Preferential DNA Cleavage under Anaerobic Conditions by a DNA Binding Ruthenium Dimer

Abstract

In the absence of O₂, the cationic complex, [(phen)₂Ru(tatpp)Ru(phen)₂]⁴⁺ (P⁴⁺), undergoes in situ reduction by glutathione (GSH) to form a species that induces DNA cleavage. Exposure to air strongly attenuates the cleavage activity, even in the presence of a large excess of reducing agent (e.g., 40 equiv GSH per P⁴⁺) suggesting the complex may be useful in targeting cells with a low oxygen microenvironment (hypoxia) for destruction via DNA cleavage. The active species is identified as the doubly reduced, doubly protonated complex H₂P⁴⁺ and a carbon-based radical species is implicated in the cleavage action.. We postulate that the pO₂ regulates the degree to which carbon radical forms and thus regulates the DNA cleavage activity.

The use of transitions metal complexes in medicine has enjoyed extensive attention given the tremendous success of cisplatin as a chemotherapeutic agent¹ and the ability of many metal complexes to interact with and damage cellular structures, particularly DNA.^2-7

A large number of DNA cleaving metal complexes function via the activation of dioxygen to generate reactive oxygen species (ROS), such as hydroxyl radical and superoxide radical.^8,9 These ROS are ultimately responsible for the DNA cleavage. Others, including cisplatin and certain photoactivated,^10-14 oxidizing^15,16 or hydrolyzing complexes,⁸ do not require O₂ to function, but they are also insensitive to the cellular [O₂]. Compounds that show enhanced cleavage activity under a low oxygen microenvironment (hypoxia) are rare^17-21 but offer a unique mechanism to target tumor cells under such conditions. These hypoxic tumor cells are often the most resistant to radiotherapy^22,23 and chemotherapy^24,25 and the most susceptible towards metastasis,^26,27 making this subpopulation a particularly attractive chemotherapeutic target.

We have discovered that the cationic ruthenium dimer, [(phen)₂Ru(tatpp)Ru(phen)₂]⁴⁺ (P⁴⁺) (tatpp = 9,11,20,22 -tetraazatetrapyrido[3,2-a: 2′,3′-c: 3″,2″-1: 2‴,3‴-n]-pentacene and phen = 1,10-phenanthroline) shown above (water soluble as the chloride salt) not only induces DNA cleavage in the presence of mild reducing agents but shows enhanced activity under anaerobic conditions. The fact that exposure to air attenuates the cleavage activity suggests that ROS are not responsible for the observed cleavage and that such a complex might be useful in targeting cells under hypoxic conditions. Complex P⁴⁺ is known to intercalate and bind DNA tightly (K_b 1.1×10⁷ M⁻¹ at 25 mM NaCl).^28,29 The strong interaction with DNA is not unusual for this class of cationic complexes and it has a number of structural similarities to many known metallointercalators^13,14,30-33 including those that are know to thread their way through the DNA double-helix.³⁴

The ability of P⁴⁺ to cut DNA was examined by following the conversion of supercoiled plasmid DNA (form I) to the circular form (form II) or linear form (form III) using agarose gel electrophoresis to separate the products (experimental details given in ESI). As shown in Figure 1, P⁴⁺ alone does not cause appreciable DNA cleavage (lane 2), however, addition of a mild reducing agent such as glutathione (GSH) leads to cleavage activity (lanes 4 &5). However the yield of cleavage products is clearly higher under anaerobic conditions (compare lane 4 vs lane 5) Yields of cleavage products (Forms II+III) under aerobic and anaerobic conditions are 55 % and 97 %, respectively.³⁵ The appearance of linear DNA in lane 5 appears to result from sequential ss cuts, not ds cleavage, thus the overall cleavage activity is single-strand (ss) sission.

Figure 1.

Cleavage of supercoiled pUC18 DNA (0.154 mM) by P⁴⁺ in 7 mM Na₃PO₄ buffer (pH 7.0) at 25°C. Lane M, marker lane containing Form I, II and III DNA; lane 1, DNA control; lane 2, DNA + P⁴⁺ (0.0128 mM); lane 3, DNA + GSH (0.513 mM); lane 4, DNA + GSH (0.256 mM) + P⁴⁺ (0.0128 mM) under aerobic conditions; lane 5, same as lane 4 under anaerobic conditions; lane 6, DNA + GSH (0.512 mM) + Fe-Blm (0.0128 mM) under aerobic conditions; lane 7, same as lane 6 under anaerobic conditions. Incubation time for all these cases was limited for 1 hour.

Given the importance of excluding trace O₂ as playing a role in the observed cleavage activity, a positive control was included. Under anaerobic conditions, Iron(II)-Bleomycin (Fe-Blm) is known to induce DNA nicks but not ds cuts. When exposed to O₂, however, Fe-Blm is an effective ds-nuclease.^36-38 As seen in Figure 1, lane 6, Fe-Blm in the presence of O₂ causes extensive DNA-ds breaks whereas when O₂ was excluded (lane 7), only ss nicking is observed. These studies were carried out side by side with the P⁴⁺ cleavage experiments demonstrating unequivocally that P⁴⁺ is a more effective DNA cleaving agent under reducing and hypoxic conditions. Due to the known photoreactivity of P⁴⁺, all experiments were conducted in the dark so that photochemically induced cleavage reactions could be ruled out. All cleavage experiments were conducted under low light conditions as P⁴⁺ is known to be photoactive under some conditions.^39-41 Control experiments conducted in the dark or under ambient laboratory lighting gave identical results, showing that this cleavage reaction is not a photochemical reaction.

We have previously examined the redox chemistry of P⁴⁺ in water at various pH's by electrochemical, spectroelectrochemical and chemical reduction methods.^39-41 Reaction 1 shows the two reduction products of P⁴⁺ in water at pH 7.0. The redox reactions are reversible and easily followed by visible spectroscopy as distinct changes in the absorption spectrum are observed for each reduction and protonation event. It is readily apparent that GSH reacts with P⁴⁺ as the solution color changes from yellow to green

$${\mathbf{P}}^{4+}\underset{-e^{-}}{\overset{+e^{-}}{\rightleftharpoons }}{\mathbf{P}}^{3+}\underset{-e^{-}-2{\mathrm{H}}^{+}}{\overset{+e^{-}+2{\mathrm{H}}^{+}}{\rightleftharpoons }}{\mathbf{H}}_2{\mathbf{P}}^{4+}$$ (1)

upon addition of the GSH. As seen in Figure 2, the absorption spectra of P⁴⁺ in aqueous buffer (pH 7.0) after addition of GSH is identical to that of H₂P⁴⁺ as prepared by stoichiometric reduction and protonation or electrochemically.^40,42 Thus, it appears that P⁴⁺ is a prodrug, which is converted to the H₂P⁴⁺ by in situ reduction.

Figure 2.

Absorption spectra of P⁴⁺ (12.8 μM) before (solid line) and after (dotted line) addition of 10 equiv. GSH in anaerobic 7 mM Na₃PO₄ buffer (pH 7.0). Dashed line is the absorption spectrum of H₂P⁴⁺ in MeCN when prepared by stioichiometeric cobaltocene reduction and trifluoroacetic acid protonation.⁴²

In order to identify the chemical species responsible for the observed anaerobic cleavage and to rule out participation by glutathyl radical species, we examined the cleavage activity of the P⁴⁺, P³⁺ and H₂P⁴⁺(see ESI for synthetic procedures) under anaerobic conditions without GSH present.^40,42

As seen in Figure 3, P⁴⁺ does not induce appreciable DNA cleavage under aerobic or anaerobic conditions (lanes 2 and 3). P³⁺ (lanes 4 and 5) does show some enhanced nicking ability, however, H₂P⁴⁺ is clearly the most potent nicking agent (lanes 6 and 7). As seen in lanes 6 and 7 (Figure 3), the amount of DNA cleavage increases with increased [H₂P⁴⁺] as would expected if this complex is the actual cleaving agent. Thus one simple explanation for the attenuated cleaving activity under aerobic conditions would be reoxidation of H₂P⁴⁺ to P⁴⁺. Exposure of an aqueous solution of H₂P⁴⁺ to air is known to result in a rapid reoxidation of this complex to P⁴⁺, as measured by UV-visible absorbtion spectroscopy.³⁹

Figure 3.

Agarose gel of supercoiled pUC18 DNA (0.154 mM) in the presence of P⁴⁺, P³⁺ and H₂P⁴⁺. All incubations were performed under anaerobic conditions with an incubation time of 2 h at 25°C. The ratio of complex to DNA-bp was (a) 0.083 or (b) 0.20 as indicated above each lane. Lane M, marker; lane 1, DNA control; lane 2, DNA + P⁴⁺ (0.0128 mM); lane 3, DNA + P⁴⁺ (0.0307 mM), lane 4, DNA + P³⁺ (0.0128 mM); lane 5, DNA + P³⁺ (0.0307 mM), lane 6, DNA + H₂P⁴⁺ (0.0128 mM); lane 7, DNA + H₂P⁴⁺ (0.0307 mM).

The mechanism of DNA cleavage is still unclear, however Yamaguchi and coworkers have shown that dihydropyrazines cleave DNA by both oxygen dependent and independent pathways.^43-46 H₂P⁴⁺ contains the dihydropyrazine substructure and thus could function in a similar manner with the high DNA binding affinity further enhancing its activity. For dihydropyrazines, the oxygen independent cleavage activity is attributed to the formation of a carbon-based radical on the drihydropyrazine moiety. Thus, we speculate that DNA bound H₂P⁴⁺ is cleaving DNA in a similar manner via the generation of reactive carbon-based radicals that are in close proximity to the DNA due to the tight DNA binding.

To test, this hypothesis, the cleavage activity of H₂P⁴⁺ was examined in the presence of various radical trapping and metal complexing reagents. DMSO is an effective scavenger of diffusible oxygen-based radicals such as ^.OH, and superoxide.^47,48 On the other hand, TEMPO (2,2,6,6-tetramethyl-1-piperdinyloxy) is an effective scavenger of carbon radicals or metal-based radicals but ineffective with oxygen-based radicals.^49,50 As seen in Figure 3, addition of up to 5% DMSO by volume has no effect on the cleavage activity (lane 3) whereas addition of 2 mM TEMPO stops most of the DNA cleavage (lane 4). This data clearly supports the role of carbon-based radicals in the cleavage mechanism. Yamaguchi and coworkers postulated that trace metals ions, such as Cu²⁺, activated the dihydropyrazines to the reactive form.^43-46 This does not seem to be the case here as added EDTA, at concentrations up to 1 mM, has little effect on the cleavage activity (Figure 4, lane 5) of H₂P⁴⁺. We are further investigating this unusual behavior and hope to elucidate the cleavage mechanism with the help of EPR spectroscopy. We note that the carbon-based radical species would likely be very reactive towards O₂ in solution and this ‘quenching’ reaction could also explain the observed sensitivity of this cleavage activity to oxygen.

Figure 4.

Agarose gel of supercoiled pUC18 DNA (0.154 mM) in the presence of H₂P⁴⁺ (0.0256 mM). All incubations were performed in 7mM Na₃PO₄ buffer (pH 7.0) at 25⁰C under anaerobic conditions with an incubation time of 2 h. The ratio of complex to DNA-bp was 0.16. Lane M, marker lane containing Form I and II DNA; lane 1, DNA control; lane 2, DNA + H₂P⁴⁺; lane 3, DNA + H₂P⁴⁺+ DMSO (0.64 M); lane 4, DNA + H₂P⁴⁺ + TEMPO (2.04 mM); lane 5, DNA + H₂P⁴⁺ + EDTA (1.02 mM).

It is interesting to observe that the loading of P⁴⁺ onto the DNA in the following studies (ranging from 1 complex per 12 DNA-bp to 5 complexes per DNA-bp) is relatively low compared to many metal-based DNA cleaving agents. In fact, the [Fe Blm] used in the cleavage experiment shown in Figure 1 was purposefully kept the same as the [P⁴⁺] in the same experiment. As can be seen, the cleavage activity of P⁴⁺ under reducing, anaerobic conditions is comparable with Fe-Blm under aerobic conditions. To our knowledge, this is the first example of a metal complex with potentiated DNA cleavage activity under hypoxic conditions, suggesting potential therapeutic applications. Future studies will establish the mode of action and it effects on tumor cells in vitro.