ferry, D.J. Kerr, Br. J. Cancer 77 (1998) 2088–2097. [6] M. Guo, P.J. Sadler, J. Chem. Soc., Dalton Trans. (2000) 7–9. [7] C.H.S. Ruxton, P.C. Calder, C.S. Reed, M.J.A. Simpson, Nutr. Res. Rev. 18 (2005) 113–129. [8] S.H. van Rijt, A.F.A. Peacock, R.D. Johnstone, S. Parsons, P.J. Sadler, Inorg. Chem. 48 (2009) 1753–1762. [9] S.H. van Rijt, A. Mukherjee, A.M. Pizarro, P.J. Sadler, J. Med. Chem. 53 (2010) 840–849. [10] H. Faneca, V.A. Figueiredo, I. Tomaz, G. Gonçalves, F. Avecilla, M.C. Pedroso de Lima, C.F.G.C. Geraldes, J.C. Pessoa, M.M.C.A. Castro, J. Inorg. Biochem. 103 (2009) 601–608. [11] I. Kostova, Anticancer Agents Med. Chem. 9 (2009) 827–842. [12] H. Keller, B. Keppler, US Patent, 1989, 4843069. [13] B.K. Keppler, K.-G. Lipponer, B. Stenzel, F. Kratzin, Metal Complexes in Cancer Chemotherapy, VCH, Weinheim, 1993, p. 187.

310