Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Cardiotoxicity is an important factor that limits the clinical use of DOX, one of the most effective anticancer agents. The proposed mechanism for DOX cardiotoxicity is the production of reactive oxygen species during its intracellular metabolism. Therefore, many studies have focused on the effects of antioxidant systems on DOX toxicity in the heart. Because GSH is involved in both enzymatic and nonenzymatic detoxification of a variety of free radicals including reactive oxygen species, several investigations have explored the role of GSH in cardiac protection against DOX toxicity (Doroshow et al., 1981; Yoda et al., 1986; Villani et al., 1990). Administration of exogenous GSH to mice significantly inhibited the acute myocardial toxicity of DOX (Yoda et al., 1986). N-Acetylcysteine, which has been shown to increase intracellular GSH concentrations, also provided protection against DOX cardiotoxicity both in vitro (Villani et al., 1990) and in vivo (Doroshow et al., 1981). It is now well documented that GSH is an important factor in cardioprotection against DOX toxicity.

Acquired drug resistance in tumor cells is another major impediment for the application of DOX. GSH has been shown to be involved in the development of this drug resistance (Shen et al., 1997). A multiple drug resistant human breast cancer cell line, MCF-7, was selectively made resistant to DOX by stepwise increases in drug concentrations. Correlating with the drug resistance, GSH concentrations in the cells were significantly increased. When these cells were pretreated with BSO, a specific inhibitor of -gutamylcysteine synthetase that catalyzes the rate-limiting reaction of GSH biosynthesis, GSH content was depleted in these cells and the resistance to DOX was reversed (Dusre et al., 1989). In our recent studies, we have found that a human lung carcinoma A549 cell line, which was selectively made resistant to cadmium, contained elevated GSH concentrations and was cross-resistant to DOX toxicity. When these cells were treated with BSO, their cellular GSH was depleted and their sensitivity to DOX toxicity was increased (Hatcher et al., 1997). Many other studies have exclusively demonstrated the importance of GSH in the acquired resistance to DOX and other anticancer drugs in a variety of tumor cells (Kisara et al., 1995; Malayeri et al., 1996; Shen et al., 1997). For this reason, trials are ongoing to use BSO to deplete GSH content in tumor cells, thereby to sensitize the cells to anticancer drugs including DOX (Bailey et al., 1994; O'Dwyer et al., 1992; Kisara et al., 1995). BSO, when administered systemically, also decreases GSH content in the heart (Friedman et al., 1990). This decrease has been shown to enhance cardiotoxicity of cyclophosphamide (Friedman et al., 1990). It is inevitable that BSO-mediated GSH depletion would enhance DOX cardiotoxicity.

MT is a low molecular weight metal-binding protein containing a high proportion of cysteine residues but no disulfide bond (Hamer, 1986). It has been revealed that MT functions in cytoprotection against free radical-induced tissue damage (Sato and Bremner, 1993). In our recent studies, we have developed a transgenic mouse model in which cardiac MT was specifically overexpressed. Using this unique experimental tool, we have demonstrated that MT plays an important role in cardiac protection against DOX toxicity (Kang et al., 1997).

It is possible that the protection by MT against DOX toxicity in the heart can compensate for the loss of protection due to GSH depletion. Both GSH and MT contain high content of cysteine residues (1/3). Importantly, MT has been shown to play a role in scavenging free radicals (Sato and Bremner, 1993). Zinc-MT reacted with hydroxyl radicals in a cell-free system and was more effective than GSH in preventing hydroxyl radical-induced DNA degradation (Abel and de Ruiter, 1989). A recent study using HL-60 cells has demonstrated a direct reaction of hydrogen peroxide with the sulfhydryl groups of MT, and the thiolate groups in the MT were the preferential attacking targets of hydrogen peroxide compared to the other sulfhydryl residues from GSH and protein fractions (Quesada et al., 1996).

Our study was therefore undertaken to take the advantage of the specific cardiac MT overexpressing transgenic mice to test the hypothesis that MT can compensate for the loss of GSH in cardiac protection against DOX toxicity. The transgenic mice and nontransgenic controls were pretreated with BSO before exposure to DOX. The modulative effects of BSO on cardiac GSH and its correlation with enhanced DOX cardiotoxicity were examined. The results demonstrate that MT overexpressing transgenic mouse heart was resistant to the BSO-enhanced DOX cardiotoxicity.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. DOX (Adriamycin), GSH, NADH, NADPH, glutathione reductase, CdCl₂, malonaldehyde-bisdiethylacetal and myristlytrimethyl-ammonium bromide (HPLC grade) were purchased from Sigma Chemical Co. (St. Louis, MO). Sulfosalicylic acid and DTNB were obtained from Aldrich Chemical Co. (Milwaukee, WI). Acetonitrile (HPLC grade) was from Fisher Chemical Co. (Fair Lawn, NJ) and BCA protein assay reagents from Pierce (Rockford, IL). All other common use reagents were from either Sigma or Fisher and of the highest purity available.

Animals and drug treatment. Specific cardiac MT overexpressing transgenic mice were produced and characterized as reported before (Kang et al., 1997). These mice were bred to the same FVB stain of nontransgenic mice and the heterozygous transgenic litters and nontransgenic littermates were used for experiments. Animals were housed in a plastic cage at 23°C on a 12-hr light/dark cycle, and were given lab food and tap water ad libitum. Adult male mice (6-7 wk old, weighing 20-30 g) were randomly assigned to one of four treatment groups. These groups include 1) saline-treatment control, 2) BSO treatment only, 3) DOX treatment only and 4) the combination of BSO and DOX treatment. BSO was given by i.p. injection at 5 mmol/kg, two times with a 12-hr interval. Four hours after the second BSO treatment, some animals received 15 mg/kg of DOX dissolved in physiological saline by a single ip injection at a volume of 5 ml/kg, the control group received a sham injection of an equal volume of saline. On the 4th day of DOX postdosing, mice were anesthetized with sodium pentobarbital (65 mg/kg, Vet Labs, Lenexa, KS). Blood was collected from abdominal vena cave and serum was separated with a Serum Separator apparatus (Becton Dickenson, Inc., Rutherford, NJ) within 30 min. The heart was perfused with cold 0.9% KCl with inferior vena cava cut open, then removed, opened, washed with cold 0.9% KCl and dried with paper tissue. Tissue samples were immediately placed in liquid nitrogen. For MDA measurement, heart tissues were homogenized immediately in 4 vol of 1.15% KCl and the protein was precipitated with equal volume of acetonitrile. The whole homogenate was centrifuged first at 10,000 × g at 4°C for 30 min and the supernatant was centrifuged again at 15,000 × g for 15 min. The clear supernatant was collected for the separation of free MDA. Other samples for cardiac GSH and MT analyses were stored at 80°C no longer than 36 hr before analysis. This time course of DOX treatment and tissue harvesting are based on previous studies by us (Kang et al., 1996, 1997) and others (Kojima et al., 1993), which have shown that the level of oxidative injury in the heart, measured by lipid peroxidation and CK release, and the overt inhibitory effect on DNA, RNA and protein synthesis peaked on the 4th day. All animal procedures were approved by the AAALAC certified Institutional Animal Care Committee.

GSH and MT concentrations in mouse heart. Cardiac GSH concentrations of mice treated with BSO or saline were measured by the method of Tietze (1969). Briefly, heart tissues were homogenized in 10 vol of 5% 5-sulfosalicylic acid at 4°C. The homogenate was centrifuged at 10,000 × g for 15 min, and the supernatant was assayed for GSH by the DTNB-glutathione reductase recycling assay. The 1.0-ml reaction mixture contained 190 µl of stock buffer (143 mM sodium phosphate and 6.3 mM Na₄-EDTA, pH 7.5), 700 µl of 0.248 mg NADPH/ml in stock buffer, 100 µl of 6 mM DTNB and 10-µl sample preparation. The assay was initiated by addition of 10 µl of 266 U glutathione reductase/ml. The standard assay was done under the same conditions including the same concentration of 5-sulfosalicylic acid. The amount of GSH was determined from the standard curve, in which the equivalents present (1, 2, 3 or 4 nmol) was plotted against the rate of change of absorbance at 412 nm, and was expressed as micromoles of GSH per gram of tissue.

Measurement of DOX cardiotoxicity. Serum CK activity, an important indicator of DOX cardiotoxicity (Kang et al., 1996, 1997), was assayed as described by Oliver (1963). A CK kit (CK-20) based on this method was obtained from Sigma, and the provided instruction was followed.

Statistical analysis. Data are expressed as mean ± S.D. and analyzed by a one-way analysis of variance followed by the Scheffe's F-test. The level of significance was set at P < .01.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of BSO treatment on cardiac GSH and MT levels. Both nontransgenic and transgenic mice were treated with BSO at 5 mmol/kg body weight for two i.p. injections with a 12-hr interval between the two injections. Four hours after the second injection, the heart was removed from the anesthetized animals and cardiac GSH and MT concentrations were determined. The BSO treatment significantly (P < .01) decreased cardiac GSH concentrations. In nontransgenic mice this decrease was from 1.03 ± 0.12 to 0.46 ± 0.05 µmol/g heart, about 55% depletion. In transgenic mice it was from 1.04 ± 0.05 to 0.40 ± 0.02 µmol/g heart, about 60% depletion. There was no significant difference in the cardiac GSH concentrations between nontransgenic and transgenic mice, treated either with or without BSO (table 1). This BSO treatment, however, did not significantly alter the concentrations of MT in either transgenic or non-transgenic mouse hearts. The cardiac MT concentration was 111.62 ± 8.65 µg/g heart in transgenic mice and 5.18 ± 0.76 µg/g heart in nontransgenic mice (table 1), i.e., about 20-fold higher in the transgenic than in the nontransgenic mouse heart (P < .01).

Serum CK activity. In our previous studies (Kang et al., 1996; 1997), we have observed that measurement of serum CK was a reliable indicator of cardiotoxicity induced by DOX in the mouse model used for this study. The serum CK activity was highly correlated with DOX-induced cardiomyopathy examined by electron microscopy (Kang et al., 1997) and with the extent of cardiac lipid peroxidation (Kang et al., 1996). An increased serum CK activity was observed in nontransgenic mice treated with DOX only and this increase was significantly (P < .01) suppressed (more than 50%) in transgenic mice (fig. 1). More dramatic difference was observed between transgenic and nontransgenic mice that were treated with DOX after BSO pretreatment. The serum CK activity in the BSO- and DOX-treated nontransgenic mice was further elevated, being about 3-fold higher than in the DOX-treated only animals (P < .01). However, BSO pretreatment did not alter the DOX-induced serum CK activity in the transgenic mice (fig. 1). BSO treatment alone did not cause significant changes in the serum CK activity in either non-transgenic or transgenic mice (fig. 1).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   DOX-induced elevation of serum creatine kinase activity. Serum creatine kinase activity was measured using a CK-20 kit from Sigma. Sera were collected from saline-treated animals (BSO/DOX), DOX-treated only (BSO/DOX+), BSO pretreatment only (BSO+/DOX) and both BSO and DOX treated animals (BSO+/DOX+). The protocol for animal treatment was described in "Materials and Methods." Data represent mean ± S.D. values from six animals for each treatment. *Significantly different from DOX treated only controls (P < .01), and **significantly different from both DOX treated only and BSO plus DOX treated controls (P < .01).

Cardiac MDA concentration. MDA is a product of lipid peroxidation induced by a diversity of oxidative injury. It has been used as a biomarker of DOX-induced cardiac oxidative damage. The MDA concentrations in non-DOX-treated animals, either transgenic or nontransgenic, were not detectable by the method employed in this study (data not shown). DOX treatment alone dramatically elevated the cardiac MDA concentrations in both nontransgenic and transgenic mice, with the transgenic mouse heart displaying lower concentration, although not statistically significant (fig. 2). Corresponding to the alteration of DOX-induced serum CK activity by BSO, the cardiac MDA concentration was also significantly (P < .01) elevated by BSO pre-treatment in the DOX-treated nontransgenic mice (fig. 2). This BSO treatment again did not alter the oxidative response of the transgenic mouse heart to DOX (fig. 2).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Cardiac MDA concentrations in DOX-treated animals with or without BSO pretreatment. The level of MDA in the heart of animals treated either with saline as control (BSO/DOX) or with BSO only (BSO+/DOX) was not detectable by the method (not shown). The experimental procedure described in "Materials and Methods" was followed. Data represent mean ± S.D. values from six animals for each treatment. *Significantly different from the BSO plus DOX-treated controls (P < .01).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Cardiotoxicity has been recognized as a complicating factor of cancer chemotherapy with DOX. There are several hypotheses to explain DOX cardiotoxicity. Among these, the free radical one is the most thoroughly investigated. DOX undergoes one-electron reduction through a metabolic activation caused by NADPH-cytochrome-P-450 reductase, or other flavin-containing enzymes (Bachur et al., 1978). This reduction generates a DOX semiquinone free radical. In the presence of molecular oxygen, the semiquinone rapidly reduces the oxygen to superoxide with regeneration of intact DOX. Superoxide is rapidly converted to hydrogen peroxide spontaneously or by superoxide dismutase. The DOX semiquinone can then react with the hydrogen peroxide to yield a hydroxyl radical (Kalyanaraman et al., 1984). These highly toxic reactive oxygen species react with cellular molecules including nucleic acids, proteins and lipids, thereby causing cell damage.

The protection by increasing cardiac GSH concentrations against DOX toxicity has been demonstrated by using either exogenous GSH or precursors for GSH synthesis, as discussed previously. However, it was not known whether depletion of cardiac GSH would sensitize this organ to DOX toxicity. This issue was addressed for the first time in our study. Treatment of nontransgenic mice with BSO significantly decreased cardiac GSH concentrations and dramatically enhanced DOX-induced cardiotoxicity, as measured by the elevation of serum CK activities and the increase in cardiac MDA concentrations. The inhibitory effect of BSO on -GCS is fairly specific both in vitro and in vivo (Griffith, 1982), and GSH is an important endogenous antioxidant in detoxification of DOX (Dusre et al., 1989; Russo and Mitchell, 1985). Therefore, the BSO-enhanced DOX cardiotoxicity in nontransgenic mice was most likely mediated by depletion of GSH in the heart.

This BSO-enhanced cardiotoxicity by DOX was completely suppressed in the MT-overexpressing transgenic mouse heart. The GSH concentration in the transgenic mouse heart was depleted by BSO to the same extent as that in the nontransgenic mouse heart. Therefore, overexpression of MT in the heart did not prevent the depletion of cardiac GSH by BSO. The inhibition of the enhanced DOX cardiotoxicity thus did not result from any possible alterations in cardiac GSH depletion by BSO. The MT concentration in the transgenic mouse heart was about 20-fold higher than that in the nontransgenic mouse heart. BSO treatment did not affect MT concentrations in either transgenic or nontransgenic mouse hearts. It seems that in the presence of high concentrations of MT, the primary antioxidant defense against reactive oxygen species would shift from GSH to MT. This is possible because MT is much more efficient than GSH in protecting DNA from hydroxyl radical-induced damage (Abel and de Ruiter, 1989). MT also more preferentially reacts with hydrogen peroxide (Quesada et al., 1996), and hydroxyl radicals and superoxide (Thornalley and Vasak, 1985). Therefore, MT can compensate for the loss of GSH in cardiac protection against DOX toxicity. The results obtained from this study clearly demonstrate this possibility.

Drug resistance in tumor cells is an important limiting factor for the clinical application of DOX. Because GSH is involved in the acquired drug resistance in many tumor cells, preclinical effort has been placed on depletion of cellular GSH concentrations in tumor cells to reverse their resistance to DOX and other antitumor drugs (Dusre et al., 1989; Friedman et al., 1990; Malayeri et al., 1996). To this end, BSO has been studied experimentally to deplete GSH concentrations in tumors and has been on trials for the development of its clinical application (Bailey et al., 1994; O'Dwyer et al., 1992). However, the BSO enhancement of DOX cardiotoxicity as demonstrated in this study should be an important side effect to be concerned in the application of BSO in combination with DOX chemotherapy.

It has been shown that MT is also involved in anticancer drug resistance in tumor cells (Kelley et al., 1988). Interestingly, in tumor-bearing mice, bismuth subnitrate treatment significantly increased the concentrations of cardiac MT and protected the heart from DOX toxicity, but did not affect the antitumor activity of DOX (Naganuma et al., 1988). This implicates that selectively increasing cardioprotection by MT induction and decreasing tumor resistance to DOX by depletion of GSH would be an alternative approach to improved chemotherapeutic efficacy of DOX. MT has been shown to be elevated in the heart by a variety of inducers (Iszard et al., 1995; Satoh et al., 1988). Development of pharmaceutical agents to selectively increase cardiac MT concentrations is therefore possible.

What is the least level to which cardiac MT has to be elevated for the protection to occur? A recent study (Kondo et al., 1995) has shown that MT-null cells were not significantly sensitive to DOX toxicity. Another study using a transgenic mouse model in which MT was elevated in multiple organs including the heart (3-fold) has shown that the mouse heart was not protected from DOX toxicity (DiSilvestro et al., 1996). In our recent studies (Kang et al., 1997), we have found 10-fold elevation of cardiac MT provided a significant protection against DOX toxicity. It seems that under a threshold level, MT may not be an important factor in affecting DOX toxicity. Therefore, it should not make a difference in cellular sensitivity to DOX between MT-null cells and wild-type cells with normally low MT concentrations.

However, controversy also exists. In the bismuth study (Naganuma et al., 1988), the elevated MT level in the heart was transgenic mice was comparable to that in the transgenic mice (DiSilvestro et al., 1996). Yet, the bismuth-treated heart was significantly protected from DOX toxicity. Several explanations can be proposed to the discrepancy between these two studies. In the latter study, the toxicity of DOX was first assessed by mortality using a normally lethal dose with no significant difference observed between the transgenic and nontransgenic mice. This endpoint, however, does not reflect the acute cardiotoxicity, which has never been shown to be the cause of mortality. The same criticism can also be applied to the peritoneal fluid accumulation, which is not a specific endpoint of cardiotoxicity.

A specific measurement for cardiac oxidative injury by DOX was performed by examining the concentrations of 4-hydroxy-2-(E)-nonenal and malonaldehyde in the previous studies (DiSilvestro et al., 1996), which showed that MT transgenic mice actually displayed higher lipid peroxide concentrations in the heart treated with DOX. This study did not offer any explanation why the transgenic mouse heart showed higher concentrations of lipid peroxide products by DOX treatment. However, the colorimetric assays for 4-hydroxy-2-(E)-nonenal and malonaldehyde are nonspecific. An important problem in using these by-products as indicators of lipid peroxidation is that the by-product formation is highly inefficient and varies according to the transition metal ion content of the sample (Esterbauer et al., 1984). Because MT binds with metals and the composition of transition metal ions may be altered in the MT overexpressing transgenic heart, the measurement thus may be interfered by the presence of high concentrations of this protein.

The mechanisms by which MT functions in protection from reactive oxygen species induced tissue damage have been proposed (Vallee, 1995; Sato and Bremner, 1993; Ebadi et al., 1996; Quesada et al., 1996). Among these is that MT directly reacts with reactive oxygen species to prevent oxidative injuries such as lipid peroxidation (Thomas et al., 1986; Quesada et al., 1996). Lipid peroxidation is widely used as an important cellular event of cardiac oxidative injury by DOX. The major product of lipid peroxidation is MDA, and a commonly-used method to estimate tissue concentrations of MDA is the thiobarbituric acid test. However, this colorimetric method is notoriously nonspecific to MDA, because other known and unknown compounds produce positive reactions and the transition metals problem discussed above. It is questionable, therefore, whether the thiobarbituric acid value can be used as a reliable index for lipid peroxidation. In biological system, the cytotoxic aldehydes including MDA are extremely active. They can diffuse within or even escape from the site of the original free radical initiated event, and therefore act as "second cytotoxic messengers" (Esterbauer and Zollner, 1989; Loidl-Stahlhofen and Spiteller, 1994). Although it is challenging, measurement of free MDA before it is bound to biological macromolecules could be a meaningful effort to explore the mechanism involved in DOX cardiotoxicity. A modified HPLC method was therefore adopted in our study to separate the free MDA and the result clearly showed that free MDA level in the heart is significantly elevated by DOX treatment, although this level was not detectable in the heart without DOX treatment. Importantly, BSO pretreatment significantly further elevated the MDA concentration in the nontransgenic mouse heart, and this elevation was completely suppressed in the MT-overexpressing transgenic mouse heart. This result is therefore correlated with that from the measurement of serum CK activity. It is likely that inhibition of reactive oxygen species induced oxidative injury in the heart is an important mechanism by which MT protects the heart from DOX toxicity.