Benzene metabolism in rodents at doses relevant to human exposure from urban air.

Benzene is both an environmental pollutant and a component of cigarette smoke, gasoline, and automotive emissions. Although occupational exposure to benzene has been shown to cause blood disorders and cancer in humans, the potential health effects resulting from exposure to low levels of benzene are not known. The goals of this project were to determine how well benzene is metabolized and to assess its binding to macromolecules in rodents at doses more closely mimicking human environmental exposure. To determine whether genotoxic metabolites of benzene are produced at environmental exposure levels. various doses of 14C-benzene were given intraperitoneally to male B6C3F1 mice at doses from 5 ng/kg to 500 mg/kg body weight. Samples of urine, plasma, liver, and bone marrow were taken at selected times up to 48 hours after exposure. Individual benzene metabolites in the samples were measured by accelerator mass spectrometry (AMS*). Metabolites were quantified by determining the area under the curve (AUC) for 24 to 48 hours. The major metabolites found in urine were an unidentified radiolabeled metabolite. phenyl sulfate, phenyl glucuronide. and muconic acid (an indicator of muconaldehyde formation). The major metabolites found in plasma, liver, and bone marrow samples were muconic acid and hydroquinone. Only liver showed a dose response for hydroquinone and muconic acid. The kinetics of both DNA and protein adduct formation were assessed over 48 hours at a 14C-benzene dose of 5 microg/kg body weight. A dose-response study was then conducted using 14C-benzene doses from 5 ng/kg to 500 mg/kg body weight in B6C3F1 mice. Adduct levels were determined by AMS in liver and bone marrow. DNA and protein adducts in liver reached maximum levels 30 minutes after benzene administration, whereas those in bone marrow reached maximum levels after six hours. Both protein and DNA adduct AUCs were greater in bone marrow than in liver. Dose-response assessments at both 1 and 12 hours showed that DNA and protein adducts in liver and bone marrow were dose dependent over doses spanning eight orders of magnitude. Consistent with the benzene metabolism data, these data show that reactive forms of benzene were present in liver and bone marrow after exposure to human-relevant benzene levels. Thus, at low doses, benzene was absorbed and metabolized into reactive intermediates capable of binding to DNA and protein. The relation between benzene metabolism and macromolecular binding was examined by comparing benzene macromolecular adduct formation among strains of male mice (B6C3F1, DBA/2, and C57BL/6) and male rats. These animals have been reported to have different metabolic capacities for benzene and also different tumorigenic and cytotoxic responses to benzene exposure. We hypothesized that differences in the capacity to metabolize benzene affect macromolecular adduct formation and that the amount of macromolecular damage is related to benzene's ability to cause cancer and other blood disorders. 14C-benzene was administered intraperitoneally to all rodents (5 microg/kg body weight) and adduct levels were determined by AMS at selected time points up to 48 hours after exposure. AUCs for protein and DNA adducts in bone marrow, the primary target organ for benzene toxicity, decreased in the following order: B6C3F1, DBA/2, C57BL/6, and rats. Similarly, adduct AUCs in liver were greater in B6C3F1 mice than in rats although the trend was less clear for the DBA/2 and C57BL/6 mouse strains. The results of this work are consistent with previously published work showing that the ability to metabolize benzene follows a similar pattern with these animals and is consistent with the tumorigenicity of benzene in mice and rats. Thus, our data suggest that benzene toxicity is related to the ability to produce macromolecular adducts. Preliminary studies were conducted to assess adduct dosimetry after low-dose inhalation of benzene. Inhalation methods were developed to administer a nominal body burden of 5 microg benzene/kg body weight. Then, 14C-benzene was administered to B6C3F1 mice and rats by both intraperitoneal (IP) administration and by inhalation, and DNA and protein adducts in liver and bone marrow were analyzed by AMS. AUCs for adduct levels were greater after IP benzene administration than after inhalation of benzene. Adduct levels were greater in DNA from B6C3F1 mouse bone marrow than in DNA from liver regardless of exposure route. Collectively, these data show that the internally reactive dose was greater when benzene exposure was by IP administration. In summary, the results suggest that benzene is metabolized to reactive forms capable of binding both protein and DNA in target and nontarget organs of rats and mice at doses encountered by humans through environmental exposure. Macromolecular binding was dose-dependent at low doses of benzene and reflected benzene toxicity, based on its carcinogenicity and ability to cause other blood-related disorders. These data are consistent with macromolecular adducts being indicative of benzene exposure and benzene toxicity although much more research is needed to validate this point. Additionally, benzene metabolism varies among species and among strains within a species of rodent. Thus data are needed in humans to understand how to use the rodent data in risk assessment and ultimately to determine whether macromolecular adducts are a useful indicator of exposure and a useful predictor of risk.