George F. Kalf

Recent Advances in the Metabolism and Toxicity of Benzene

Benzene is a heavily used industrial chemical, a petroleum byproduct, an additive in unleaded gas, and a ubiquitous environmental pollutant. Benzene is also a genotoxin, hematotoxin, and carcinogen. Chronic exposure causes aplastic anemia in humans and animals and is associated with increased incidence of leukemia in humans and lymphomas and certain solid tumors in rodents. Bioactivation of benzene is required for toxicity. In the liver, the major site of benzene metabolism, benzene is converted by a cytochrome P-450-mediated pathway to phenol, the major metabolite, and the secondary metabolites, hydroquinone and catechol. The target organ of benzene toxicity, the hematopoietically active bone marrow, metabolizes benzene to a very limited extent. Phenol is metabolized in the marrow cells by a peroxidase-mediated pathway to hydroquinone and catechol, and ultimately to quinones, the putative toxic metabolites. Benzene and its metabolites appear to be nonmutagenic, but they cause myeloclastogenic effects such as micronuclei, chromosome aberrations, and sister chromatid exchange. It is unknown whether these genomic changes, or the ability of the quinone metabolites to form adducts with DNA, are involved in benzene carcinogenicity. Benzene, through its active metabolites, appears to exert its hematological effects on the bone marrow stromal microenvironment by preventing stromal cells from supporting hemopoiesis of the various progenitor cells. Recent advances in our understanding of the mechanisms by which benzene exerts its genotoxic, hematotoxic, and carcinogenic effects are detailed in this review.

A Perspective on Benzene Leukemogenesis

Although benzene is best known as a compound that causes bone marrow depression leading to aplastic anemia in animals and humans, it also induces acute myelogenous leukemia in humans. The epidemiological evidence for leukemogenesis in humans is contrasted with the results of animal bioassays. This review focuses on several of the problems that face those investigators attempting to unravel the mechanism of benzene-induced leukemogenesis. Benzene metabolism is reviewed with the aim of suggesting metabolites that may play a role in the etiology of the disease. The data relating to the formation of DNA adducts and their potential significance are analyzed. The clastogenic activity of benzene is discussed both in terms of biomarkers of exposure and as a potential indication of leukemogenesis. In addition to chromosome aberrations, sister chromatid exchange, and micronucleus formation, the significance of chromosomal translocations is discussed. The mutagenic activity of benzene metabolites is reviewed and benzene is placed in perspective as a leukemogen with other carcinogens and the lack of leukemogenic activity by compounds of related structure is noted. Finally, a pathway from exposure to benzene to eventual leukemia is discussed in terms of biochemical mechanisms, the role of cytokines and related factors, latency, and expression of leukemia.

Covalent binding of benzene and its metabolites to DNA in rabbit bone marrow mitochondria in vitro

Rabbit bone marrow mitochondria, stripped of their outer membrane (mitoplasts), have been shown to carry out the NADPH-dependent bioactivation of radiolabelled benzene in vitro to metabolites capable of covalently binding to mtDNA, thereby inhibiting transcription. The metabolites of benzene produced in bone marrow cells by the microsomal cytochrome P-450 are thought to be phenol, catechol, hydroquinone and p-benzoquinone (Andrews et al., Life Sci., 25 (1979) 567; Irons et al., Chem.-Biol. Interact., 30 (1980) 241). Incubation of mitoplasts from rabbit bone marrow cells in vitro with varying concentrations of the putative microsomal metabolites showed a concentration-dependent inhibition of RNA synthesis. The 50% inhibitory molar concentration (IC50) for each metabolite was determined to be: 1,2,4- benzenetriol , 6.3 X 10(-7); p-benzo-quinone, 2 X 10(-6); phenol, 2.5 X 10(-5); hydroquinone, 5 X 10(-5); catechol, 2 X 10(-3); benzene, 1.6 X 10(-2). DNA, isolated from rabbit bone marrow cell or rat liver mitoplasts prelabelled in DNA with [3H]dGTP and exposed to [14C]benzene in vitro, was enzymatically hydrolyzed to nucleosides which were chromatographed on a Sephadex LH-20 column to separate free nucleosides from nucleoside-adducts. The elution profiles indicated that rat liver mtDNA contained six guanine nucleoside-adducts and rabbit bone marrow cell mtDNA contained seven guanine nucleoside-adducts. Incubation of bone marrow mitoplasts in vitro in the presence of benzene and the hydroxyl radical scavenger, mannitol, resulted in the inhibition of formation of four of the guanosine-adducts. When [3H]dATP was substituted as the prelabelled precursor nucleotide, the LH-20 column profile indicated that two adenine nucleoside-adducts were also formed from benzene in vitro. Furthermore, a comparison of the Sephadex LH-20 column profiles of purine adducts derived from [14C]benzene- and [3H]dGMP-labelled mtDNA with profiles generated by individually incubating each of the putative unlabelled metabolites with bone marrow mitoplasts in vitro has indicated that p-benzoquinone, phenol, hydroquinone and 1,2,4- benzenetriol form adducts with guanine. One of the two adenosine-adducts may arise from hydroquinone; the compound forming the other adduct is unknown at the present time. Exposure of mitoplasts to catechol in vitro resulted in the formation of a guanine nucleoside-adduct that was present in rat liver mtDNA but absent from the DNA isolated from rabbit bone marrow cell mitoplasts exposed to [14C]benzene in vitro. This suggests that catechol is probably not a major metabolite of benzene formed in bone marrow cell mitochondria.

Inhibition of human DNA topoisomerase II by hydroquinone and p-benzoquinone, reactive metabolites of benzene

Chronic exposure of humans to benzene (BZ) causes acute myeloid leukemia (AML). Both BZ and therapy-related secondary AML are characterized by chromosomal translocations that may occur by inappropriate recombinational events. DNA topoisomerase II (topo II) is an essential sulfhydryl (SH)-dependent endonuclease required for replication, recombination, chromosome segregation, and chromosome structure. Topo II cleaves DNA at purine(R)/pyrimidine(Y) repeat sequences that have been shown to be highly recombinogenic in vivo. Certain antineoplastic drugs stabilize topo II-DNA cleavage complexes at RY repeat sequences, which leads to translocations of the type observed in leukemia. Hydroquinone (HQ) is metabolized to p-benzoquinone (BQ) in a peroxidase-mediated reaction in myeloid progenitor cells. BQ interacts wit SH groups of SH-dependent enzymes. Consequently, the aims of this research were to determine whether HQ and BQ are topo II inhibitors. The ability of the compounds to inhibit the activity of topo III was tested using an assay system that depends on the conversion, by homogeneous human topo II, of catenated kinetoplast DNA into open and/or nicked open circular DNA that can be separated from the catenated DNA by electrophoresis in a 1% agarose-ethidium bromide gel. We provide preliminary data that indicate that both HQ and BQ cause a time and concentration (microM)-dependent inhibition of topo II activity. These compounds, which potentially can form adducts with DNA, have no effect on the migration of the supercoiled and open circular forms in the electrophoretic gradient, and BQ-adducted KDNA can be decatenated by topo II. Using a pRYG plasmid DNA with a single RY repeat as a cleavage site, it was determined that BQ does not stimulate the production of linear DNA indicative of an inhibition of topo II religation of strand breaks by stabilization of the covalent topo III-DNA cleavage complex. Rather, BQ most probably inhibits the SH-dependent topo II by binding to an essential SH group. The inhibition of topo II by BQ has implications for the formation of deleterious translocations that may be involved in BZ-induced initiation of leukemogenesis. ImagesFigure 1.Figure 2.Figure 3.Figure 4.Figure 5.

Hydroquinone, a Bioreactive Metabolite of Benzene, Inhibits Apoptosis in Myeloblasts

Hydroquinone (a major marrow metabolite of the leukemogen, benzene) induces incomplete granulocytic differentiation of mouse myeloblasts to the myelocyte stage, and also causes an increase in the number of myelocytes. This was confirmed using the normal interleukin 3 (IL‐3)‐dependent mouse myeloblastic 32D cell line. The hydroquinone‐induced twofold increase in the number of IL‐3‐treated myelocytes does not result from stimulation of IL‐3‐induced proliferation. Hydroquinones ability to effect this increase through an inhibition of apoptosis was investigated using mouse 32D and human HL‐60 myeloblasts. Apoptosis induced by staurosporine treatment (0.5‐1.0 μM) of HL‐60 cells (50%) and 32D cells (15%) or by IL‐3 withdrawal from 32D myeloblasts was determined by monitoring the development of characteristic morphological features and confirmed by the appearance of a typical nucleosomal DNA ladder upon agarose gel electrophoresis. Concentration of hydroquinone (1‐6 μM) that induce differentiation in 32D myeloblasts caused a concentration‐dependent inhibition of staurosporine‐induced apoptosis in both cell lines, with a 50% inhibitory concentration of 3 μM, and prevented apoptosis in IL‐3‐deprived 32D cells. Hydroquinone inhibition of apoptosis in myeloblasts, like hydroquinone‐induced granulocytic differentiation, required myeloperoxidase‐mediated oxidation of hydroquinone to its reactive species, p‐benzoquinone, and was inhibited 50% by the peroxidase inhibitor, indomethacin (20 μM). p‐benzoquinone (3 μM) was shown to cause a 50% inhibition of CPP32, an IL‐1β converting enzyme/Ced‐3 cysteine protease involved in the implementation of apoptosis and present in myeloid cells. The ability of hydroquinone to induce a program of differentiation in the myeloblast that proceeds only to the myelocyte stage coupled with its ability to inhibit the CPP32 protease and, thereby, apoptosis of the proliferating myelocytes, may have important implications for benzene‐induced acute myeloid leukemia.

Inhibition of RNA synthesis and interleukin-2 production in lymphocytes in vitro by benzene and its metabolites, hydroquinone and p-benzoquinone

The effects of benzene and its metabolites, hydroquinone and p-benzoquinone (PBQ) on RNA synthesis in mouse spleen lymphocytes in vitro were studied. Benzene and the quinones were shown to inhibit RNA synthesis in a dose-dependent manner at concentrations which had no significant effect on lymphocyte viability. Furthermore, 5 microM PBQ, the putative toxic metabolite of benzene, was shown to inhibit the formation of the T-cell growth factor IL-2. These results suggest that inhibition of RNA synthesis in lymphocytes by benzene may prevent the production of factors required for hemopoiesis and thus contribute to the aplastic anemia caused by benzene.

Metabolic activation of hydroquinone by macrophage peroxidase.

Lysates from macrophages, cells involved in hematopoiesis and immunological responses, catalyzed the metabolic activation of the benzene metabolite, hydroquinone, to protein-binding compounds and to free 1,4-benzoquinone. This reaction is mediated by a peroxidase since activation was dependent upon H2O2 and was prevented by the inhibitors aminotriazole and azide. Activation of hydroquinone was independent of HO. radicals since protein binding occurred in the presence of the HO. scavengers mannitol and dimethyl sulfoxide. In reactions with macrophage lysates, phenol, another hepatic metabolite of benzene, stimulated the production of 1,4-benzoquinone as well as the amount of hydroquinone equivalents bound to protein in a dose-dependent manner. Addition of cysteine to incubations with macrophage lysates resulted in a dose-dependent decrease in hydroquinone equivalents bound to protein. At 100 microM cysteine, protein binding was inhibited by 63% and this decrease was recovered as the monocysteine-hydroquinone conjugate. Macrophages catalyzed the arachidonic acid-mediated activation of hydroquinone to metabolites which bound to cellular macromolecules. This activation was inhibited by indomethacin indicating the action of prostaglandin synthase in hydroquinone metabolism by macrophages. The results of these experiments demonstrate that macrophage peroxidase catalyzes the metabolic oxidation of hydroquinone to 1,4-benzoquinone and that 1,4-benzoquinone and/or its semiquinone intermediate are binding to protein and cysteine. Hydroquinone activation by macrophages and subsequent macromolecular binding may be associated with the immunologic and hematopoietic toxicity of benzene.

DNA replication by a membrane-DNA complex from rat liver mitochondria.

The results presented here indicate that mitochondrial DNA (mtDNA) synthesis occurs on the inner mitochondrial membrane and that a membrane-DNA complex, enriched in newly synthesized DNA, can be isolated. The complex is able to synthesize DNA in vitro. Enrichment studies demonstrated that mtDNA synthesis occurs on an intact membrane-DNA complex in vitro and that pulse-labeled mtDNA could be chased from the membrane-DNA complex to the top fraction of the discontinuous sucrose gradient. The membrane-DNA complex was also shown to carry out replicative synthesis of mtDNA in vitro. Replication was shown to be asynchronous with heavy-strand synthesis preceding light-strand synthesis. The progression of mtDNA replication by the membrane-DNA complex was shown to be from small fragments (<13 S) to larger fragments (14–24 S) liberated from closed circular molecules, to a heat-stable 27 S molecule, and finally to a 38 S heat-stable molecule. The time estimated to progress from small fragments to the 38 S molecule is 120 min.

The incorporation of leucine-1-C14 into the protein of rat heart sarcosomes: An investigation of optimal conditions

The optimal conditions for the incorporation, of leucine-1-C14 into the protein of rat sarcosomes have been investigated. The incorporation is dependent upon ATP generated within the sarcosome by oxidative phosphorylation. The antibiotics puromycin and chloramphenicol suppress this incorporating ability almost completely, while RNase, DNase, and the soluble fraction of the cell have little or no effect on the process. Evidence has been presented that the capacity to incorporate labeled leucine is a property of the mitochondria per se and is not the result of contamination by either microsomes or bacterial growth. Treatment of the radioactive protein with 1-fluoro-2,4-dinitrobenzene indicates that the leucine-1-C14 most probably is incorporated into the interior of the peptide chain.

The inhibition of mitochondrial DNA replication in vitro by the metabolites of benzene, hydroquinone and p-benzoquinone*

Rat liver mitochondria incubated with the metabolites of benzene, p-benzoquinone or 1,2,4-benzenetriol, showed a dose-dependent inhibition of [3H]dTTP incorporation into mtDNA with median inhibitory concentrations of 1 mM for each compound. Benzene and the metabolites phenol, catechol and hydroquinone did not inhibit at concentrations up to 10 mM. Similarly, incubation of p-benzoquinone or hydroquinone with rabbit bone marrow mitochondria showed a dose-dependent inhibition of mtDNA synthesis with 50% inhibition at 1 mM and 10 mM, respectively. That these metabolites inhibit mitochondrial replication was evidenced by the fact that [3H]dTTP incorporation into characteristic 38S, 27S and 7S mitochondrial replication intermediates was decreased by the quinones, as analyzed on 5-20% neutral sucrose velocity gradients. p-Benzoquinone, hydroquinone and 1,2,4-benzenetriol inhibited the activity of partially purified rat liver mtDNA polymerase gamma using either activated calf thymus DNA or poly(rA) X p(dT)12-18 as primer/template, with 50% inhibitory concentrations of 25 microM, 25 microM and 180 microM, respectively. Preincubation of the metabolites with polymerase gamma or primer/template, followed by removal of the unreacted metabolite by gel filtration, indicated that inhibition resulted from interaction of the metabolites with the enzyme, rather than with the template. Binding appeared to involve a sulfhydryl residue on the enzyme since the binding of [14C]hydroquinone was prevented by N-ethylmaleimide. The ability of hydroquinone or p-benzoquinone to inhibit binding of [14C]hydroquinone to the enzyme suggests that the compounds bind to a common site or are converted to a common intermediate. Inhibition of, or changes in, replication in mitochondria of bone marrow cells by hydroquinone and p-benzoquinone may explain the changes in the mitochondrial genome observed in marrow stem cells in acute myelogenous leukemia and may suggest a mechanism for benzene leukemogenesis.