Multi-component redox system for selective and potent antineoplastic activity towards ovarian cancer cells

Abstract

Ovarian cancer is the deadliest gynecological cancer which rarely causes symptoms, and goes undetected until reaching the advanced stage of drug-resistant metastases. The cationic porphyrin meso-tetra(4-N-methylpyridyl)porphine (TMPyP) is a well-known photosensitizer (PS) used in photodyamic therapy (PDT) for curing cancer due to its strong affinity for DNA and high yield of reactive oxygen species (ROS) upon light activation. The practicality to irradiate tumor cells alone in the physiological system being slim (due to the close proximity of healthy cells and tumors), we looked for a variation in the PDT using a mixture of TMPyP with 1,5-dihydroxynapthalene (DHN) and Fe(III) ions at a mole ratio of 1:20:17 (drug combo) respectively in aqueous solution. The drug combo needs no photoactivation in H₂O₂ rich environment (mimicking the microenvironment of cancer/tumor), where it generates ȮH and juglone, the latter being a known potent anticancer agent. In vitro studies of the drug combo in drug resistant and sensitive ovarian cancer cell lines showed drastic growth inhibition and cell death compared to normal epithelial cells. The drug combo provides an effective and non-invasive alternative to conventional PDT, exploiting the cytosolic carcinogenic H₂O₂ to produce an efficient anticancer treatment. The unique action of cancer-specific cytotoxicity arises from the redox chemistry involving activation of Fe(III) as the oxidizing agent to generate juglone, which utilizes the cytosolic ROS in cancer cells against itself.

1. Introduction

The most common cause of gynecological cancer-related death is ovarian cancer, a malignant tumor that represents a serious threat to women’s health, and has the highest mortality among all gynecological tumors [1–4]. Among 295,000 new cases of ovarian cancer in 2018 worldwide, the death rate was 62.73% [3,4]. While the 5-year survival rate for breast cancer increased to 90% in 2015 compared to 1985, it only increased by 45% in ovarian cancer during the same time interval [5]. Absence of specific clinical symptoms in the early stages, along with lack of effective early screening result in the diagnosis of 70% of ovarian cancer patients only at the advanced stage (stage III–IV) [6]. Like most tumor tissue, OVCA manifests excessive oxidative stress [7] which further creates an opportunity to therapeutically target the redox signaling and reactive oxidative stress (ROS) mediated cell death. ROS alters metabolic pathways, (including glycolysis, and lipid metabolism) in cancer cells [7,8]. Therapeutic strategies to target oxidative stress in a selective manner is therefore an effective approach to promote growth inhibition and promote apoptosis.

Porphyrins, derived from porphines [9], are tetrapyrollic organic pigments which play key roles in diverse biological processes, including hemoproteins (involved in oxygen transport and storage) [10], chlorophylls (vital in photosynthesis) [11], and photosensitizers (for biomedical applications) [12]. Porphyrins are known to produce reactive oxygen species (ROS) upon light activation, a modality used widely in Photodynamic Therapy (PDT) [12]. The high hydrophobicity of porphines has prompted the synthesis of cationic porphyrin meso-tetra(4-N-methylpyridyl)porphine (TMPyP) (Fig. 1) which has improved aqueous solubility [13]. In the last two decades, the cationic porphyrin TMPyP has come into prominence as a novel PDT agent to treat cancers [14] including cervical [15], breast [16], melanoma [17], leukemia [18], lung [19], ovary [20], colon [21] and larynx models [22]. In cellular studies TMPyP is found to accumulate into lysosomes [23] when used at low concentrations. Upon illumination, TMPyP produces ROS, particularly, singlet oxygen, which immediately reacts with lysosomes [24] and ruptures the lysosomes releasing TMPyP into the nucleus where it binds with DNA with a high affinity and cleave DNA [25].

Fig. 1.

Top. 3-D structures for TMPyP, DHN, and Juglone; atoms are color-coded as grey (carbon), white (hydrogen), blue (nitrogen), red (oxygen). A. Absorption spectra of TMPyP (alone green), with DHN (red) and with DHN and FeCl₃ (black) in water (solid line). Figure Ainset highlights the absorption region between 500 and 700 nm. B. Fluorescence emission (upper panel) and excitation (lower panel) spectra of TMPyP (6 μ M, green line); TMPyP (6 μ M) with DHN (120 μ M) (red line); drug combo (TMPyP (6 μ M) with DHN (120 μ M) and FeCl3 (100 μ M) (black line)) in aqueous solution. There are 3 distinct regions (A, B, C) of fluorescence spectra of the drug combo where A denotes λ_ex = 350 nm (top) emission and λ_em = 420 nm (bottom) excitation, B denotes λ_ex = 420 nm (top) emission and λ_em = 715 nm (bottom in the Soret band region) excitation and C denotes λ_ex = 520 nm (top) and λ_em = 715 nm (bottom in the Q band region) excitation spectra.

Some combination treatment strategies using TMPyP conjugated with photosensitizers and other anticancer drugs/agents have been found effective [23,26] to obtain synergistic anti-tumor effects and reduce resistance to therapy [27]. However, these combinations have limitations which include: high systemic toxicity, development of resistance to therapy, complex synthetic procedures involved for preparation, limited aqueous solubility, and inability to produce ROS [28]. Thus, the development of a suitable nontoxic combination system based on TMPyP is highly desirable.

Zamadar and coworkers have recently developed a method to control the growth of pathogen using the photosensitization reaction of TMPyP with DHN [29]. It is reported that cancer cells have intrinsically elevated levels of H₂O₂ compared to the normal cells due to its fast proliferation and altered metabolism [30]. H₂O₂ creates significant molecular changes to the nucleic acids, proteins, and lipids [31] in diseased cells. In this paper we report the design and therapeutic effect of a novel porphine-based catalytic redox multi-component system that utilizes the carcinogenic oxidative stressor itself to generate a potent anticancer agent in situ. The therapeutic design is based on taking advantage of the propensity of metallo porphyrins (TMPyP) to engage in a catalytic redox chemical process (with Fe(III) and DHN) to generate an anticancer agent which is the product of a redox reaction. This approach highlights viability of a novel multi-component drug system wherein the active anticancer agent generation occurs selectively in cancer cells due to elevated levels of cytosolic ROS compared to normal cells.

2. Materials & methods

Triple distilled water, TMPyP, DHN, FeCl3 were obtained from Sigma Aldrich Inc. (MO, USA). Stock solutions of TMPyP and FeCl₃ were made in water. DHN stock solution was made in 9:1H₂O: acetonitrile solution. The working concentrations of the drug combo was 6, 120 and 100 μM of TMPyP, DHN and FeCl3 respectively.

2.1. Cell culture studies

Human ovarian cancer cell lines, HeyA8 and HeyA8MDR cell lines were purchased from MD Anderson cancer center, TX. Mouse embryonic fibroblast (MEF) cell was a generous gift from Professor Reuben Shaw of Salk Institute, La Jolla, CA. Cells were grown in RPMI-1640 supplemented with 10% fetal bovine serum (FBS) and 1% antibiotic in a humidified atmosphere at 37 °C, with 5% CO2.

2.2. Cell growth and ROS assay

Cells (0.5–1 × 10⁴) were seeded in 96-well plates and exposed to drug combo for 24 h, along with the individual drugs. The inhibitory concentrations of 50% (IC50) values were determined by MTT assays as previously described [32]. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT, 100 μg per well) was added to culture medium after drug treatment and further incubated for 4 h at 37 °C. Generated formazan crystals were dissolved in DMSO and optical density was measured at 550 nm in a multimode reader (Varioskan, USA). Cells were imaged post-treatment using a brightfield inverted microscope (Thermo Fisher) in 20× magnification. Multiple images were taken from each well to study the effects of drug on the cellular morphology. Reactive oxygen species (ROS) were measured in control and treated cells by the H2DCFDA dye as described earlier [32].

2.3. Steady state absorption and fluorescence

Steady state absorption and fluorescence measurements were performed to characterize the TMPyP in free and with DHN and with both DHN and FeCl₃ in aqueous system. Steady state absorption spectra were recorded with a Shimadzu UV 2550 spectro-photometer. Steady state fluorescence measurements were carried out with a PerkinElmer FL 6500 fluorimeter. Excitation and emission slit widths were mainly 2.5/5 (5/10 nm for studying the Q-bands [33]) as indicated in Fig. 1.

2.4. Agarose gel electrophoresis

The p202 plasmid (size 6517 base pairs) was used as a model DNA to study the effect of the drug combo in the dark and upon photo activation. 2 μL of the stock DNA (0.4 μg/mL) was added to 98 μL of mixture containing the following: i. TE buffer, ii. drug combo mixture (1X) (TMPyP with DHN and Fe(III) ions at a mole ratio of 1:20:17), and iii. 50% (v/v) of drug combo with water (0.5X). After thorough mixing, the solutions were divided into two parts, one was left in the dark while the other was irradiated with a blue (447 nm) laser for 10 min. Samples were then analyzed on a 2% agarose gel at 100 V for 30 min.

3. Results and discussion

3.1. Fluorescence spectroscopic studies on the drug combo system

The aqueous solution of TMPyP strongly absorbs in the blue wavelength zone of the visible spectrum where its Soret band [29] appears at 422 nm and weak Q-bands at 518 nm, 554 nm, and 590 nm, as is shown in Fig. 1A (green profile). Upon addition of DHN and FeCl₃, there is a slight decrease in the overall absorbance including that in the Q-band region, with the occurrence of the absorption bands of DHN around 300 nm (the red and black profiles respectively in Fig. 1A). The absorbance of the Q-bands at 550 and 586 nm regions increased in the combo state compared to TMPyP alone.

The typical fluorescence emission (top) and excitation (bottom) profiles of TMPyP and the drug combo is provided in Fig. 1B. Figure 1BI (top) provides the emission profile with λ_ex = 350 nm where the drug combo (black line) showed emission at −410 nm. The excitation spectra (BI bottom) at λ_em = 350 nm shows the excitation at −300 nm which indicates that coming from DHN as was observed in the absorption profile in Fig. 1A. Upon excitation at the Soret band region of 420 nm (BII top), TMPyP (alone, green) and in combo mixture (black and red (no Fe)) showed the main emission −712 with a shoulder at −660 nm where the fluorescence intensities reduced between free and combo states. This is indicative of some interactions between the drug combo ingredients. The lower panel in BII shows the excitation spectra at λ_em = 715 nm with the ${\lambda }_{ex}^{max}$ at −420 nm, indicative of TMPyP responsible for the emission at 715 nm (as is seen in Fig. 1A). Excitation of Q-band of TMPyP at λ_ex = 520 nm produced a spectrum with emission bands at 666 nm and 696 nm as is displayed in Fig. 1BIII (top). Using a higher slit, the excitation spectra of the drug combo with λ_em = 715 nm showed the excitation bands at around 592 (Q-band), 623, 655 nm between the region 600–700 nm. According to De Baróid et al. [33], the Q-band around 590 nm is responsible for producing more ¹O₂ compared to the Soret band. It is evident from the Fig. 1A inset and BIII excitation spectra that the drug combo still has the ability of producing ¹O₂ by the 590 nm Q band.

Generation of Hydroxyl radical (ȮH), and Juglone and Removal of Excess H₂O₂ by the Drug Combo in Dark H₂O₂ rich environments.

We carried out experiments to determine the optimum concentration of H₂O₂ and Fe(III) ions to optimize Fenton-like [34] reaction for producing ȮH in situ in aqueous solution under dark condition. Fig. 2A showed the decrease of absorption of DHN at 301 nm when the solution of TMPyP/DHN/Fe(III) was treated with varying amounts of H₂O₂ while the concentrations of Fe(III) ions, DHN and TMPyP were kept at 100 μM, 100 μM, 6.0 μM respectively. A maximum decrease of absorption of 1,5-DHN at 301 nm was observed when 400 μM H₂O₂ was used (see in Fig. 2A and B). Similarly, experiments were carried out to find out an optimum concentration of Fe(III) ion by varying the concentration of Fe(III) ions while the concentrations of TMPyP (6.0 μM), DHN (100 μM), and H₂O₂ (400 μM) were kept constant. Fig. 2C shows the changes of absorptions of DHN at 301 nm with varying concentrations of Fe(III) ions. A maximum decrease of absorption of DHN at 301 nm was observed when Fe(III) concentration was 25 μM (see Fig. 2D). As observed in a standard Fenton-like reaction [34] (see Fig. 2E), H₂O₂ reacts with Fe(III) ions and forms Fe(III)-peroxo complexes, which later decomposes into Fe(II) ions and HȮ₂ radicals. The produced Fe(II) ions then reacts with H₂O₂ to form reactive ȮH radicals via standard Fenton reaction. On the other hand, the produced HȮ₂ radicals react with another HȮ₂ radicals or Fe(III) ions or Fe(II) ions and produce O₂, or Fe(II) and O₂, or [Fe^IIIHO₂]²⁺, respectively [35]. The reaction of 1,5-DHN and TMPyP/Fe(III) ions in H₂O₂ aqueous solution suggests a Fenton-like reaction that generates ȮH in situ and oxidizes 1,5-DHN in the dark conditions forming Juglone or derivatives of Juglone in solution. The control reactions of 1,5-DHN with TMPyP and H₂O₂ and of 1,5-DHN with H₂O₂ alone revealed no detectable 1,5-DHN oxidation in the dark conditions (data not shown). This data suggest that Fe(III) ions and H₂O₂ are the required reagents for the generation of ȮH radicals in aqueous solution.

Fig. 2.

A. Optimization of H₂O₂ concentration in the presence of Fe(III) ions (1 × 10⁻⁴ M), DHN (1.0 × 10⁻⁴ M) and TMPyP (6.0 × 10⁻⁶ M) in aerobic, aqueous solution under dark conditions using the absorbance of DHN at 301 nm. B. The UV–vis spectra of DHN recorded at the optimum H₂O₂ concentration of 400 μM where a maximum decrease of absorption of DHN at 301 nm was observed in Fig. 2A. C. Optimization of iron (III) concentration in the presence of 400 μM iron (III) in aerobic, aqueous solution under dark conditions. D. The UV–vis spectra of the recorded optimum iron (III) concentration at 25.0 μM. E. Possible mechanism for Fenton-like reaction.

3.2. Drug combo inhibits cell proliferation and triggers acidification of cells in the cancer cells

We treated the cancer cells (HeyA8 and HeyA8MDR) and non-cancerous mouse embryonic fibroblast (MEF) cells with an increasing concentrations of drug combo (1/32x, 1/16x, 1/8x, 1/4x, 1/2x and 1x dilution of the drug). Cells were treated for 24 h followed by MTT assay to detect cell viability. Results show that cancer cells have a growth inhibition with only 1x dilution of the drug whereas non-cancerous MEF cells are not affected by this treatment (Fig. 3A). This led to further investigate the morphology of the cells after drug treatment (see Fig. 3B), which depicted morphological changes in cancerous cells indicating early apoptotic events, where the non-cancerous cells remain unaffected. Next, we stained the drug combo treated cancer cells with LysoTracker dye, which clearly shows an increase in acidic compartments in the cells, which is an indirect indication of autophagic up-regulation in the cells (Fig. 3C).

Fig. 3.

A. MTT assay was performed to detect HeyA8, HeyA8MDR and MEF cell viability after treating with drug combo. B. Microscopic images to detect any variation in the cell morphology (magnification: 20x). C. Effects of drug combo on intracellular acidic compartments were evaluated by Lysotracker Deep Red staining after 24 h of treatment. Nuclei were stained using Hoechst 33342. Pictures were acquired by fluorescence microscopy (magnification: 40x).

3.3. Drug combo mediated cell growth inhibition is reactive oxygen species mediated

We analyzed ROS levels in HeyA8, HeyA8MDR and MEF cells using H2DCFDA dye after treating cells with combo drug using 0.5x and 1x dilution. Fig. 4A shows elevated levels of ROS in HeyA8 and HeyA8MDR cells when treated with 0.5x and 1x drug combo. MEF cells did not show any elevated ROS level when treated. To prove if combo mediated ROS generation is critical for cell growth inhibition we have pre-treated cells with a ROS inhibitor N-acetyl-l-cysteine (NAC). Pretreatment with NAC (ROS inhibitor) significantly reduced drug combo mediated cell growth inhibition in HeyA8MDR but not in HeyA8 cells (see Fig. 4B).

Fig. 4.

A. Reactive oxygen species are detected in untreated and drug combo-treated HeyA8 (i), HeyA8MDR (ii) and MEF (iii) cells using DCFDA. B. Cell viability is tested in HeyA8 and HeyA8MDR cells by MTT assay in the presence of NAC or drug combo or NAC + drug combo. Statistical analysis for A and B done using one way ANOVA with Tukey’s Multiple comparison test. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. C. DNA gel electrophoresis of 1. p202 in TE buffer unexposed to light; 2: p202 in TE buffer, exposed to blue laser; 3: p202 treated with 0.5x combo, unexposed to light; 4: p202 treated with 0.5x combo, exposed to blue laser; 5: p202 treated with 1x combo exposed to blue laser; 6: p202 treated with 1x combo, unexposed to light; 7: p202 in TE buffer, unexposed to light.

Our results depict that drug combo induced cell death is triggered by elevated oxidative stress which damages DNA. As TMPyP is known to have a strong affinity for DNA [23,24], we explored the effect of drug combo on DNA. Fig. 4C presents the results of blue laser irradiation of the plasmid p202 DNA in absence or presence of (0.5x and 1x) drug combo. Partial and complete DNA damage were observed for irradiated samples at 0.5x and 1x dose, as seen in lanes 4 and 5 respectively in Fig. 4C. The same unexposed solutions did not show any change compared to the control (DNA in buffer). The drug combo produced free radicals and oxidative stress upon excitation at the Soret band region (around 420 nm) which cleaved the DNA.

The chemical mechanism that is responsible for generation of juglone is based on the redox chemistry between TMPy-Fe³⁺ … OOH and DHN. This quaternary system needs the TMPyP chelator, Fe(III) oxidizer, ROS as electron mediator, and DHN as reducing agent. Specific ratiometric combination of all four components in definite proportion leads to remarkable cancer cell death caused by juglone. This suggests that juglone is not produced unless DHN is oxidized by H₂O₂ in presence of Fe³⁺ and TMPyP. Based on our own experimental data and knowledge of metallo-porphyrin redox chemistry [36,37], it could be predicted that Fe³⁺ chelates into porphyrin binding site, which in the presence of hydrogen peroxide forms a complex with higher oxidation potential that oxidizes DHN to juglone. Though H₂O₂ offers several advantages including a moderate oxidation potential, reaction selectivity, bio-friendly byproduct (water), it is met with two main disadvantages in the form of lack of chemical stability and low reaction rate [37]. In presence of TMPyP and Fe³⁺, H₂O₂ serves as the electron acceptor resulting in efficient generation of juglone. Till date, there is no such study on the use of metallo-TMPyP along with ambient H₂O₂ as oxidants, making this a pioneering work on using redox chemistry in treating cancer.

4. Conclusions

In conclusion, we report a concoction of four easily accessible substances (Fe³⁺, porphine derivative, hydrogen peroxide, and dihydroxynaphthalene) to create a potent anticancer agent that is selective in its cytotoxic activity. Through systematic variation in relative proportions of the individual components, we were able to formulate a specific molar ratio of the components that exerts its highest biological effect. Spectroscopic study of the quaternary system yielded insights into the chemical dynamic aspects of the four-component mixture, which is responsible for its anticancer effect towards ovarian cancer, which is one of the most aggressive forms of cancer. The biological effect of this drug combo was evaluated through analyzing its selective cytotoxic potentials and morphometric analysis in chemosensitive and chemoresistant ovarian cancer cells. We intend to explore the influence of other metal ions, porphine and DHN derivatives in their ability to exploit intracellular ROS to generate anticancer agents within the cancer cells. It is anticipated that the drug combo effects described herein can be modified appropriately to screen against other types of cancer.