Chemically reactive intermediates and pulmonary xenobiotic toxicity.

After reading this lengthy review, the reader may consider it foolhardy to attempt to summarize the data contained herein. If so, the author and the reader are in complete accord. I set out to integrate, wherever possible, pathological changes in tissues and cells as determined by light, scanning, and TEM with chemical, biochemical, molecular, biological, and genetic techniques. I am not so naive as to equate correlation with causation, but in experimental biology, one is often compelled to rely on a number correlations to suggest causation. These data may also act as a stepping-off point helping the investigator design experiments to support/confirm or refute/disprove the hypothesis under investigation. Science is rarely black and white but rather gray, and the young investigator must be wary of scientists who argue their points too vociferously, too loudly, or too selectively. The truth in science is often a Gordian knot.

A possible role for membrane lipid peroxidation in anthracycline nephrotoxicity.

Adriamycin causes both glomerular and tubular lesions in kidney, which can be severe enough to progress to irreversible renal failure. This drug-caused nephrotoxicity may result from the metabolic reductive activation of Adriamycin to a semiquinone free radical intermediate by oxidoreductive enzymes such as NADPH-cytochrome P-450 reductase and NADH-dehydrogenase. The drug semiquinone, in turn, autoxidizes and efficiently generates highly reactive and toxic oxyradicals. We report here that the reductive activation of Adriamycin markedly enhanced both NADPH- and NADH-dependent kidney microsomal membrane lipid peroxidation, measured as malonaldehyde by the thiobarbituric acid method. Adriamycin-enhanced kidney microsomal lipid peroxidation was diminished by the inclusion of the oxyradical scavengers, superoxide dismutase and 1,3-dimethylurea, and by the chelating agents, EDTA and diethylenetriamine-pentaacetic acid (DETPAC), implicating an obligatory role for reactive oxygen species and metal ions in the peroxidation mechanism. Furthermore, the inclusion of exogenous ferric and ferrous iron salts more than doubled Adriamycin-stimulated peroxidation. Lipid peroxidation was prevented by the sulfhydryl-reacting agent, p-chloromercuribenzenesulfonic acid, by omitting NAD(P)H, or by heat-inactivating the kidney microsomes, indicating the requirement for active pyridine-nucleotide linked enzymes. Several analogs of Adriamycin as well as mitomycin C, drugs which are capable of oxidation-reduction cycling, greatly increased NADPH-dependent kidney microsomal peroxidation. Carminomycin and 4-demethoxydaunorubicin were noteworthy in this respect because they were three to four times as potent as Adriamycin. In isolated kidney mitochondria, Adriamycin promoted a 12-fold increase in NADH-supported (NADH-dehydrogenase-dependent) peroxidation. These observations clearly indicate that anthracyclines enhance oxyradical-mediated membrane lipid peroxidation in vitro, and suggest that peroxidation-caused damage to kidney endoplasmic reticulum and mitochondrial membranes in vivo could contribute to the development of anthracycline-caused nephrotoxicity.

In vitro studies on the metabolism and covalent binding of [14C]1,1-dichloroethylene by mouse liver, kidney and lung.

The metabolism and covalent binding of 1,1-dichloro[1,2-14C]ethylene (DCE) to subcellular fractions of liver, kidney and lung of C57BL/6N mice have been investigated in vitro. Covalent binding was NADPH- and cytochrome P-450-dependent. The microsomal fraction bound more radiolabel than any other subcellular fraction, and the levels of covalent binding in cell fractions correlated well with their cytochrome P-450 content. Covalent binding by mouse liver and lung microsomes also reflected their cytochrome P-450 content. However, although mouse kidney microsomes contained twice as much total cytochrome P-450 as the lung, no detectable covalent binding of DCE-derived radioactivity occurred in kidney. Omission of NADPH, heat inactivation of microsomes, carbon monoxide, addition of SKF-525A, piperonyl butoxide or reduced glutathione (GSH), all inhibited (40-90%) covalent binding of radiolabel to liver and lung microsomes. The absence of O2 (incubation under N2) did not greatly affect the metabolism and covalent binding. Pretreatment of mice with various inducers, phenobarbital (PB), beta-naphthoflavone (beta-NF), pregnenolone 16 alpha-carbonitrile (PCN) and 3-methylcholanthrene (3-MC), evoked increases in total liver microsomal cytochrome P-450 content (2-fold) and corresponding increases in covalent binding (3-fold). However, microsomes from PCN-treated mice showed only a 50% increase in DCE binding. Kidney microsomes from control, PB-, and beta-NF-pretreated mice were incapable of covalent binding of radiolabel but those from PCN- and 3-MC-pretreated mice showed levels of binding similar to untreated mouse lung microsomes. It is proposed that the nephrotoxicity of DCE may be due to translocation of reactive metabolites from the liver to the kidney.

Protection by methylprednisolone against butylated hydroxytoluene-induced pulmonary damage and impairment of microsomal monooxygenase activities in the mouse: lack of effect on fibrosis.

The effects of the synthetic corticosteroid methylprednisolone (MP; 30 mg/kg, s.c. given twice daily for 3 days), on the pneumotoxic effects of a single dose of butylated hydroxytoluene (BHT; 400 mg/kg, i.p.) over a 10 day experimental period was investigated in male C57BL/6N mice. BHT alone caused time-dependent alveolar hypercellularity, inflammatory infiltration, alveolar septal thickening and hypercellularity of the bronchiolar epithelium, reaching a maximum by day 5 with some degree of recovery by day 10. The pulmonary monooxygenase activities reflected the degree of alveolar damage and Clara cell abnormality with time; reductions in monooxygenase activities occurred and minimum levels (7-15% of control) were reached by day 5 and again a trend towards recovery by day 10. MP administered 0, 24 and 48 hr after BHT treatment partially protected mice from these effects of BHT in a distinctly time-dependent fashion; the degree of protection decreased as the time between BHT challenge and MP treatment increased. Although MP alone did not morphologically affect Clara and alveolar cells, it increased, decreased or had no effect on the monooxygenase activities. About 25% of the mice that received BHT alone died by day 5 and 50% by day 10. MP completely blocked the lethal effects of BHT by day 5 and reduced the deaths to between 15% and 25% by day 10. Interestingly, MP did not protect against the BHT-induced pulmonary fibrosis, measured as total lung hydroxyproline content, irrespective of the time between BHT challenge and MP treatment. MP alone did not cause any deaths nor increase lung hydroxyproline content.

Studies on the distribution and covalent binding of 1,1-dichloroethylene in the mouse. Effect of various pretreatments on covalent binding in vivo.

The distribution and covalent binding of a single dose of [1,2-14C] 1,1-dichloroethylene (DCE; 125 mg/kg, i.p.) was studied in male C57Bl/6N mice. Total radioactivity was distributed in whole homogenates of all tissues studied, with peak levels occurring within 6 hr. Covalent binding of radioactive material peaked at 6-12 hr in all tissues, and highest levels were found in kidney, liver, and lung with smaller amounts in skeletal muscle, heart, spleen, and gut. Covalent binding in kidney, liver, and lung fell to 50% of peak levels in about 4 days. Between 12 hr and 4 days after DCE administration, 70-100% of total radioactivity present in homogenates of kidney, liver, and lung was covalently bound. The three tissues showed a similar spread in total radioactivity in subcellular fractions 24 hr after exposure to DCE; most of the radioactivity was covalently bound (60-100%) and distributed fairly uniformly with a slight tendency to concentrate in the mitochondrial fraction. Phenobarbital (PB) and 3-methylcholanthrene (3-MC) pretreatments increased the covalent binding in the liver and lung but had no effect in the kidney. Piperonyl butoxide and SKF-525A decreased the covalent binding in liver and lung, but the latter increased binding in the kidney while the former decreased it. Diethylmaleate administration increased the covalent binding (2- to 3-fold) in all three tissues as well as increasing lethal toxicity. These results are consistent with the view that DCE is metabolized to some reactive intermediate(s) which may be detoxified by conjugation with glutathione.

The utility of changes in cardiac weight as an index of drug-induced cardiotoxicity.

The effects of toxic doses of various drugs and of food or water deprivation upon heart weights of mice were evaluated over a four day period to test the validity of the hypothesis that changes in cardiac weights are indicators of cardiotoxicity. Drugs included in the study were actinomycin-D, methotrexate, 5-fluorouracil, adriamycin, daunomycin, N-dimethyladriamycin, N-trifluoroacetyladriamycin-14-valerate, isoproterenol, atropine, and acetylsalicylic acid. Additional groups of mice served as vehicle controls, or were deprived of food or water for the duration of the experiment to control for the anorexia and dehydration accompanying treatment with antineoplastic drugs. Body weights were taken at the start of the experiment (day 0), day 2, and day 4 (just prior to sacrifice). Heart ventricle wet weights were determined immediately, and dry weights after thorough desiccation of the samples. Statistical evaluation of the weights revealed that there were no ventricular weight changes unique to any particular drug, and that decreases in heart weights correlated well with decreases in body weights, thereby reflecting the general toxicities of the drugs, including inanition, and not any specific cardiotoxicities.

Enhancement of reactive oxygen-dependent mitochondrial membrane lipid peroxidation by the anticancer drug adriamycin.

Mitochondrial degeneration is a consistently prominent morphological alteration associated with adriamycin toxicity which may be the consequence of adriamycin-enhanced peroxidative damage to unsaturated mitochondrial membrane lipids. Using isolated rat liver mitochondria as an in vitro model system to study the effects of the anticancer drug adriamycin on lipid peroxidation, we found that NADH-dependent mitochondrial peroxidation--measured by the 2-thiobarbituric acid method--was stimulated by adriamycin as much as 4-fold. Marker enzyme analysis indicated that the mitochondria were substantially free of contaminating microsomes (less than 5%). Lipid peroxidation in mitochondria incubated in KCl-Tris-HCl buffer (pH 7.4) under an oxygen atmosphere was optimal at 1-2 mg of mitochondrial protein/ml and with NADH at 2.5 mM. Malonaldehyde production was linear with time to beyond 60 min, and the maximum enhancement of peroxidation was observed with adriamycin at 50-100 microM. Interestingly, in contrast to its stimulatory effect on NADH-supported mitochondrial peroxidation, adriamycin markedly diminished ascorbate-promoted lipid peroxidation in mitochondria. Superoxide dismutase, catalase, 1,3-dimethylurea, reduced glutathione, alpha-tocopherol and EDTA added to incubation mixtures inhibited endogenous and adriamycin-augmented NADH-dependent peroxidation of mitochondrial lipids, indicating that multiple species of reactive oxygen (superoxide anion radical, hydrogen peroxide and hydroxyl radical) and possibly trace amounts of endogenous ferric iron participated in the peroxidation reactions. In submitochondrial particles freed of endogenous defenses against oxyradicals, lipid peroxidation was increased 7-fold by adriamycin. These observations suggest that some of the effects of adriamycin on mitochondrial morphology and biochemical function may be mediated by adriamycin-enhanced reactive oxygen-dependent mitochondrial lipid peroxidation.

Selective induction of renal microsomal cytochrome P-450-linked monooxygenases by 1,1-dichloroethylene in mice.

Intraperitoneal administration of a single dose of 1,1-dichloroethylene (DCE) to C57 B1/6N mice (125 mg/kg) caused a selective 6- to 10-fold increase in renal microsomal 7-ethoxyresorufin O-deethylase ( EROD ) and 7-ethoxycoumarin O-deethylase ( ECOD ), without affecting benzo[a]pyrene hydroxylase activity (AHH) or total microsomal cytochrome P-450 content. The observed increases did not result from in vitro activation of the enzymes or from any analytical artifact. Moreover, studies with actinomycin D and cycloheximide demonstrated that the increases resulted from de novo enzyme synthesis. Maximal enzyme induction was observed after a DCE dose of approximately 125 mg/kg, and the induced enzyme decayed rapidly, returning to control levels in about 3 days. Compared to female mice, male mice had higher basal levels of renal EROD and ECOD and were more responsive to the inductive effects of DCE; this correlated with corresponding differences in microsomal cytochrome P-450 levels. Starvation of mice for 24 or 48 hr increased renal EROD and ECOD activities in both male and female mice, but not the extent observed after DCE. The present results support the view of multiple renal cytochrome P-450 isozymes.

Selective damage to nonciliated bronchiolar epithelial cells in relation to impairment of pulmonary monooxygenase activities by 1,1-dichloroethylene in mice.

A single ip dose of 1,1-dichloroethylene (DCE) to mice (125 mg/kg) caused a reduction within 24 hr in cytochrome P-450 and related monooxygenases in lung microsomes, with no corresponding changes in liver and kidney microsomes. Light microscopy revealed that at 24 hr, DCE caused a highly selective and complete loss of the bronchiolar nonciliated (Clara) cells at all levels of the tracheobronchial tree. Electron microscopy showed that at this time, the bronchiolar luminal surface was covered by flattened, elongated ciliated cells. Within 24 hr total microsomal cytochrome P-450 and NADPH cytochrome c reductase were maximally reduced to about 50% of control and cytochrome P-450-dependent enzyme activities decreased to about 60% of control. By contrast, coumarin 7-hydroxylase was reduced to approximately 10% of control within 4 days. Since pulmonary coumarin 7-hydroxylase has been shown to reside almost exclusively in the Clara cells, this finding is in agreement with the observed extensive necrosis of the Clara cells. The return of lung microsomal P-450-linked enzyme activities took between 3 and 6 weeks and was paralleled by a corresponding slow reappearance of the bronchiolar Clara cells.

Interaction of the oncolytic drug, 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea with the mixed-function oxidase system in rats.

Effects of 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU) on the hepatic mixed-function oxidase system in male rats were studied both in vivo and in vitro. A single dose of CCNU (40 mg/kg) caused a significant reduction in hepatic mixed-function oxidase activities within 3 days after administration. The depression was prolonged for cytochrome P-450, total haem and the metabolism of several type I substrates lasting up to 10 weeks after a single dose. By contrast, aniline hydroxylase, cytochrome b5 and NADPH-cytochrome c reductase activities returned to near control levels after week two. Microsomal enzymes in the kidneys of treated animals however, were unaltered. Serum glutamic pyruvic and glutamic oxaloacetic transaminase and bilirubin levels, indicators of hepatotoxicity, were greatly elevated 3 days after CCNU treatment. These parameters fell rapidly but were still above control levels to the end of the 10-week study. When added in vitro, CCNU reduced apparent cytochrome P-450 content and the metabolism of type I substrates in microsomes from untreated, phenobarbital (PB) and 3-methylcholanthrene (3-MC)-pretreated rats. Total haem and NADPH-cytochrome c reductase were not affected whereas aniline hydroxylase activity was activated. CCNU interacted with hepatic microsomes to produce a type I difference spectrum.

A comparison of the pulmonary toxicity and chemotherapeutic activity of bleomycin-BAPP to bleomycin and pepleomycin.

The pulmonary toxicity and antitumor activity of a new bleomycin analog butylamino-3-propylamino-3-propylamine (Blm-BAPP) was investigated and compared with bleomycin and pepleomycin. Blm-BAPP was significantly more pulmonary toxic than bleomycin and had no greater activity against B16 melanoma than either bleomycin or pepleomycin. Although pepleomycin was as equitoxic as bleomycin in producing pulmonary fibrosis, doses of pepleomycin greater than 5 mg/kg were more lethal than bleomycin. Not only did the three drugs function similarly in vivo, but they behaved similarly in two in vitro test systems: microsome-catalyzed drug-mediated DNA deoxyribose cleavage and binding to DNA.

Stimulation of mouse heart and liver microsomal lipid peroxidation by anthracycline anticancer drugs: characterization and effects of reactive oxygen scavengers.

The interaction of adriamycin or several other quinone-containing anthracycline anticancer drugs with mouse heart or liver microsomes resulted in greatly enhanced NADPH oxidation and 5- to 10-fold stimulation of NADPH-dependent, reactive oxygen-mediated microsomal lipid peroxidation. Adriamycin, daunorubicin, desacetyladriamycin and aclacinomycin A all stimulated lipid peroxidation maximally when included at concentrations of 100 to 200 microM, whereas carminomycin and 4-demethoxydaunorubicin produced equivalent enhancement of peroxidation at lower concentrations of 30 to 50 microM. Mitomycin C markedly increased heart microsomal lipid peroxidation, but, interestingly, had little effect when liver microsomes were used. In contrast, 5-imino-daunorubicin and alkylaminoanthracenedione inhibited both heart and liver microsomal lipid peroxidation. NADPH supported anthracycline-stimulated lipid peroxidation, however, an NADPH-generating system provided greater activity. NADH was only about 50% as effective as NADPH and served as cofactor with liver microsomes but not with heart microsomes. Scavengers of reactive oxygen such as superoxide dismutase, reduced glutathione and 1,3-dimethylurea diminished the anthracycline-stimulated heart and liver microsomal lipid peroxidation, indicating that superoxide radical and hydroxyl radical participated in the peroxidation reactions. Ethylenediaminetetraacetic acid was also inhibitory, which suggests trace amounts of iron were also required. Adriamycin, 4-demethoxydaunorubicin, carminomycin, aclacinomycin A and mitomycin C strikingly enhanced peroxidation of unsaturated lipids in mouse lung and kidney microsomes, indicating that anthracycline-mediated enhanced reactive oxyradical generation may have toxic consequences in those organs as well as in the heart. These observations support the proposal that anthracycline-accentuated membrane lipid peroxidation may be relevant to the pathogenesis of anthracycline cardiotoxicity.

Enhancement of rat heart microsomal lipid peroxidation following doxorubicin treatment in vivo.

Cardiac microsomes from rats injected with doxorubicin (5 mg/kg ip x 3) were susceptible to dramatically enhanced NADPH-dependent lipid peroxidation in the presence of doxorubicin in vitro (18-fold increases over control peroxidation). These results suggest that doxorubicin treatment in vivo caused biochemical alterations which predisposed cardiac microsomal membranes to attack by anthracycline free radical-generated oxyradicals.

Bleomycin-metal interaction: ferrous iron-initiated chemiluminescence.

Chemiluminescence often accompanies the spontaneous degradation of intermediates in an electronically excited state. The interaction of iron with bleomycin results in the activation of bleomycin to a reactive intermediate which can alter DNA or undergo self-inactivation. This report demonstrates that the interaction of ferrous iron with bleomycin results in chemiluminescence, that this response is iron-specific and that the presence of DNA prevents the generation of chemiluminescence. These observations suggest that the activated bleomycin intermediate may be in an electronically excited state.

Lung-selective impairment of cytochrome P-450-dependent monooxygenases and cellular injury by 1,1-dichloroethylene in mice.

The acute toxic effects of 1,1-dichloroethylene (DCE; 125 mg/kg, i.p.) on mouse lung, liver and kidney were investigated 24 hr after its administration. DCE caused a reduction of cytochrome P-450 levels and related monooxygenases in lung microsomes with no corresponding changes in liver and kidney. Examination of the lung tissue by light microscopy revealed necrosis restricted to the Clara cells. In contrast, liver and kidney were relatively unaffected by DCE treatment, as indicated by lack of changes in microsomal monooxygenase activities and morphology.