Thioredoxin decreases anthracycline cardiotoxicity, but sensitizes cancer cell apoptosis

Abstract

Cardiotoxicity is a major limitation for anthracycline chemotherapy although anthracyclines are potent antitumor agents. The precise mechanism underlying clinical heart failure due to anthracycline treatment is not fully understood, but is believed to be due, in part, to lipid peroxidation and the generation of free radicals by anthracycline-iron complexes. Thioredoxin (Trx) is a small redox-active antioxidant protein with potent disulfide reductase properties. Here, we present evidence that cancer cells overexpressing Trx undergo enhanced apoptosis in response to daunomycin. In contrast, cells overexpressing redox-inactive mutant Trx were not effectively killed. However, rat embryonic cardiomyocytes (H9c2 cells) overexpressing Trx were protected against daunomycin-mediated apoptosis, but H9c2 cells with decreased levels of active Trx showed enhanced apoptosis in response to daunomycin. We further demonstrate that increased level of Trx is specifically effective in anthracycline toxicity, but not with other topoisomerase II inhibitors such as etoposide. Collectively these data demonstrate that whereas high levels of Trx protect cardiomyocytes against anthracycline toxicity, it potentiates toxicity of anthracyclines in cancer cells.

Introduction

Anthracyclines are highly potent chemotherapeutic agents that have been widely used in the treatment of leukemia, lymphoma, breast cancer, lung cancer, ovarian cancer, thyroid cancer, neuroblastoma, Wilms’ tumor, acute lymphoblastic leukemia and many other types of cancers including treatment of childhood cancers (1–3). In addition, anthracyclines are major components of several multidrug regimens for metastatic breast carcinoma (MBC), Hodgkin’s disease (HD), lung cancers and leukemia’s. For example, the most widely used combination for MBC being doxorubicin with cyclophosphamide and 5-fluorouracil (5-FU), commonly designated FAC or CAF (4), and epirubicin with cyclophosphamide and 5-FU, commonly designated FEC (5). Additionally, ABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine) is both more effective and less toxic than other regimens used in HD disease (6). This regimen has become the current standard for the treatment of advanced HD. In addition, anthracyclines are also used in childhood cancers; however, some of survivors of childhood cancers treated with anthracyclines develop clinical heart failure (CHF) or other cardiac complications in later part of their lives due to cardiotoxicity of anthracyclines (7, 8).

A particular limitation of anthracycline chemotherapy is the dose-dependent cardiotoxicity that can lead to CHF (9, 10). The problem of anthracycline-induced CHF is an important public health concern as it may not be seen for many years and remains a life-long threat (11). Additionally, among serious cardiac complications that have been reported are: 1) arrhythmias; 2) myocardial necrosis causing dilated cardiomyopathy; 3) vaso-occlusion or vasospasm resulting in angina or myocardial infarction (10). It is of particular importance in children who may survive for decades after successful antineoplastic treatment. The precise mechanism underlying CHF is not fully understood, but is believed to be due, in part, to lipid peroxidation and the generation of free radicals by anthracycline-iron complexes. The therapeutic activity of anthracyclines is mediated by their insertion into the DNA of replicating cells, causing DNA fragmentation, inhibition of polymerases and decreased DNA, RNA and protein synthesis. The mechanism of myocardial damage is unlikely to involve the same mechanism, since myocytes are not actively replicating (9).

Trx is a low molecular weight protein (12kD) that is widely distributed; Trx is found within the cytoplasmic, membrane, extra cellular and mitochondrial cellular fractions (12, 13). The Trx system includes Trx and Trx reductase (TrxR1). We have previously shown that besides being an antioxidant itself (14–16) Trx plays an important role in regulating the expression of other antioxidant genes, such as manganese superoxide dismutase (17). Trx has also been shown to be freely permeable to cells and could be secreted using a leaderless pathway (18). We have previously shown that overexpression of Trx increases apoptosis, and enhance the generation of O₂^.− in the presence of daunomycin or doxorubicin in lung and breast cancer cells (19, 20). Trx induces redox-cycling of anthracyclines and enhances p53-dependent apoptosis of cancer cells (20). Additionally, Trx also induces the redox-cycling of 1,2-Naphthoquinones via its redox active Cys32 and Cys35 (21). Based on these reports we hypothesize that overexpression of Trx would protect cardiomyocytes, but will enhance toxicity of anthracyclines in cancer cells.

In this report we present evidence that whereas cancer cells overexpressing Trx undergo enhanced apoptosis by daunomycin, cells overexpressing of redox-inactive mutant Trx show decreased apoptosis in response to daunomycin. In contrast, rat neonatal cardiomyocytes, H9c2 cells were protected against daunomycin-mediated apoptosis. However, H9c2 cells with decreased levels of active Trx undergo increased apoptosis in response to daunomycin. Taken together, these data suggest that Trx enhances the apoptotic death of cancer cells in response to anthracycline treatment, but protects cardiomyocytes against daunomycin toxicity. This pro-oxidant and pro-apoptotic role of Trx in presence of anthracyclines is not only novel, but also quite intriguing because Trx is widely accepted as an antioxidant and protects cells and tissues against oxidant-mediated apoptosis.

Materials and Methods

Reagents, cell culture and adenovirus production

MCF7 cells were cultured in DMEM with 10% fetal bovine serum and 100 units of penicillin/streptomycin. MCF7 clones expressing Trx, dominant negative redox inactive Trx and only vector (Vector) were cultured in DMEM containing G418 (300 μg/ml). HCT116 cells (colon cancer), A431 cells (epidermoid cancer) and U937 cells (leukemic cells) were obtained from ATCC and propagated in F12K or DMEM medium. AdenoX system was obtained from Stratagene Corporation (La Jolla, California) and Trx or mutant Trx ORF (22) were cloned into pAdenoX vector. Recombinant virus was allowed to infect HEK293 cells for generation of viral particles. For transfection, cells were infected with approximately 1X10⁸ infectious units (per million cells) and after 48 hours protein expression was determined using ELISA. Stable clones of H9C2 cells were produced by transfecting pCMV-Trx and pCMV-dnTrx constructs and selecting the stable clones with G418 selection. The stable clones of other cell lines were generated in a similar manner and clones were cultured in the presence of various concentrations of G418 as done previously (20). Human bronchial epithelial cells (BEAS-2B) were obtained from ATCC and were grown in RPMI1640 and stable clones of vector and clones expressing high levels of Trx or dnTrx were generated using G418 selection (300–400μg/mL). Normal mammary epithelial cells (HMEC) were obtained from ATCC and were maintained in RPMI 1640 media. HMVEC were obtained from Clonetics, Inc and were cultured in endothelial basal medium with addition of Bullet kit (Clonetics). H1299 cells were obtained from ATCC and were transfected with pCMV-p53 expression vector to overexpress the expression of p53 using Fugene 6 reagent (Roche) and the clones were selected using G418 selection.

Electrophoresis and Western Analysis

Protein lysates were prepared using radio immunoprecipitation assay (RIPA) buffer containing 5% sodium deoxycholate, 1% SDS, 1% Igepal in PBS with protease inhibitors and protein concentration was determined using Biorad protein assay reagent (Biorad). Equal amounts of protein was resolved on 10% SDS- polyacrylamide gel electrophoresis, and transferred onto nitrocellulose membrane (Hybond-ECL, Amersham Pharmacia Biotech). The blot was treated with appropriate dilutions of primary antibody (PARP, Trx, p53, β-actin) and visualized using either Lumiglo (Cell Signaling Technology, Beverly, MA) or ECL plus system (Amersham Pharmacia Biotech, Piscataway, New Jersey) with appropriate HRP conjugated secondary antibody.

Apoptosis measurement by Annexin V labeling in Flowcytometry

To detect the apoptosis of H9c2 cells by flowcytometry, we have used Annexin V-FITC Apoptosis Detection Kit from Abcam (Cat. No. ab14085). Briefly, after treating cells with daunomycin (1.0 μM) for 16h, they were trypsinized, washed with 1x annexin-V binding buffer. Then the cells were incubated with Annexin V-FITC and propidium iodide (PI) for 5min, washed, fixed in 2% PFA for 20 min and analyzed by FACScalibur

Measurement of Apoptosis by Electron Paramagnetic Resonance Spectrometry

To determine the apoptosis in cultured H9c2 cells, we used EPR based annexin-V-paramagnetic iron detection assay method using annexin-V magnetic microbeads kit from Miltenyi Biotec GmbH, Germany (Cat. No. 130–090-201). After treating cells with daunomycin, they were trypsinized, washed with 1x annexin-V binding buffer supplied by the manufacturer. Then the cells were incubated in 100 μL of annexin-V microbead suspension and incubated at 2–4°C for 20 minutes. At the end of incubation period, cells were loaded in a capillary tube and the annexin-V bound to cells was quantified by measuring conjugated iron spins using Bruker EMX Nano spectrometer at room temperature. EPR spectra were acquired under following scan conditions: microwave frequency, 9.63 GHz; power, 0.32 mW; attenuation 25 dB; modulation frequency, 100 kHz; modulation amplitude, 4.00 G; sweep time, 60 s; time constant, 20.48 s; receiver gain, 40 dB; magnetic field, 1610–4610 G. Absolute spin counts from spectra were calculated using Quantitative EPR module of Bruker Xenon Nano 1.3 software.

Statistical Analysis

All cell culture studies were performed in triplicate and repeated at least twice. The results are expressed as mean ± SEM (standard error of mean). Data were statistically analyzed by analysis of variance for multiple means with Tukey’s post hoc analysis. Student’s t-test was used to compare two means. Prism software (Version 8.0) was used for all statistical analyses.

RESULTS

Cancer cells from different tissues overexpressing Trx undergo increased apoptosis in response to daunomycin

We overexpressed Trx in HCT116 (colon carcinoma cells), A431 (epidermoid cancer cells), and U937 cells (leukemic cell line) by either transient methods using adenoviral delivery of Trx gene as done before (20), or by generating stable clones of Trx as described previously (23). The level of Trx activity has been described in the clones of HCT116 and MCF-7 cells (23). We used cleavage of PARP (89Kd product) as a marker of apoptosis in these cells. Since PARP is cleaved by caspases that has been used as a hallmark of apoptosis (24, 25) we used this marker as a determinant of caspase-dependent apoptosis in cells. As shown in Fig 1A & B, HCT116 cells stably expressing Trx showed higher level of PARP cleavage compared to vector alone or dnTrx cells. In contrast, cells deficient in Trx showed less PARP cleavage compared with vector control cells. To demonstrate whether other cancer cells also demonstrate similar response to daunomycin in the presence of high levels of Trx we used U937 and A431 cells. As shown in Fig 1C and ID, these cell lines also demonstrated similar response to daunomycin in the presence of high levels of Trx. As shown in Fig 1E, the expression of Trx in Trx or dnTrx overexpressed cells was increased in the U937, HCT116 or A431 clones compared to vector-only clones. Overexpression of dnTrx produces mutant redox-inactive Trx, but both Trx and dnTrx are immunoreactive to Trx antibody. Taken together, data presented here show that overexpression of Trx in cancer cells of various tissue origin undergo significantly higher levels of apoptosis compared to cells with vector control cells or cell overexpressing redox-inactive mutant Trx, suggesting increased activity of Trx is required for increased apoptosis of daunomycin treated cells.

Figure 1. Thioredoxin overexpression increases PARP cleavage in response to daunomycin in cancer cells.

(A) HCT116 stable clones expressing Trx and dnTrx were generated using G418 selection. These cells were treated with 1 μM daunomycin for 16 hours. Cell lysates were prepared and PARP and cleaved PARP were detected using PARP specific antibodies (Cell Signaling Technologies) in western analysis. (B) Densitometry of cleaved PARP. *P<0.05 vector vs. vector plus daunomycin, ANOVA; **P<0.05, Trx vs. Trx plus daunomycin (C) A431 cells were infected with AdenoX-lacz, Adx-Trx or Adx-dnTrx as mentioned for Fig 1B. After 48 hours cells were treated with 1 μM daunomycin for 16 hours. Following incubation, cell lysates were prepared and PARP and cleaved PARP were detected using PARP specific antibodies (Cell Signaling Technologies) in western analysis; (D) U937 cells were infected with adenox-lacZ or AdX-Trx. After 48 hours cells were treated with indicated concentration of daunomycin for 8 hours. Following which PARP western analysis was performed as described for Fig 1A. (E) Expression of Trx in cell types shown in response to either stable expression or adenovirus expression of Trx using western analysis;

Effect of Trx overexpression on daunomycin-mediated p53 expression in MCF-7, MDA-MB231, and HMEC

Daunomycin has been shown to induce p53-dependent apoptosis of cancer cells. We determined whether Trx-mediated enhanced apoptosis of cancer cells would be dependent on p53. MCF-7 cells contain wildtype p53 (p53^+/+) and are estrogen positive (ER^+/+−). MDA-MB-231 cells are ER^−/− and contains mutated p53. We evaluated whether breast cancer cells with ER^−/− and p53 (non-functional p53) would show similar response to daunomycin in the presence of high levels of Trx. In addition, we also used normal mammary epithelial cells (HMEC), and MCF-7 cells for transient transfections of Trx to determine whether these cells would produce similar results in transient transfections, as obtained with stable clones. As shown in Fig 2A, transient transfection of Trx or dnTrx increased Trx expression in MCF-7 and MDA-MB-231 cells, but HMEC cells showed relatively a small increase in Trx expression. As shown in Fig 2B MCF-7 cells transiently overexpressing Trx showed higher p53 levels in response to daunomycin. In contrast, cells overexpressing mutant Trx showed decreased p53 expression in response to daunomycin (Fig 2B). Similar results were obtained with MDA-MB-231 breast cancer cells (Fig. 2B, middle panels). However, HMEC did not show significant change in the levels of p53 in vector or Trx overexpressed cells. Our data with HMEC indicate a differential role of Trx in the potentiation of anthracyclines in cancer cells to that of non-transformed cells. Additionally, we tested whether Trx overexpressing clones of MCF-7 cells, Trx9(20) would undergo enhanced apoptosis by evaluating the release of cytochrome c levels in the cytosol. As shown in Fig 2C, stable clones of MCF-7 cells overexpressing Trx (Trx9) showed significant release of cytochrome c in the cytosol compared to vector only stable clones. Collectively, these data further show that enhanced daunomycin toxicity in response to high levels of Trx is independent of ER status or p53 status, but depends upon transformed cells, but not normal cells.

Figure 2. Overexpression of Trx modulates p53 expression in MCF-7, MDA-MB231 and normal human mammary epithelial cells (HMEC) in response to daunomycin.

(A) MCF-7, MDA-MB231 or HMEC were transiently transfected with pCMV-Trx or pCMV-dnTrx or empty vector. After 48 hours cells were treated with daunomycin (0.1 or 0.05 μM) for 16 hours. Cell lysates were prepared and p53 western blotting was performed as described earlier. (B) The expression of p53 and β-actin in MCF7 (top panel), MDA-MB321 (middle panel) and HMEC cells (lower panel); (C) The release co cytochrome c in MCF-7 cells with increased levels of Trx.

Cells with wildtype p53 undergo enhanced PARP cleavage, whereas p53^−/− cells show a delayed response to daunomycin-mediated PARP cleavage

In all of our experiments we used cells that are positive for wildtype p53 and found that upregulation of p53 is an important response to anthracycline toxicity. Therefore, we sought to determine whether cells lacking p53 such as H1299-p53^−/− (homozygous deletion of p53) would undergo apoptosis in response to daunomycin. As shown in Fig 3A, cells lacking p53 did demonstrate cleaved PARP after 24 hours in response to daunomycin suggesting that p53-independent pathways in cells constitutively lacking p53 protein. To determine the role of p53 in daunomycin-mediated apoptosis, we stably transfected H1299 cells with wildtype p53 expression vector and selected p53 overexpression clones using G418. H1299 cells lacking p53 and complemented with p53 were exposed to a range of concentration of daunomycin for 8 hours. As shown in Fig 3B cells lacking p53 failed to show PARP cleavage (except a minor increase in cleaved PARP in 1μM daunomycin) in 8 hours treatment with daunomycin. In contrast, cells complemented with wildtype p53 showed strong PARP cleavage in response to daunomycin treatment. These results show that while p53 could rapidly induce apoptosis in H1299 cells (PARP cleavage), cells lacking p53 have a delayed effect in the onset of apoptosis. These data suggest that an alternate delayed pathway become operational in the absence of p53 as a compensatory mechanism.

Figure 3. Daunomycin induces PARP cleavage at higher concentration in p53^−/− cells, but lower concentrations of daunomycin induces PARP cleavage in p53^+/+ cells:

(A) H1299 cells lacking p53 were treated with daunomycin (1μM) for 24 hours. Following treatment western analysis for PARP was done as described previously. (B) H1299 cells were stably transfected with a p53 expression vector and H1299^−/− cells and H1299 cells complemented with wildtype p53 were treated with daunomycin and PARP western analysis was performed as mentioned previously.

Trx specifically enhances anthracycline-mediated PARP cleavage via caspase activation, but this effect does not occur in other topoisomerase II inhibitors

To understand whether apoptosis is the major mechanism of cell death in response to anthracyclines in the presence of Trx, we inhibited caspase activity by broad-spectrum caspase inhibitor Z-VAD-FMK, and treated these cells with daunomycin, and evaluated the expression of cleaved PARP as a marker of apoptosis. As shown in Fig 4A, inhibition of caspases inhibited PARP cleavage in HCT116-Trx cells, demonstrating that apoptosis is the principal mechanism of cancer cell death by anthracyclines in Trx enriched cells. One of the most widely accepted mechanisms of action of anthracyclines is their inhibitory effect on topoisomerase II. Thus, we determined whether Trx would affect the activity of topoisomerase II that will increase the potency of anthracyclines. To analyze the effect of Trx on topoisomerase II, we used another widely used topoisomerase II inhibitor etoposide and determined PARP cleavage in HCT116 cells overexpressing Trx or its mutant form. As shown in Fig 4B, etoposide induced PARP cleavage in vector only transfected cells. However, there was a significant increase in PARP cleavage in cells expressing mutant Trx. In contrast, Trx overexpressed cells demonstrated protection against PARP cleavage, demonstrating that high levels of Trx in fact protect against etoposide-induced apoptosis. These data show that the effect of Trx is not at the level of topoisomerase II. Thus, the mechanism of action of Trx is independent of inhibitory action of anthracyclines on topoisomerase II action. Rather, increased levels of Trx may protect cells from damaging action of topoisomerase II inhibitory action as shown by our etoposide data (Fig 4B).

Figure 4.

(A) Etoposide-induced PARP cleavage does not increase in presence of higher levels of Trx, but PARP cleavage increases in cells overexpressing mutant Trx. HCT 116 cells stably expressing Trx or dnTrx were either untreated or treated with 100 μM etoposide for 16 hours. Following incubation cell lysates was prepared and PARP western analysis was performed as described earlier (B) Anthracycline-mediated PARP cleavage is inhibited by caspase inhibitor Z-VAD-FMK: HCT116-Trx cells were pre-treated with Z-VAD-FMK (20 μM) for 2-hours, followed by treatment with 1 μM daunomycin for 16 hours. Cell lysates were prepared and PARP western analysis was performed as described earlier.

Rat embryonic cardiomyocytes H9C2 cells overexpressing high levels of Trx show decreased PARP cleavage in daunomycin treatment, but increased PARP expression occurs in cells deficient in Trx

Cardiotoxicity is a paramount concern in anthracycline chemotherapy and is a limiting factor in the use of anthracyclines. To evaluate the role of Trx in anthracycline toxicity in cardiomyocytes we generated stable clones of H9C2 cells overexpressing either Trx or dnTrx and treated these cells with anthracyclines (Fig 5A). Cardiomyocytes did not show increased PARP cleavage in response to daunomycin with higher level of Trx (Fig 5A & B). In contrast, H9C2 cells with Trx deficiency (dnTrx) showed increased PARP cleavage in presence of high level of mutant Trx, demonstrating that active Trx protects cardiomyocyte against daunomycin-mediated apoptosis (Fig 5A & B). We also found significant cell death in H9c2-dnTrx cells in compared to H9C2-Trx cells in response to H₂O₂ (Fig 5C). The level of Trx and dnTrx in stable clones of H9c2 cells is shown in Fig 5C. Next we determined the effect of higher levels of Trx and dnTrx on PARP cleavage or p53 expression in non-malignant cells to evaluate whether cancer cells are specifically sensitive to daunomycin in presence of Trx. For these experiments we used primary cell lines of human microvascular endothelial cells (HMVEC) or an airway epithelial cell line not derived from cancer tissue (BEAS-2B). As shown in Fig 5E, BEAS-2B cells with increased Trx levels showed decreased PARP cleavage compared to dnTrx-expressing BEAS-2B cells. We also determined the effect of daunomycin on p53 expression in HMVEC overexpressing Trx or dnTrx to demonstrate whether p53 expression would remain unaffected due to Trx or dnTrx overexpression. As demonstrated in Fig 5F, increased Trx and mutant protein expression was observed in HMVEC transfected with Trx or dnTrx constructs (Fig 5F). However, increased levels of Trx or dnTrx protein had no effect on p53 expression in response to daunomycin in HMVEC. These data show that Trx specifically enhances the PARP cleavage or p53 expression in cancer cells, but not in normal cells in response to daunomycin.

Figure 5. Decreased level of PARP cleavage in H9c2 and other primary cell lines expressing higher level of Trx in response to daunomycin.

(A) We generated stable clones of H9c2 cells using pCMV-Trx and pCMV-dnTrx constructs and selecting them in G418 as mentioned in the methods. These cells were treated with 1 μM daunomycin for 16 hours following which PARP western analysis was performed as described in the methods. (B) Densitometry of cleaved PARP depicted in Fig 5A. *Significantly higher, P<0.05, Vec + 1μM Dan vs. Trx+Dan; **Significantly higher, P<0.05, Trx+Dan vs. dnTrx+Dan (C) Photomicrograph of stable clones of H9c2 cells treated with 100 μM H₂O₂, demonstrating enhance damage to dnTrx-H9c2 cells; (D) expression of Trx or dnTrx in stable clones of H9c2; (E) PARP cleavage in BEAS-2B cells with high levels of Trx or dnTrx; (F) Expression of Trx or dnTrx in HMVEC; (G) the expression of p53 in HMVEC overexpressing Trx or mutant Trx in response to daunomycin.

H9C2 cells overexpressing high levels of Trx are protected against daunomycin-induced apoptosis, but cells with decreased level of redox active Trx undergo enhanced apoptosis

Although PARP cleavage suggests enhanced apoptosis, we further performed the effect of daunomycin on apoptosis of H9c2 cells by annexin V labeling assay by flow cytometry, and also by EPR technique using Fe²⁺-annexin labeling. As demonstrated in Fig 6A, H9c2 cells overexpressing Trx demonstrated decreased levels of apoptosis compared to vector only or dnTrx transfected H9C2 cells (33.6% for Trx; 54.6% for vector-only cells and 44.5%% for dnTrx cells). To further confirm the effect of Trx on daunomycin mediated toxicity of cardiomyocytes, we utilized a novel and sensitive EPR assay as described in the “methods” section. As shown in Fig 6B–C, high levels of Trx decreased daunomycin-mediated apoptosis of H9c2 cells compared with either vector or dnTrx transfected cells as shown by absolute spin counts of EPR spectra that shows binding of magnetic beads to annexin-V. The level of PARP cleavage was also decreased in Ad-Trx infected cells compared to either vector or dnTrx infected H9c2 cells (Fig 6D). Taken together, these experiments establish that H9c2 cells are protected against daunomycin-mediated apoptosis.

Figure 6. hTrx over expression protects H9c2 cells from daunomycin-induced cell apoptosis:

(A) H9c2 cells were infected with Ad-LacZ, Ad-Trx and Ad-dnTrx adenovirus for 36h, then treated with daunomycin (0.1 μM) for 16h. Apoptosis was detected by flow cytometry with Annexin V-FICT apoptosis detection Kit from Abcam as mentioned in the methods; (B) To quantitate apoptotic cells, they were incubated with annexin-V-paramagnetic iron conjugate and the annexin-V bound cells were quantified by measuring electron paramagnetic resonance of annexin-V conjugated iron as described in the methods. (C) Absolute spin count was calculated from EPR spectra of Fe bound to Annexin-V and plotted as bar graph. Values are means ± SD. *, p < 0.01 versus Ad-LacZ; **, p < 0.01 versus Ad-LacZ + daunomycin. (D) hTrx overexpression prevents daunomycin induced PARP cleavage. H9c2 cells were infected with Ad-LacZ, Ad-Trx and Ad-dnTrx adenovirus for 36h, then treated with daunomycin (0.1 μM) for 16h. Cell lysate was collected and analyzed for PARP, Trx and β-actin by western blotting.

Discussion

In this report, we have established that cancer cells from various tissue origins undergo enhanced apoptosis due to daunomycin treatment in the presence of high levels of Trx. However, normal cells including embryonic cardiomyocytes did not undergo enhanced apoptosis due to daunomycin treatment in the presence of high levels of Trx as determined by PARP cleavage. This apoptosis potentiation by Trx is limited only to anthracyclines, as etoposide, a topoisomerase II inhibitor did not increase apoptosis in the presence of increased Trx levels. In contrast, Trx protected against etoposide-induced apoptosis. Collectively, these findings indicate a specific effect of Trx in potentiating anthracycline toxicity in cancer cells, but not in cardiomyocytes.

The heart is particularly vulnerable to free radical injury because protective enzymes are present in lower levels than in other tissues (26). Consequently, the damage to myocardial cells may eventually lead to irreversible heart failure (26). It has been proposed that an effective way to avoid the cardiotoxic effects of anthracyclines is to prevent cardiac injury during chemotherapy. Three approaches have been attempted: first; decreasing myocardial concentrations of anthracyclines and their metabolites by dose limitations or schedule limitation (27); second, by developing less cardiotoxic anthracycline derivatives and formulations (28); and third, by administration of cardioprotective agents during or after chemotherapy to attenuate the effects of anthracyclines on the heart (29). Our studies with rat embryonic cardiac myocytes, H9C2 cells show that overexpression of Trx protects against daunomycin-induced PARP cleavage and apoptosis. Additionally, a study has shown that Trx overexpression in the heart of Trx-transgenic mice protects against anthracycline cardiotoxicity (30).

Our observations contradict a previous report that Trx confers resistance to anticancer drugs due to its antioxidant properties (31, 32). This study published by Wang et al is a correlative study that demonstrated a correlation between increased Trx expression in various leukemia cell lines and the drug resistance to adriamycin. However, there was no cause and effect relationship between high levels of Trx and adriamycin-related drug resistance. Therefore, there was no mechanistic evaluation of the role of enhanced expression or depletion of Trx in drug resistance or sensitivity. In contrast, Berggren et al (33) has clearly demonstrated that although Trx protects cells against H₂O₂-mediated apoptosis, it does not protect against adriamycin-induced apoptosis. In the same study they have shown that Trx overexpression increased the expression of peroxiredoxin (Prx), which could remove H₂O₂ (33). Thus, it is clear that while H₂O₂ could be removed in Trx-overexpressed cells, thereby preventing H₂O₂-mediated apoptosis, Trx overexpression failed to protect cells against adriamycin-induced apoptosis. In our studies we have observed that Trx-overexpression not only failed to protect against daunomycin-mediated apoptosis, but also caused enhanced apoptosis in cancer cells (20). Additionally, other studies have shown that Trx overexpression could result in enhanced cell death in response to interferon (34). Further, Trx has been shown to increase the activity of caspases (35), which may also contribute to increased apoptosis of cancer cells via activation of caspases.

The most plausible explanation for the differential role of Trx to daunomycin toxicity could be explained by the specific mode of action by which anthracyclines kill cancer cells. The major mechanism of action of anthracyclines is to inhibit topoisomerase II that could prevent DNA unwinding for replication and trigger an apoptotic response. Second, the quinone moiety of anthracyclines intercalates with the DNA that inhibits DNA replication. The semiquinone radical that is produced due to Fe²⁺-mediated redox-cycling of anthracyclines produces increased ROS that may also trigger DNA damage (Fig 7) and induction of apoptotic signaling. All of these biochemical effects could occur in cancer cells as they are rapidly dividing and could intercalate quinone into their DNA while replicating (Fig 7). However, these mechanisms do not occur in cardiomyocytes as these cells do not divide and would therefore not intercalate with quinone moiety of anthracyclines. However, they may undergo enhanced production of ROS due to Fe²⁺ mediated redox-cycling of anthracyclines (Fig 7). High levels of Trx in cardiomyocytes would induce MnSOD expression and activity (17, 36), resulting in rapid dismutation of O₂^.− to H₂O₂. High levels of Trx could prevent enhanced H₂O₂ production and subsequent hydroxyl radical formation due to removal of H₂O₂ by increased peroxiredoxins expression (33). Thus, whereas Trx could enhance the anthracycline redox cycling by activating a bio reductive enzyme such as Cyt p450 reductase in cancer cells (19), the mechanism of redox-cycling of daunomycin would not occur in cardiomyocytes as they are non-cycling cells. Consistent with this concept, a study has shown redox-cycling of 1,2-Naphthoquinone by Trx via Cys32 and Cys35, which are catalytic cysteines on Trx (21). On the contrary, Trx could protect against daunomycin-mediated ROS generation due to scavenging of superoxide anion and H₂O₂-Fe²⁺ -mediated hydroxyl radicals via increased peroxiredoxins expression (Fig 7).

Figure 7. Possible mechanisms associated with Trx-mediated enhanced anthracycline toxicity in cancer cells, but decreased toxicity in cardiomyocytes:

Anthracyclines redox cycles within the cancer cells due to autooxidation or via bio reductive enzymes resulting in semiquinone radical formation, which intercalates with the DNA and causes DNA damage and apoptosis in proliferating cancer cells. The redox cycling of anthracyclines is increased in the presence of Trx, which could be an electron donor for bio reductive enzymes resulting in high levels of semiquinone radical formation and apoptosis in these cells. The adult cardiomyocytes are non-cycling, and therefore intercalation of anthracyclines will be much lower, although redox cycling would be enhanced with high levels of Trx. However, increased superoxide anion (O₂^.−) production will occur in presence of high levels of Trx due to redox cycling. In the cardiomyocytes high levels of Trx would produce increased MnSOD and peroxiredoxins (Prx), which would neutralize superoxide anion as well as H₂O₂ resulting in decreased oxidative stress in cardiomyocytes. Since cardiomyocytes are terminally differentiated and do not divide the intercalation process does not occur in these cells. Therefore, the damage to cardiomyocyte only occurs by increased superoxide anion production due to redox cycling. High levels of Trx would protect cardiomyocytes against increased production of H₂O₂ and iron-catalyzed hydroxyl radical (.OH) formation via MnSOD and Prx.

In conclusion, we have shown that increased levels of Trx promotes the apoptosis of cancer cells with concomitant protection cardiomyocytes, suggesting that Trx or its derivatives could be used as a potential adjuvant in anthracycline chemotherapy with reduced cardiotoxicity. One of the limitations of this study is the lack of animal experiments to conclusively establish cancer cell apoptosis with decreased cardiotoxicity in a specific cancer such as leukemia, breast cancer or other. However, we are currently testing the efficacy of anthracyclines in animal models with increased Trx levels to establish our findings in vivo.