The electrochemistry of quinizarin revealed through its mediated reduction of oxygen

The reduction of quinone functionality either by a one or two electron processes is known to produce reactive species, which are able to readily reduce oxygen (Fig. 1A) either to superoxide () or hydrogen peroxide (H₂O₂). When the reduction of the species occurs at an electrode surface this mechanism results in a catalytic cycle and voltammetrically, a large irreversible redox signal is observed (13). The reduction of the QZ species upon a BDD electrode was investigated in both the presence and absence of dissolved oxygen, where the potential was swept at a scan rate of 0.5 V s^-1 between -0.2 V and -0.85 V (vs. SCE), as shown in Fig. 2, where the QZ concentration is 480 nM. It should be noted that because of the low solubility of the QZ molecule within solution the use of a Teflon cell was required to avoid loss of material through adsorption onto glass surfaces. In the absence of oxygen a small surface-bound voltammetric peak was observed at -0.535 V (vs. SCE), and this peak is ascribed to the reversible two electron, two proton reduction of the quinone species (Fig. 1A). It is proposed that because of the low solubility of QZ it readily adsorbs upon the BDD electrode surface, and hence the direct reduction of adsorbed QZ is measurable until a solution concentration of roughly 200 nM. From the charge under the surface-bound voltammetric peak in the presence of 480 nM QZ it is possible to estimate the surface coverage of the QZ molecule as being roughly 2.5 ± 0.5 × 10¹² molecules cm^-2. From X-ray crystallographic data the molecular volume is found to be ca. 260 Å³ (18). Using estimated geometric dimensions we can calculate the expected full monolayer coverage of QZ upon the BDD surface to be 1.5–4.0 × 10¹⁴ molecules cm^-2, depending on the orientation of the species (vertical or horizontal). Hence, this value suggests that at 480 nM the surface coverage of the QZ species is roughly 0.5–2% of a full monolayer coverage.

In the presence of oxygen, a large irreversible voltammetric feature is observed because of the mediated reduction of oxygen (Fig. 1A). In this study pure oxygen has been used, the application of the following procedures to in vivo studies will likely require the use of lower oxygen concentrations. Significantly the voltammetric response exhibits a sharp switching-off of the current at -0.776 V (vs. SCE). This rapid shift is indicative of the presence of a phase transition; given that material is known to be adsorbed upon the electrode it is reasonable to conclude that this sharp voltammetric feature is related to a change in the orientation of the surface-bound groups at more negative potentials. Further this phase transition also clearly causes the catalytic redox cycle to become nonoperative. This phase transition may be either the desorption of material or a change in alignment of the QZ upon the electrode surface, as is known for quinoline, aniline, and other organic species (19, 20). As the potential is decreased during the reverse scan the material returns to its original orientation switching on the mediated reduction pathway, leading to a rapid increase in current [-0.717 V (vs. SCE)]. Moreover, the voltammogram exhibits hysteresis in the potentials at which the phase transition occurs upon the forward and reverse scan. Given that the surface coverage of the QZ species is known to be very low (0.5–2%) and that if the QZ distribution upon the surface is assumed to be homogenous the redox centers may be expected to be ca. 2 nm apart. At such large interatomic distance, interactions between the redox centers will be minimal, thus the rapidity of the phase change is likely caused by the presence of negative differential resistance leading to bistability within the electrochemical system (21).

The mediated peak current was measured as a function of QZ concentration (Fig. 3), where it was found that measurable peak currents were found for concentrations as low as 5 nM (13.2 ± 3.6 × 10^-6A). At higher concentrations the peak current shows a plateau; this limiting of the current is not related to a limitation in the adsorption of the QZ on to the electrode surface. It was found that the voltammetric peak area for the nonmediated process varied linearly with QZ concentration beyond a value of 480 nM. The limitation in the mediated process is more likely related to the consumption of the available oxygen. From literature it is known that the solubility of oxygen within the aqueous solutions is approximately 1.24 mM (22) and the diffusion coefficient has been measured as 1.77 × 10^-5 cm² s^-1 (23). Consider the Randles–Ševčík equation for an irreversible process:where i_p is the peak current (amperes), n is the total number of electrons, n^′ is the number of electrons transferred before the rate determining step (RDS), α_RDS is the transfer coefficient for the rate determining step, A is the area of the electrode (cm²), C_o is the bulk concentration of the analyte (moles cm^-3), D_o is the diffusion coefficient (cm² s^-1), and ν is the scan rate (V s^-1). The value of (n^′ + α_RDS) has been set as 0.5 as measured experimentally from Tafel analysis (i.e., the first electron transfer is the rate determining step). The equation predicts that the maximum peak for the reduction of oxygen is approximately 1.1 × 10^-4A (this value is depicted as the horizontal line on the inlay of Fig. 3). As can clearly be seen at relatively high concentrations of QZ the voltammetric peak current approaches the theoretical maximum. This conclusion is further corroborated by the fact that the peak current for the mediated oxygen reduction peak at 480 nM is found to vary with the square root of scan rate (Fig. 3, Inset). As discussed above, the surface coverage of the QZ at the higher concentrations is known; consequently, it is possible to estimate through measurement of the charge passed for the mediated reduction pathway, that on the forward scan each QZ molecule present upon the surface is on average reduced approximately 300 times.

Having investigated the fundamentals of the system, we address the QZ chemistry. The main accepted mode of chemotherapeutic action for the anthracycline antibiotics is through intercalation of the quinonal moiety into dsDNA and subsequent inhibition of the topoisomerase II enzyme (24). To demonstrate the intercalative abilities of QZ, the voltammetric response of an oxygenated solution containing 60 nM QZ was measured, in the presence of increasing concentrations of solution-phase DNA (Fig. 4). The inlay of Fig. 4 depicts the associated voltammograms under conditions in which there is zero (red) and 25 μM (black) DNA. Clearly as the DNA concentration increases the measured voltammetric response for the mediated oxygen reduction decreases. This observed decrease is due to the sequestration of the QZ through intercalation. It is to be noted that this decrease in voltammetric response is not due to blocking of the electrode with increasing DNA concentration, as confirmed through the addition of higher concentrations of QZ to a 25 μM solution of DNA upon which the mediated voltammetric signal returned.

It is of interest, that given the reported high binding constant for QZ with DNA (25), a relatively large excess of DNA is required to titrate the QZ. This observation suggests that the adsorption of the QZ to the electrode surface is causing a decrease in the measured sensitivity of the system. This interference of the BDD electrode may be best understood in terms of both the preconcentration of QZ onto the electrode surface and also the high number of available adsorption sites present upon the BDD. The preconcentration of the QZ species upon the electrode increases the local effective concentration to roughly 0.6 μM (based on an approximate diffusion layer thickness of 10 μm). This value is consistent with the DNA binding to roughly 1 in 40 base pairs. Moreover, the QZ should not be expected to bind to the DNA in a 1∶1 ratio because of the likely influence of longer-range allosteric effects (26, 27). Further, if we assume that it is possible to attain a near monolayer coverage of QZ, then within the diffusion layer (10 μm) the number of BDD sites, as compared to the intercalative sites in the DNA, will be greater even at DNA concentrations of 25 μM and above. The presence of the nonzero peak current for the mediated oxygen reduction at high concentrations of DNA is also likely a result of the QZ adsorption. Even given this lower sensitivity toward DNA, this current electrochemical method exhibits significantly lower levels of detection than obtainable by comparable UV-visible-based methods (25, 28). This situation arises predominantly because of the ability to study far lower concentrations of QZ in this work.

The main metabolic route for the removal of anthracycline antibiotic from the body is via reduction of the carbonyl group (C-13, as labeled on Fig. 1) (29), but because of the presence of the hydroquinone group within the anthracycline (situated on the quinizarin) the species is also susceptible to oxidative degradation (Fig. 1B). Biologically this reaction is known to occur both directly via oxidation through a peroxidase (30) or alternatively indirectly via the oxidized products of the enzymes (31, 32). Further it has been proposed that the oxidative degradation pathway of the anthracycline species may provide a possible route by which cardiac protection may be attained (8).

The direct oxidation of the QZ species was investigated at the BDD electrode in the presence of dissolved oxygen (0.5 V s^-1). At relatively high QZ concentrations (480 nM) the irreversible oxidation of surface-bound species is observed with a half-wave potential of +0.386 V (vs. SCE) (Fig. 5, Inset, blue dotted line). This voltammetric feature corresponds to the 2e^-, 2H⁺ oxidation of the hydroquinone group within the QZ structure (Fig. 1B). In most experimentally measured scan rates no back-peak is observed as the succeeding chemical steps leading the decomposition of the species are rapid. Only at higher scan rates does the evidence of a minimal reverse peak become apparent (as can be seen in Fig. 5, Inset), which may be due to the direct reduction of the products or may also be related to the presence of electroactive decomposition products. The potential of the irreversible oxidative wave was measured as a function of pH where it was confirmed that the peak varied with approximately 59 mV/pH, demonstrating that during the oxidation an equal number of protons and electrons are transferred, further corroborating the conclusion that the peak relates to the 2e^-, 2H⁺ oxidation of the hydroquinone group.

To demonstrate that the oxidation of the species is both related to the QZ and further causes loss of the quinone species, the reductive voltammetric scan was performed (compare to Fig. 2) but here the system was preconditioned at a more positive potential for 15 s prior to running the voltammogram. The results of this experiment can be seen in Fig. 5, where the conditioning potential has been varied between -0.1–0.9 V (vs. SCE). As the potential is systematically increased there is a clear decrease in the measured reduction peak current upon the conditioning potential reaching the potential for the oxidation of the QZ species. This decrease occurs over a potential range of 130–170 mV, hence it may be concluded that the electron transfer is quasireversible in nature. It should be noted that even at high conditioning potentials a nonzero peak current is measured. This current is related to the presence of unreacted QZ present upon the electrode surface, which accumulates via diffusional replenishment after the oxidative potential has been applied. Futhermore, as demonstrated above only nanomolar quantities of QZ are required for a mediated reduction peak to be observed. Hence this experiment has demonstrated that the observed voltammetric feature at +0.386 V (vs. SCE) corresponds to the irreversible oxidation of the QZ species leading to its decomposition.

The use of electrochemical methods for the investigation of biologically relevant systems is often compromised by levels of low sensitivity and selectivity. We have experimentally demonstrated how, through harnessing the catalytic oxygen reduction mechanism, it is possible to observe the presence of the poorly soluble molecule QZ in solution down to 5 nM (∼100 ppt). Further, as QZ forms an active part of the anthracycline antibiotics, the measured electrochemistry gives insight into its biological function. Specifically, this work confirms that oxidation of the quinone functionality cause the degradation of the species and results in the mechanistic pathway that leads to the formation of ROS to be switched off. This oxidation mechanism was found to be pH-dependent confirming that the electrochemical mechanism is associated with proton transfer. This work has wider significance in that it provides direct evidence of the redox properties of the quinone functionality present within anthracycline antibiotics, if this oxidative degradation pathway can be exploited biologically and within the heart it has potential to decrease the cardiotoxicity of future chemotherapy treatments.

All chemicals were purchased from Sigma-Aldrich. Solutions were prepared using deionized water of resistivity not less than 18.2 MΩ cm^-1 at 298 K (Millipore). Because of the low solubility of the QZ, stock solutions of 0.2 and 0.02 mM QZ in ethanol were prepared. Aliquots of these solutions were added to the aqueous buffer solution (0.1 M phosphate, pH 6.84) to achieve the low QZ concentrations required. The calculation of the low molecular weight DNA (salmon sperm) concentration was done upon the basis of the average base pair molecular weight being 650 gM^-1.

The voltammetric measurments were recorded using an Autoloab PGSTAT20 computer-controlled potentiostat (Eco-Chemie). A standard three-electrode configuration was used throughout, consisting of a BDD electrode, a platinum wire (99.99%, GoodFellow), and a saturated calomel electrode (SCE, Radiometer) acting as the working, counter, and reference electrodes, respectively. The BDD electrode was polished by using alumina (1.0, 0.3 μm, Buehler), rinsed, and sonicated for 1 min in deionized water between each experiment. The electrochemical experiments were performed in a Teflon beaker and the solution was either fully oxygenated or deoxgenated by bubbling O₂ or N₂ through the solution for 15 min, prior to experimentation.