Air Pollution, the Automobile, and Public Health (1988)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Health Effects of Alclehydes d Al h 1 M b 1 Source Emissions LAWRENCE J. MARNETT Wayne State University Ambient Levels and Production by Mobile Sources / 580 Health Effects / 582 Aldehydes / 582 Methanol / 587 Phenols and Catechols / 587 Cocarcinogenic Effects of Aldehydes, Alcohols, and Phenols / 588 Metabolism / 589 Methanol and Formaldehyde / 589 Acrolein / 590 Measurement of Inspired Methanol and Formaldehyde / 590 Chemical Reactions / 591 Quantification of Exposure and Estimation of Human Risk / 595 Summary / 597 Summary of Research Recommendations / 598 Exposure / 598 Health Effects / 598 Cellular Effects / 599 Molecular Dosimetry / 599 Air Pollution, the Automobile, and Public Health. (a) 1988 by the Health Effects Institute. National Academy Press, Washington, D.C. 579

580 Aldehydes and Alcohols in Mobile Source Emissions Aldehydes are oxidation products of alco- hols; phenols contain an alcohol function- ality attached to an aromatic ring. Al- though they are structurally related, the chemistry and toxicology of the three classes of compounds are different. From a toxicologic standpoint, aldehydes have been more extensively investigated than alcohols and phenols and constitute the most important health hazard. Thus, most of the emphasis of this chapter is placed on them, but the literature on exposure, health effects, metabolism, and chemistry of al- dehydes, alcohols, and phenols are also examined. Recommendations are made for research necessary to fill important gaps in our knowledge of these subjects. The liter- ature review highlights key experiments but is not comprehensive; recent review articles and monographs are cited and can be con- sulted for more complete information. Ambient Levels and Production by Mobile Sources Estimates of the atmospheric levels of some common pollutants present in mobile source emissions and the series of com- pounds that are discussed in this chapter are presented in table 1. Formaldehyde and acetaldehyde have been monitored more extensively than acrolein, alcohols, and phenols. The data for the latter three com- pounds represent episodic reports rather than averages of extensive compilations. Nevertheless, they are useful for assigning an order of magnitude to the levels likely to be found in urban air. Formaldehyde is usually the most abundant of the com- pounds of interest and acetaldehyde is next most abundant. Their concentrations are between 100 and 1,000 times lower than carbon monoxide (CO), the principal pol- lutant in auto exhaust. Although ambient formaldehyde levels range from 4 to 86 parts per billion (ppb), occasional levels in excess of 1,000 ppb 1 part per million (ppm) have been reported (Goldsmith and Friberg 1977~. Acrolein is usually de- tected at levels below 10 ppb. Most of the aldeh.ydes in urban air are present as gases. Estimates suggest the percentage of al- dehydes bound to particles is approxi- mately 1 percent of the concentration in the gas phase (Grosjean 1982~. Acrolein is also produced In fires, but by far the highest concentrations are present in cigarette smoke (12 ppm) (Treitman et al. 1980; Carson et al. 1981~. Acetaldehyde is also present in very high concentrations in cig- arette smoke (1,65~2,500 ppm) (Elmen- horst and Schultz 1968~; the formaldehyde  concentration in cigarette smoke is approx- imately equal to that of acrolein (Newsome et al. 1965~. Alcohol concentrations in the atmo- sphere have only been determined in a cursory fashion. In a recent study in urban and rural sampling stations in Arizona, Snider and Dawson (1985) reported values of 7.9 ppb for methanol in Tucson and 2.6 Table 1. Atmospheric Levels of Compounds Present in Mobile Source Emissions Compound Level % from Mobile Source Reference Carbon monoxide 5-13 ppm 70 Graedel, this volume Nitrogen oxides 20-30 ppb 50 Graedel, this volume Nonmethane hydrocarbons 1-2 ppmC 37 Graedel, this volume Sulfur dioxide 9 ppb <5 Graedel, this volume Formaldehyde 086 ppba 55-75 Grosjean (1982) Acetaldehyde 2-39 ppba 55-75 Hoshika (1977) Grosjean (1982) Acrolein 2-7 ppb NA Beauchamp et al. (1985) Alcohols 8-100 ppb NA Snider and Dawson (1985) Phenols NA NA a Los Angeles. NOTE: NA = not available.

Lawrence I. Marnett 581 Table 2. Concentrations of Formaldehyde and Acetaldehyde in Exhaust Diluted Gases of Internal Combustion Engines Fueled by Gasoline or Ethanol" Gasoline-Fueledb Ethanol-Fueled Formaldehyde Acetaldehyde Formaldehyde Acetaldehyde Cold starts 540 <8 67019,800 Hot starts 97 <8 100550 40-mph cruise 48 <8 360840 a Concentrations are given in ppb, and dilution factor is 10:1. b Equipped with a catalyst. c Federal Test Procedure driving cycle. SOURCE: Based on data from Swarin and Lipari 1983. ppb at a location 50 km from Tucson. Atmospheric sampling was by a condensa- tion procedure. The concentrations of eth- anol at the two stations were reported to be 3.3 and 0.4 ppb, respectively. The origin of the alcohols was not clear. The authors were unable to detect propanol, butanol, or acrolein. Phenol, as well as o- and m-cresol, have been reported in undiluted auto ex- haust at levels of 1.4, 0.2, and 0.3 ppm, respectively (Kuwata et al. 1981~. The ambient levels of aldehydes reported to exist in Los Angeles appear to be higher than levels in other cities in the United States and Japan. Therefore, upper limits can be approximated by studies performed in the Los Angeles metropolitan area. Data gathered at a reporting station in Clare- mont (~50 km east of Los Angeles) indi- cate that diurnal patterns exist for formal- dehyde and acetaldehyde levels (Grosjean 1982~. There is a close correlation between variations in the aldehyde levels and the diurnal fluctuations of ozone (O3) as well as the movement of smog banks. Other re- ports establish a correlation of fluctuations in aldehyde levels to diurnal variations in CO levels (Cleveland et al. 19774. Diurnal and seasonal fluctuations of formaldehyde and acetaldehyde levels have been observed at a monitoring station on Long Island (Tanner and Meng 1984~. These data sug- gest that mobile sources contribute sign111- cantly to ambient concentrations. Esti- mates of the percentage contribution made by mobile emission sources to the levels of the various compounds are presented in table 1. Such estimates have not been made for alcohols and phenols. Analysis of the levels of aldehydes in exhaust gases provides evidence for a de- pendence on the type of engine and fuel (Swarin and Lipari 1983~. Table 2 lists the concentrations of formaldehyde and acetal- dehyde detected in a 10:1-diluted sample of exhaust gas from an internal combustion engine fueled by gasoline or ethanol. Dra- matic increases in acetaldehyde concentra- tions are detected in ethanol-fueled engine  exhaust, as might be expected since acetal- dehyde is the two-electron oxidation pro- duct of ethanol. Presumably a similar in- crease in formaldehyde concentrations would be observed in exhaust from engines fueled by methanol. Diesel engines produce significantly higher amounts of aldehydes than do gas- oline engines (table 3~. Ratios of acetal- dehyde to formaldehyde also increase if catalysts are placed in the exhaust stream (Swarin and Lipari 1983~. Taken together, these findings suggest that the character of automotive emissions varies dramatically with the type of engine and fuel. This Table 3. Concentrations of Selected Aldehydes in Diluted Diesel Exhausta Aldehyde Cold Startb Formaldehyde Acetaldehyde Propionaldehyde Acrolein Crotonaldehyde Benzaldehyde Hot Startb 428 80 24 24 6 NA 539 115 57 57 11 9 a Concentrations are given in ppb, and dilution factor is 10:1. b Federal Test Procedure driving cycle. NOTE: NA = not available. SOURCE: Based on data from Swarin and Lipari 1983.

582 Aldehydes and Alcohols in Mobile Source Emissions implies that a major shift to, for example, alcohol-containing fuels would have a sig- nificant effect on the concentration of cer- tain aidehydes and alcohols in urban air. The levels of alcohols in urban air are rather low, and it is difficult to imagine that they are high enough to exert any health effects. Levels of aldehydes are normally well below levels at which they induce hazardous effects, although they can occa- sionally reach high ambient concentrations (Grosjean 1982; Beauchamp et al. 1985~. Results of test burns indicate that dramati- cally increased levels of aldehydes are produced from alcohol-containing fuels (Swarin and Lipari 1983). These higher levels could be well within the range that induces adverse health effects. If methanol- and ethanol-based fuels are widely adapted, it will be important to be able to determine the levels of atmospheric aldehydes and to have a baseline value against which to compare them. Thus, routine monitoring should be initiated now. · Recommendation 1. Routine moni- toring of atmospheric alcohol and aldehyde levels should be performed in regions where alcohol-based fuels are or will be in heavy use. Reports of the detection of phenols and catechols from mobile source emissions are extremely limited and usually do not pro- vide quantitative information. As a result, knowledge of their levels in urban air and the contribution made by mobile sources is totally inadequate. No realistic risk assess- ment can be undertaken for this class of compounds without such information. ~ Recommendation 2. Methods should be developed to routinely analyze phenols and catechols in urban air. Health Effects Aldehydes The literature on the health effects of al- dehydes and alcohols is enormous, but several recent reviews by Consensus Work Table 4. Acute Effects of Acrolein on Human Volunteers Concentration Effects (ppb) 30-34 14~150 300-500 800-900  1,200 Odor threshold for most acrolein sensitive people. Some eye irritation in 2 mini in- creased annoyance and almost no eye or nose irritation during re peated exposures. Slight eye and nose irritation; no ef- fect on respiratory frequency or amplitude; odor perceived. Changes in amplitude of respiratory movements; slightly increased res- piratory frequency; decreased eye sensitivity to light. Extremely irritating to all mucous membranes in 5 mini lacrimation. 1,000-23,000 Medium to severe eye irritation in 5 man. SOURCE: Based on data from Beauchamp et al. 1985. shop on Formaldehyde (1984), Beauchamp et al. (1985), and Tephly (1985) are partic- ularly appropriate. The present discussion is restricted primarily to inhalation toxicol- ogy. Acute Elects in Humans. Acute irritant effects of aldehydes on human volunteers have been documented. In general, acrolein is the most potent of the series acrolein, formaldehyde, acetaldehyde, crotonalde- hyde. For example, acrolein is approxi- mately two to three times more potent than formaldehyde as an irritant (Beauchamp et al. 1985). Table 4 is a compilation of the doses at which various short-term re- sponses to acrolein have occurred in human volunteers (Carson et al. 1981). Ocular and olfactory irritation is the first detectable response and occurs at doses that are 1 (}20 times higher than the ambient levels in urban air (table 1). Extreme irritation to mucous membranes and alteration in respi- ration occur at doses approximately 100 times ambient. At such levels, it is likely that irreversible epithelial damage occurs on chronic exposure (see Acute Effects in Rodents, below). A similar profile of ef- fects is observed for formaldehyde in hu

Lawrence]. Marnett 583 mans at somewhat higher doses. Irritation occurs at 0.1-3.0 ppm, and respiratory difficulties are evident at 10-20 ppm (Fas- sett 1963~. Acetaldehyde and crotonal- dehyde are 10-100 times less active than acrolein and formaldehyde. Allergic responses to aldehydes have been reported. Hendrick and Lane (1975) documented a case of asthma induced by exposure of a hospital staff member to formalin vapor. A pronounced decrease in respiratory performance was observed after exposure to a 25 percent solution for 15 mini however, the ambient levels of form- aldehyde present were not measured. Symptoms were prevented by pretreat- ment of the patient with betamethasone. Skin allergies have been induced by topical application of solutions of formaldehyde but the dose responses have not been ex- tensively determined (Maibach 1983~. Der- mal but not respiratory sensitivity has been observed in guinea pigs exposed to 10 ppm formaldehyde for 6 or 8 hr/day for 5 con- secutive days. (Lee et al. 1984~. Chronic Effects in Humans. Numerous groups of individuals are occupationally exposed to formaldehyde, acetaldehyde, acrolein, and crotonaldehyde. Epidemio- logic studies of the chronic effects of form- aldehyde have been conducted with several of these groups, but the results are incon- clusive. Despite the ability to identify ex- posed individuals, there is little information on their smoking and drinking habits, which confounds the interpretation of any detected alterations in disease incidence. The Consensus Workshop on Formalde- hyde (1984) evaluated several epidemio- logic studies of professional and industrial workers exposed to formaldehyde and con- cluded that there are insufficient data to establish whether or not it is a human carcinogen. The level of atmospheric expo- sure in those workers was approximately 0.1-1.0 ppm. In the same groups, there appeared to be no excess mortality associ- ated with formaldehyde exposure. Chronic exposure of humans to high levels of acro- lein is considered unlikely because of its extreme irritation. At levels below those that cause olfactory or respiratory damage (~1 ppm), prolonged exposure to acrolein is intolerable, causing individuals to leave the contaminated environment. Conse- quently, there is no information on human carcinogenicity or other chronic effects of acrolein, nor are data on potential human carcinogenicity of acetaldehyde or croton- aldehyde available. Acute and Chronic Effects in Rodents.  Pathological changes occur in the upper respiratory, especially nasal, epithelium of rodents exposed to aldehydes. The site and severity are dose dependent (Kutzman et al. 1985~. Acute effects have also been noted, and the lesions observed include exfolia- tion, ciliastasis, cell erosion, ulceration, ne- crosis, squamous metaplasia, and inflam- mation (Dalhamn and Rosengren 1971; Buckley et al. 1984~. Most of the damage is reversible but some is irreversible. Signifi- cantly, these effects are detected when ro- dents are exposed to the RD50s for formal- dehyde and acrolein (Buckley et al. 1984~. (The RD50 is defined as the level at which a 50 percent reduction in respiratory rate occurs. This level reflects the stimulation of sensory receptors that attempt to limit ex- posure to irritants.) The RD50s for formal- dehyde and acrolein are 3.1 and 1.7 ppm in mice (Buckley et al. 1984~; the RD50 for acrolein is 6.0 ppm in rats (Babiuk et al. 1985~. Since pathological changes occur in experimental animals as a result of expo- sure to the RD50 levels, proposals have been made that the RD50 be used to esti- mate "safe" exposure levels for humans (possibly 0.01-0.1 x RED (Kane et al. 1979; Alarie 1981~. Whether damage occurs in response to exposure to these much lower levels is unknown. a, ~ Recommendation 3. Chronic low- dose inhalation toxicology studies should be undertaken to determine if tissue dam- age occurs in response to exposure to levels of formaldehyde, acetaldehyde, and acro- lein that are 10-100 times lower than their RDsos. Although pure formaldehyde and acro- lein do not cause neutrophil recruitment, an inflammatory response has been observed

584 Aldehydes and Alcohols in Mobile Source Emissions in response to carbon particles coated with either compound (Kilburn and McKenzie 1978~. Paradoxically, exposure of suspen- sions of neutrophils to formaldehyde and acrolein results in lowered responsiveness to soluble stimuli, such as phorbol esters, and decreased generation of superoxide an- ion (Witz et al. 1985~. This may be respon- sible for the decreased in vivo killing of bacteria by mice treated with either com- pound Jakab 1977~. Carcinogenicity. Exposure of 232 Fischer 344 rats to 14.3 ppm formaldehyde (6 hr/day, 5 days/week for 24 months fol- lowed by 6 months of nonexposure) in- duced squamous cell carcinoma in the nasal epithelium of 103 animals (Kerns et al. 1983~. Exposure to 5.6 ppm formaldehyde induced squamous cell carcinoma in only 2 of 235 animals, and at 2.0 ppm no response was observed in 236 animals. In mice (B6C3F~), exposure to 14.3 ppm induced nasal tumors in only 2 of 215 animals (Kerns et al. 1983~. This figure did not represent a statistically significant increase but is notable because of the rarity of nasal tumors in mice. Exposure of Syrian golden hamsters to 10 ppm formaldehyde (5 hr./ day, 5 days/week for 120 weeks) did not induce any airway tumors (Dalbey 1982~. Mixtures of formaldehyde and hydrochlo- ric acid induced nasal cancer in Sprague- Dawley rats that was entirely due to the formaldehyde; no enhancing effect of hy- drochloric acid was seen (Albert et al. 1982; Sellakumar et al. 1985~. In all of these chronic exposure studies, clearcut evidence was acquired for reversible as well as irre- versible damage to respiratory epithelium. Chronic exposure of groups of Wistar rats (110 animals/group) to acetaldehyde at initial doses of 750, 1,500, or 3,000 ppm (6 hr/day, 5 days/week for 27 months) in- duced 14, 34, and 38 nasal tumors, respec- tively, compared to 1 in controls (Wou- tersen et al. 1984~. Severe irreversible degenerative changes of the upper respira- tory tract were observed in the high-dose group so the acetaldehyde concentration had to be reduced repeatedly throughout the course of the experiment. The tumors observed at the low and moderate doses of acetaldehyde occurred in the olfactory epi- thelium (Woutersen et al. 1984), whereas nasal tumors induced by low levels of formaldehyde occur in the respiratory epi- thelium (Kerns et al. 1983~. Nasal tumors induced by high-level exposure to acetal- dehyde and formaldehyde occur in the ol- factory and respiratory epithelium. The results at low levels suggest that acetal- dehyde is better able to penetrate to remote  anatomic locations than formaldehyde. Further evidence for differential effects of aldehydes is provided by the observation that acetaldehyde at levels of 1,65~2,500 ppm (7 hr/day, 5 days/week for 52 weeks) induces tracheal, but not nasal, tumors in Syrian golden hamsters (Feron et al. 1982~. Exposure of Syrian golden hamsters to 4 ppm acrolein (7 hr/day, 5 days/week for 52 weeks) induced a number of pathological changes in the upper respiratory tract, par- ticularly the nasal epithelium, but no tu- mors were observed in any organs (Feron and Kruysse 1977~. Acrolein exhibits po- tent teratogenic and embryolethal effects when it is administered to rats intraamni- otically but not by inhalation (Slots and Hales 1985~. Neither acrolein nor formaldehyde was carcinogenic in Syrian golden hamsters (Feron et al. 1982~. Acrolein is similar to formaldehyde in chemical reactivity, irri- tant activity, and retention in the respira- tory tract, so it should be tested in the same species in which formaldehyde has been detected as a carcinogen the rat (Kerns et al. 1983~. Attention should be paid to the development of nasal tumors. ~ Recommendation 4. A chronic inha- lation toxicology study of acrolein should be undertaken in rats, with emphasis on . . . carclnogenlclty. The species and organ specificities of different aldehydes with respect to their ability to induce respiratory tumors on inhalation exposure is fascinating and has been discussed (Kerns et al. 1983; Swen- berg et al. 1983~. Stimulation of nasal re- ceptors may play a key role in the difference in the higher sensitivity of rats relative to mice. Rodents attempt to restrict their in

Lawrence I. Marnett 585 take of irritants by reducing their respira- tory minute volume (Chang et al. 1981~. This response is more pronounced in mice than in rats so, for example, at the same level of exposure to formaldehyde, rats breathe approximately twice as much formaldehyde as mice (Chang et al. 1981~. Indeed, the tumorigenic response of mice to the effects of 14.3 ppm formaldehyde is roughly the same as the response of rats to 6 ppm formaldehyde (Kerns et al. 1983~. The importance of effective dose on tissue specificity is further emphasized by the observation that no tumors have been de- tected outside of the respiratory tract with any aldehyde. Studies of the retention (that is, the amount of compound bound to tissue) of various aldehydes in the respira- tory tract of dogs indicate that formalde- hyde is completely retained and acrolein is nearly completely retained in the upper tract whereas propionaldehyde is much less retained (Egle 1972b). Acetaldehyde is the least retained of all the aldehydes tested in the upper respiratory tract, which is con- sistent with its ability to induce tumors in hamster trachea (Egle 1972a). Decreases in minute volume cannot ex- plain the sharpness of the dose response of rats to formaldehyde. However, Swenberg and colleagues (1983) proposed that effects on mucociliary activity may play a role. The nasal respiratory epithelium is nor- mally covered by a dynamic protective layer of mucus. The carbohydrate and pro- tein in this layer may react with molecules such as formaldehyde, preventing their ac- cess to epithelial tissue. Interruption of mucous flow might saturate the capacity of these macromolecules to react with form- aldehyde over certain anatomic locations, thereby increasing the effective dose. Formaldehyde increases mucous flow at low exposure levels but reduces it at high levels (Swenberg et al. 1983~; this may result from the ciliastatic activity exhibited by formaldehyde (Morgan 1983; Morgan et al. 1983~. Inhibition of mucociliary clear- ance introduces an additional step in the carcinogenic process, suggesting that short exposures to high concentrations would be more effective for compound delivery than long exposures to low doses. This is consist ent with the nonlinear dose response ob- served for formaldehyde carcinogenicity in rats. It also suggests that occasional high levels of exposure might exert biological effects not expected from extrapolation of dose responses obtained by chronic low- level exposure (Swenberg et al. 1983~. Fur- ther support for nonlinearity of formalde- hyde action on respiratory epithelium is provided by the dose dependence for in-  duction of squamous metaplasia in the nasal cavity of Fischer 344 rats and B6C3F~ mice (Kerns et al. 1983~. Formaldehyde at 2 ppm only induces squamous metaplasia in the anterior-most regions of the nasal cavity in rats. Extensive metaplasia in midlevel and posterior portions of the cavity are ob- served with 5.6 and 14.3 ppm, respectively. By comparing the extent of squamous meta- plasia in mice and rats, it is possible to approximate doses that exert similar patho- logical effects. Using this criterion, the extent of penetration by formaldehyde appears equivalent in rats and mice at doses of 5.6 and 14.3 ppm, respectively. This result correlates well to the difference in sensitivity of the two species to the carcinogenic action of formal- dehyde in the respiratory tract. Studies indicate that alterations of mu- cous flow and ciliatoxicity are important components of the nonlinear dose response for the carcinogenic action of formaldehyde (Swenberg et al. 1983~. Acrolein exhibits the most potent ciliatoxic activity of any volatile aldehydes (Beauchamp et al. 1985~. This may enhance the carcinogenic re- sponse to other less ciliatoxic aldehydes such as formaldehyde or acetaldehyde. The most likely combination to test first is acrolein and formaldehyde because they are the most potent ciliatoxins and carcino- gens, respectively, in mobile source emis- s~ons. ~ Recommendation 5. A chronic inha- lation toxicology study of mixtures of formaldehyde and acrolein should be un- dertaken in rats and hamsters, with empha- s~s on carc~nogen~c~ty. How one extrapolates the results of car- cinogenicity studies in rodents to human exposure is uncertain. Humans are rou

586 Aldehydes and Alcohols in Mobile Source Emissions finely exposed to atmospheric levels of formaldehyde that are 10~1, 000 times lower than the doses that induce nasal tumors in rodents. However, individuals in certain cities are intermittently exposed to much higher levels. Whether lon~-term damage results from these episodic expo- sures is uncertain, although there is little doubt that acute effects, such as irritation, occur. High intermittent exposure might serve as an initiating event that provides a focus of transformed cells sensitive to pro- motion by other pollutants or environmen- tal agents. Epidemiologic data do not pro- vide evidence for a significant contribution of air pollution to human cancer but one might suggest that the combination of ex- posure to aldehydes in automotive emis- sions with other environmental agents is important in some individuals such as smokers (Doll and Peto 1981~. This seems reasonable enough, but the concentrations of aldehydes in cigarette smoke are several orders of magnitude higher than their con- centrations in urban air. Therefore, the significance of the contribution of al- dehydes in mobile source emissions to health effects in smokers is uncertain. The other complication of extrapolating results from rodent bioassays to humans is the difference in anatomy and physiology of the two species. Rodents are obligate nose breathers whereas humans are not. This has obvious implications for the amounts of toxic agents that reach respiratory tissues by inhalation. Cultured Cells. Formaldehyde and other aldehydes exert numerous effects on iso- lated cells in culture. They are toxic to normal as well as tumor cells and, in fact, certain cr,,~unsaturated aldehydes were used in human clinical trials as potential chemotherapeutic agents (Schauenstein et al. 1977; Krokan et al. 1985~. The genotox- ic effects of formaldehyde have long been recognized (Auerbach et al. 1977~. It in- duces single-strand breaks, DNA-protein cross-links, sister chromatic exchanges, and chromosome aberrations (Ross and Shipley 1980; Bedford and Fox 1981; Ross et al. 1981; Fornace 1982; Fornace et al. 1982; Levy et al. 1983~. It is mutagenic in a variety of prokaryotic and eukaryotic cells including human fibroblasts (Chanet and van Borstel 1979; Boreiko et al. 1982; Goldmacher and Thilly 1983; Szabad et al. 1983), transforms rodent cells (Ragan and Boreiko 1981), and enhances viral transfor- mation of Syrian hamster embryo cells (Hatch et al. 1983~. Formaldehyde-induced DNA lesions appear to be repaired, but formaldehyde itself inhibits the ability of  human bronchial epithelial cells and fibro- blasts to repair damage by x rays and methylating agents (GraEstrom et al. 1983, 1984~. A similar constellation of events occurs in response to treatment of cells with acrolein (Schauenstein et al. 1977; Beauchamp et al. 1985~. Despite the extensive documentation of the cellular effects of formaldehyde and other aldehydes, the understanding of their actions at the molecular level is incomplete. For example, the critical targets that lead to various cellular pathologies are, for the most part, unknown. Evidence exists link- ing the toxicity of cr,,~unsaturated al- dehydes to modification of a critical sulf- hydryl protein, ., , , ~. but its identity is unspec~ea t~chauenstein et al. 1977~. Cer- tain DNA polymerases contain important sulfhydryl groups that are sensitive to modification, so these are likely candidates (Kornberg 1980~. Sulfhydryl reactivity may also contribute to the inhibition of DNA repair by methyl transferases caused by formaldehyde (Krokan et al. 1985~. Recommendation 6. Experiments should be undertaken in cells cultured from various segments of the upper respiratory tract to determine the mechanisms by which aldehydes exert pathological changes such as toxicity, hyperplasia, ciliatoxicity, and so on. Such experiments should concentrate on identifying the critical targets for modifica- tion by each compound and the extent of modification that triggers the response. For example, despite the extensive literature on killing of prokaryotic and eukaryotic cells by a,,l3 unsaturated aldehydes, the precise mechanism of killing and the macromol- ecules involved are uncertain. Does modi

Lawrence.~. Marnett 587 fication of DNA polymerases lead to tox- icity or does toxicity result from inhibition of enzymes of ATP generation? At what level of modification does toxicity result? Such knowledge will be important for basic biology as well as for risk assessment based on molecular dosimetry (see below). A major unresolved question is how one extrapolates the results of experiments demonstrating pathological effects of al- dehydes on cultured cells to risk assessment for human exposure. For most of the in vitro experiments, aldehydes are added in solution, whereas in animal exposure ex- periments they are administered by inhala- tion. How one relates molar concentrations of liquids to dosages of a gas that may accumulate in a target cell is unknown. Methanol Methanol is rapidly absorbed following oral, cutaneous, or respiratory exposure and undergoes general distribution to body water (Yant and Schrenck 1937; Haggard and Greenberg 1939~. Its biological half-life is 1.5-2 hr (Sedivec et al. 1981), which means that many of the toxicologic effects triggered by inhalation exposure may be similar to those observed following oral administration. Methanol's oral toxicity to humans has been known for over 100 years (Tephly 1985~. Considerable variability is observed in the dose at which toxicity results but best estimates of a dose required for severe intoxication and death are around 1 g/kg (Roe 1982~. A lag phase of 12-24 hr is observed before any symptoms of toxicity are seen, which implies that a metabolite is involved in the toxicity (Tephly 1985~. Metabolic acidosis occurs followed by visual effects that can lead to blindness. Ocular toxicity is occasionally followed by coma, other central nervous system effects, and death. Ethanol antago- nizes the effects of methanol and it may be that varying amounts of ethanol contami- nation account for the variability in dose at which methanol is toxic to individuals (Roe 1955~. Rodents are not susceptible to the toxic effects of methanol but nonhuman primates are; for example, methanol exhib- its ocular toxicity in monkeys (Roe 1982) (see Metabolism, Methanol and Formalde- hyde). Exposure of human volunteers to an atmosphere containing 200 ppm methanol results in accumulation of 750 mg of which 50 60 percent is retained in the lung (Se- divec et al. 1981~. Considering the dose of methanol estimated to be toxic to humans (1 g/kg), it is unlikely that a normal human being could ever be exposed to enough of it by inhalation to experience acute toxicity. This author was unable to find carcino-  genicity studies of methanol by inhalation exposure. Methanol is metabolized slowly during systemic circulation to formalde- hyde, which is quickly metabolized to formic acid (Tephly 1985~. A remote possi- bility is that methanol is oxidized to form- aldehyde in the respiratory epithelium which is sensitive to its carcinogenic action. If so, methanol may act as a latent form of formaldehyde leading to accumulation in tissues that formaldehyde is ordinarily in- accessible to. Similar considerations hold for ethanol with respect to acetaldehyde. Taken with the potential importance of methanol and ethanol as alternate fuels, it seems important to test them thoroughly for carcinogenicity via the inhalation route. It is less important to test ethanol in inha- lation studies because its oxidation product, acetaldehyde, is 100 times less active as a carcinogen than formaldehyde. · Recommendation 7. A chronic inha- lation toxicology study of methanol should be undertaken in rats and hamsters, with . . . . . emphasis on carclnogenlclty. Phenols and Catechols No information is available on the inhala- tion toxicology of phenols or catechols. Most toxicologic studies have been per- formed by oral or intravenous administra- tion, so the concentrations used are difficult to relate to inhalation exposure. Phenols are not strongly toxic, and substituted phenols such as butylated hydroxy toluene and butylated hydroxy anisole are used as pre- servatives in food. There is some specula- tion that the presence of phenolic antioxi- dants in food accounts for the steady decrease in stomach cancer in developed

588 Aldehydes and Alcohols in Mobile Source Emissions countries since 1945 (Doll and Peto 1981~. Indeed, phenolic antioxidants such as buty- lated hydroxy anisole inhibit chemical car- cinogenesis and appear to act at the promo- tion stage (Slaga et al. 1983~. However, high doses of phenolic antioxidants actually appear to be tumor promoters themselves (Ito et al. 1982~. Cocarcinogenic Effects of Aldehydes, Alcohols, and Phenols Cocarcinogenic effects have been reported for formaldehyde and acetaldehyde (Dal- bey 1982; Feron et al. 1982~. Lifetime ex- posure of Syrian golden hamsters to 30 ppm formaldehyde concomitant with sub- cutaneous administration of 0.5 mg dieth- ylnitrosamine resulted in an enhancement of the number of tracheal tumors over treatment with diethylnitrosamine alone (Dalbey 1982~. As mentioned above, form- aldehyde does not induce tracheal tumors in hamsters. No enhancement of tumori- genesis was seen in the larynx or lung, and the effect on the trachea was only observed when formaldehyde exposure began before diethylnitrosamine administration. En- hancement did not result when formalde- hyde exposure began after diethylnitrosa- mine injections were completed. A similar experiment was performed in Syrian golden hamsters with acetaldehyde (1,65() 2,500 ppm) and benzota~pyrene adminis- tered by intratracheal instillation (Feron et al. 1982~. At a dose of 36.4 mg but not 18.2 mg benzota~pyrene, enhancement of tra- cheal and bronchial tumorigenesis was ob- served after 52 weeks. When a similar experiment was performed with injection of diethylnitrosamine no enhancement of tracheal tumorigenesis was observed. In fact, there appeared to be a decrease over controls but this was considered a casual association. Formaldehyde has been re- norted to exhibit "initiating" and "pro- moc~ng achy in the C3H/lOT1/2 in vitro transformation system (Ragan and Boreiko 1981; Frazelle et al. 1983~. Methanol has not been tested for cocar- cinogenicity by the inhalation route. An epidemiologic association has been estab- lished between consumption of alcoholic beverages and esophageal cancer in smok- ers, but there is no evidence for direct carcinogenicity of ethanol (Doll and Peto 1981~. Its role in enhancing the carcinogen- icity of cigarette smoke is uncertain. Cat- echol has been identified as the major co- carcinogenic component of cigarette smoke (Van Duuren and Goldschmidt 1976; Hecht et al. 1981~. However, bioassays were per- formed using the initiation-promotion model on mouse skin so the importance of  catechol as an inhalation cocarcinogen is uncerta~n. · Recommendation 8. Attempts should be made to develop an initiation-promo- tion protocol for carcinogenesis testing of aldehydes and other components of mobile . . source em1sslons. The two-stage mouse skin model has been very useful for detection of potential carcinogens, tumor initiators, and tumor promoters. There is no analogous model that can be used to screen compounds for their effects on respiratory tissues. When aldehydes were administered to rodents simultaneously or after administration of benzota~pyrene or diethylnitrosamine (Dal- bey 1982; Feron et al. 1982), some stimu- latory and inhibitory effects were noted but they were not dramatic, and it was diff~cult to speculate whether the aldehydes were acting as cocarcinogens or promoters based on the design of the experiments. A repro- ducible initiation-promotion model would enable rapid testing of mixtures of mobile source emission components by the inhala- tion route and would provide useful mech- anistic information. Considering that formaldehyde and acrolein exert most of their effects on the respiratory epithelium of the nasal tract of rats and that acetal- debyde is a nasal carcinogen in rats and a tracheal carcinogen in hamsters (Feron et al. 1982; Woutersen et al. 1984), efforts should be directed toward developing a model in which the biological effects are monitored in the upper respiratory tract. It appears that most of an inspired dose of these compounds does not reach the bron- chi and lungs, so the model should be designed with this in mind. In other words,

Lawrence J. Marnett 589 it would not seem prudent to perform developmental experiments using com- pounds that exert possible initiating or pro- moting effects in the lungs. Metabolism Although some adverse effects of aldehydes and alcohols have been described in hu- mans, experimental animals, and cell sys- tems, quantification of risk, especially at ambient concentrations, is difficult with the current data base. Additional research is necessary to better estimate the potential toxicity of these compounds. The design and interpretation of experiments will be aided by understanding their metabolism and chemical reactions. Methanol and Formaldehyde All of the alcohols and aldehydes consid- ered here are soluble in aqueous and or- ganic solutions, which means they distrib- ute rapidly throughout the body and within cells (Beauchamp et al. 1985~. The major pathway of metabolism is oxidative with alcohols oxidized to aldehydes and al- dehydes oxidized to acids. For example, methanol is oxidized to formaldehyde, which is oxidized to formic acid: OH O C ~ // \ H H H H H Methanol Formaldehyde o 11 11 C ~C / \ / H OH Formic Acid ( 1 ) Metabolism of alcohols and aldehydes can result in either detoxification or metabolic activation. The fact that a lag phase is observed before the onset of clinical symp- toms of methanol toxicity, coupled with the findings that ethanol and alcohol dehy- drogenase inhibitors antagonize methanol toxicity, suggests that a metabolite of methanol is responsible for its observed toxicologic effects. Alcohol dehydrogenase appears to be primarily responsible for the oxidation of methanol (McMartin et al. 1975~. Its binding constant for methanol is approximately six times lower than its binding constant for ethanol, which ac- counts for the ability of ethanol to antago- nize methanol's effects (Maker and Tephly  1975~. Catalase is important for methanol metabolism in rats but not in monkeys (Mannering and Parks 1957~. The major metabolite of methanol in monkeys is formic acid (eq. 1~. Formate also exhibits ocular toxicity in monkeys (Martin-Amat et al. 1978~. It accumulates in monkeys following methanol adminis- tration, thereby resulting in metabolic aci- dosis, but does not accumulate in rats. This is consistent with the differential sensitivity of these species to methanol toxicity and implies a role for formate as a toxic metab- olite. Methanol oxidation by alcohol dehy- drogenase is the rate-limiting step of me- tabolism and appears to be equally rapid in rats and monkeys (Watkins et al. 1970; Clay et al. 1975~. The accumulation of formate in monkeys relative to rats appears to be due to a decreased rate of its oxidation to carbon dioxide (CO2) in monkeys (Mc- Martin et al. 1977~. Formate metabolism occurs by a folic acid-dependent pathway, so folate deficiency renders monkeys ex- tremely sensitive to methanol toxicity (McMartin et al. 1977~. Conversely, folate supplementation lowers their sensitivity (Noker et al. 1980~. It appears that folic acid levels are rate-limiting for formate metab- olism to CO2 in monkeys. The toxic effects of methanol may be enhanced by simultaneous exposure to other compounds. For example, antifolates are used clinically for treatment of psoriasis and cancer, and it is conceivable that indi- viduals undergoing treatment could exhibit enhanced sensitivity to methanol. Acute methotrexate treatment of monkeys does not decrease their rate of formate oxida- tion, but the effects of chronic treatment are unknown (Noker et al. 1980~. Perhaps more important is the observation that nitrous oxide (N2O) is an inhibitor of an enzyme of folic acid metabolism and leads to folate depletion (Eells et al. 1982~. En- hanced sensitivity to methanol toxicity is observed following exposure of monkeys to N2O, and metabolic acidosis is induced in rats, a species normally resistant to

590 methanol effects (Eells et al. 1981~. This raises the possibility that nitrogen oxides (NOX) in auto exhaust might enhance any toxic effects of methanol and formaldehyde that are mediated by formic acid. The oxidation of methanol to formic acid most likely involves formaldehyde as an intermediate (eq. 1~. In contrast to the slow rate of its formation from methanol, form- aldehyde is oxidized quite rapidly. Its half- life is estimated to be 1 min following intravenous infusion (Reitbrock 1969; McMartin et al. 1979~. Formaldehyde does not accumulate following methanol ad- ministration but is rapidly metabolized to CO2 (85 percent) and expired (Neely 1964; Mashford and Jones 1982~. The initial oxidation of formaldehyde appears to be catalyzed by a formaldehyde dehydroge- nase that is glutathione-dependent (Stritt- matter and Ball 1955; Goodman and Tephly 1971~. The enzyme is quite specific for formaldehyde (Strittmatter and Ball 1955~. The remaining 15 percent of form- aldehyde that is not oxidized to CO2 may bind to protein or enter pathways of one- carbon metabolism. Formaldehyde does not appear to contribute significantly to methanol toxicity following oral adminis- trat~on. Acrolein The major pathway of acrolein metabolism appears to involve conjugation with gluta- thione followed by conversion to S-car- boxyethyl-mercapturic acid (Draminski et al. 1983~. to H~C=CH-C + GSH ~ GS~H~CH~ C H o Ac~ys CHIN CH, C Reaction of acrolein with glutathione is a rapid chemical reaction but is also catalyzed by glutathione transferases Jakoby and Ha- big 1980~. The half-life for conjugation of glutathione with 4-hydroxy-nonenal, a molecule structurally related to acrolein, is Aldehydes and Alcohols in Mobile Source Emissions approximately 4 sec in perfused rat heart (Ishikawa et al. 1986~. Acrolein is also oxidized to acrylic acid in vivo, but this only accounts for approximately 15 percent of the administered dose (Draminski et al. 1983~. o ~ [OX] H~C=CH-C ~ H~C=CH-CO'H  H (3) Oxidation of acrolein to acrylic acid by rat liver subcellular fractions is inhibited by diethyldithiocarbamate, an inhibitor of al- dehyde dehydrogenase (Patel et al. 1983~. Epoxidation of acrolein to glycidaldehyde occurs in rat liver subcellular fractions but it is not known if this transformation takes place in viva (Patel et al. 1983~. O 0 0 ~ /\ ~ H~C=CH-C ~ H~C~CH-C H H (4) This is a potentially important metabolite of acrolein because it has been classified as an animal carcinogen (International Agency for Research on Cancer 1976). It induced malignant tumors in rats following subcu- taneous injection and papillomas in mice following skin painting (Van Duuren et al. 1965, 1966, 1967). Its role in acrolein me- tabolism and potential carcinogenicity is uncertain. Measurement of Inspired Methanol and Formaldehyde O Methanol is produced as a result of normal human metabolism; it is detectable in hu man breath and urine. The best way to H assay for inhaled methanol is to quantitate increases in its urinary levels by gas chro matography. Approximately 1 percent of the inspired dose is excreted in human H (2) urine. Because there is a significant back ground level of methanol in human urine due to metabolism or diet, it has been estimated that the lower limit of exposure to methanol that could be detected by an increase in urinary levels would result from inhalation for 8 hr of air containing 100 ppm methanol (Heinrich and Angerer

Lawrence I. Marnett 591 1982~. On this basis, it is unlikely that one could ever detect increases in methanol inhalation resulting from exposure to mo- bile source emissions. Considering that formaldehyde is an intermediate in metha- nol and carbohydrate metabolism, one can extend this analysis to reach the conclusion that it is also impossible to detect inhala- tion exposure to formaldehyde by moni- toring urinary levels of it or its initial metabolite formic acid. This conclusion is strengthened by the realization that form- aldehyde does not escape the respiratory tissue to which it is administered and, therefore, never achieves appreciable sys- temic levels. Chemical Reactions Aldehydes. Adduct Formation. Aldehydes are reactive electrophiles that add reversibly to nucleophiles to form covalent hydroxy- methyl and imine adducts. OH (5) H.C=0 + Nu ~ H,C-Nu ~ H~C=Nu + H'O For simple aliphatic aldehydes such as formaldehyde or acetaldehyde, these ad- ducts are unstable and readily revert to starting materials (Fraenkel-Conrat 1954; Hoard 1960~. Thus, despite the fact that formaldehyde is mutagenic, reacts with nu- cleic acid, and induces major alterations in nucleic acid structure, hydroxymethyl or imine adducts to nucleic acid bases or nu- cleotides have only recently been isolated and identified (McGhee and von Hippel 1975; Beland et al. 1984~. Reaction of hy- droxymethyl derivatives with a second nu- cleotide base produces cross-linked pro- ducts that are stable to hydrolysis (Feld- man 1967~. OH HO-Nu + Nu H H \ / C N / \N Dimeric adducts have been isolated follow- ing the reaction of formaldehyde with de- oxyguanosine, deoxyadenosine, and de- oxycytidine, and mixtures thereof (figure 1) (Feldman 1967; Chaw et al. 1980~. Chemical reduction of imines generates amines, which are stable. H'C=NH + NaBH4 , H3C-NH-R \  R (7) Imine reduction probably does not occur in biological systems in vivo but is often used in in vitro experiments to trap unstable aldehyde/nucleophile addition products (Chio and Tappel 1969~. It may be useful as a derivatizing reaction for quantitation of aldehyde/protein or aldehyde/nucleic acid adducts that would otherwise decompose during tissue processing and sample prep aration. Addition of nucleophiles to the ,~carbon of cY,,B unsaturated aldehydes generates products that are considerably stabler than hydroxymethyl compounds or imines. o o H~C=CH-C + Nu ~ Nu H.C CH-C (8) H H The most reactive nucleophile in proteins toward c~,,(3 unsaturated aldehydes is the sulfhydryl group (Schauenstein et al. 1977~. Many proteins and enzymes contain sulf hydryl groups, and a detailed study of the toxicity of a,,l3 unsaturated aldehydes to virus particles, bacteria, and mammalian cells indicates a correlation of their reactiv ity toward sulfhydryl groups (Schauenstein et al. 1977~. The identity of the critical cellular protein inactivated by a,,~unsat urated aldehydes has not been established, but it is noteworthy that two DNA poly merases are reactive toward sulfhydryl re agents (Kornberg 1980~. Another candidate is ribonucleotide reductase, which catalyzes the rate-limiting step in cellular DNA syn thesis and contains an unusual dithiol group that is sensitive to c~,,~unsaturated al dehydes and sulDhydryl reagents (The lander and Reichard 1979~. Finally, it <6' should be mentioned that the trinentide glutathione contains a sulfhydryl group that reacts readily with ct,,l3 unsaturated aldehydes. In fact, an important physiolog ical role of glutathione is the scavenging of electrophiles such as c~,,B unsaturated al dehydes Jakoby and Habig 1980~. Reaction

592 Aldehydes and Alcohols in Mobile Source Emissions Formaldehyde o 0 N (Ni N X No: N R OH O ant R R Crotonaldehyde OH O ~N>N? R Acrolein o HO;N-1~) R Figure 1. Adducts formed by reaction of guanosine with formaldehyde, croton- aldehyde, and acrolein. The formaldehyde adduct is a dimer of two guanosines with a single molecule of formaldehyde. Analogous dimeric adducts are formed by reaction of formaldehyde with adenosine and cytidine. Unsymmetrical dimeric adducts are also formed when formaldehyde is added to mixtures of guanosine, adenosine, and cytosine. R = deoxyribose. of glutathione with these compounds oc- curs spontaneously or is enzyme-catalyzed. Although intracellular glutathione concen- trations are high (~6 mM), acute exposure to c~,,B unsaturated aldehydes can result in significant glutathione depletion, which lowers the cell's defenses toward electro- philic agents Jakoby and Habig 1980). This is known to potentiate toxicity of xenobi- otics in the short term, but the long-term effects of glutathione depletion are un- known. Adducts between c~,,l3 unsaturated al- dehydes and amine groups also form and are biologically very important (eq. 8~. Formation of amine-a,,~unsaturated al- dehyde adducts is reversible but sufficiently slow that the adducts can be isolated and identified. The structures of several a,,B unsaturated aldehyde/deoxynucleoside ad- ducts have been elucidated and are listed in figure 1. Adducts to deoxyguanosine have been detected following the reaction of acrolein or crotonaldehyde with DNA (Chung and Hecht 1983; Chung et al. 1984~. These adducts survive hydrolysis of  DNA to deoxynucleosides, which is a key step in the isolation of any DNA adducts. They may represent useful potential indi- cators of DNA damage in cells exposed to high concentrations of a,,~unsaturated al- dehydes. It is interesting that cyclic adducts result from the reaction of cr,,l3 unsaturated aldehydes with DNA bases. Cyclization of the aldehyde group of the initial adduct to a suitably disposed amine group of deoxy- guanosine is favored by entropy. O 0 1 N N N H R HOH ~ N JAN \ H R (9) A cyclic adduct also forms between deoxy- guanosine and glycidaldehyde, an in vitro metabolite of acrolein (Van Duuren and Loewengart 1977).

Lawrence i. Marnett Reaction of aldehydes with nucleic acids is believed to be responsible for the muta- genic and carcinogenic effects of the com- pounds. However, saturated and unsaturat- ed aldehydes also inhibit repair of certain adducts formed by methylating agents (Grafstrom et al. 1983, 1984~. Inhibition of DNA repair appears to be a result of the covalent reaction of aldehydes with the sulfhydryl group of the methyl acceptor protein that removes the methyl group from 0-6-methylguanine residues in DNA exposed to methylating agents. Inactiva- tion of the methyl acceptor protein en- hances the mutagenicity of methylating agents. Thus, aldehydes can enhance mu- tagenesis by a mechanism that does not involve modification of nucleic acid. Free-Radical Formation. Aldehydes un- dergo enzyme-catalyzed oxidation and re- duction in biological systems. The enzymes involved appear to be dehydrogenases, which implies mechanisms involving hy- dride transfer a relatively innocuous transformation. Aldehydes are also prone to autoxidation triggered by one-electron oxidation of the aldebydic carbon-hy- drogen bond (Lloyd 1973). o 11 c / \ o [OX] ~ R H R/ (l O) The resultant radical couples to O2 under aerobic conditions to form a peroxyl radical. 0 0 11 11 cool ~ c o R R O (ll) The peroxyl radical then carries out one- electron oxidation of another molecule of aldehyde. 11 c o + / \ / R O o 11 C O + / \ / \ R O H o 11 /c R (12) The latter reactions constitute the propaga-  tion steps of a free-radical chain autoxida 593 tion. Numerous molecules of aldehyde are oxidized and peroxyl radicals produced as a result of a single initiation event. Peroxyl radicals are relatively stable free radicals that are selective in their reaction with cellular constituents (Willson 1985). They have the ability to diffuse far from the site of their generation to react with specific molecules (Pryor 1984). Acyl peroxyl rad- icals, the type produced by aldehyde autox- idation, are significantly more reactive than alkyl peroxyl radicals and may not be able to diffuse as far intracellularly. The cellular targets for peroxyl radical reactions are unknown, but they epoxidize isolated dou- ble bonds and abstract hydrogen atoms from polyunsaturated fatty acid residues in phospholipids (Willson 1985). The latter reaction results in lipid peroxidation, which can disrupt membrane structure, lead to cell death, and release soluble mediators of toxicity and chemotaxis. Peroxyl radicals also oxidize sulfides to sulfoxides. ROO S ~ , / \ o 11 S / \ , ~, ~ R. R. R. R~ c}-1-Proteinase inhibitor contains a critical methionine residue close to the site that combines with a variety of proteinases to inhibit their action (Travis and Salvesen 1983). Oxidation of the methionine residue abolishes the ability of the protein to inhibit catalysis by proteolytic enzymes. This al- ters the balance of proteolysis and can have dramatic effects on lung function. The physiological target for a-1-proteinase in- hibitor appears to be elastase which de- grades pulmonary connective tissue by vir- tue of its action on elastin. Elastase is secreted by neutrophils in response to their activation during inflammation. The activ- ity of elastase is regulated in part by the amount of a-1-proteinase inhibitor that is available to combine with and inactivate it. Certain genetic diseases that are character- ized by pulmonary emphysema are associ- ated with decreased amounts of a-1-pro- teinase inhibitor (see Wright, this volume). The critical methionine residue of cY-1- proteinase inhibitor that controls its activ- ity toward proteinases is oxidized by per

594 Aldehydes and Alcohols in Mobile Source Emissions oxides and hypochlorous acid, both products of activated neutrophils. It is also oxidized by free radicals such as the one present in cigarette smoke (Pryor 1984~. The latter reaction may be especially im- portant in the genesis of diseases associated with chronic cigarette smoking such as emphysema. Although it has not been tested, it seems quite likely that acyl per- oxyl radicals formed by autoxidation of aldehydes inactivate cz-1-proteinase inhibi- tor by oxidizing its critical sulfide to a sulfoxide. This provides a mechanism by which one-electron oxidation of aldehydes could lead to pulmonary emphysema. Alcohols. Alcohols are relatively unreac- tive chemically with nucleophiles and elec- trophiles. They can be converted to more reactive derivatives by conjugation with functionalities (for example, glucuronate, sulfate) that render the hydroxyl groups more reactive Jakoby et al. 1980; Kasper and Henton 1980~. H3C-OH + ·OH , H~C OH + HERO (14) These conjugates could conceivably act as electrophiles and alkylate nucleophiles, but there is no evidence that the toxicity exhib- ited by, for example, methanol is a result of such reactions. Furthermore, short-chain alcohols are highly water-soluble, which removes much of the driving force for their conjugation with polar moieties. Alcohols are oxidized to aldehydes and acids by dehydrogenases, which is a reaction of pri- mary importance in the metabolism of al- cohols. Alcohols are not oxidized by one electron to free radicals very readily, al- though they will react with hydroxyl rad- ical to form hydroxylmethyl radicals. Phenols and Catechols. Attachment of a hydroxyl group to an aromatic ring greatly enhances the reactivity of the O H bond. By comparison to aliphatic alcohols, the chemistry of phenols is rich. Phenols are more acidic than alcohols and possess sig- nificant nucleophilicity toward reactive elec- trophiles. For this reason they can serve a protective role as scavengers of metaboli- cally generated electrophiles. They are also conjugated readily, which effects a marked change in their polarity. The most important reaction of phenols is probably with one- electron oxidants (Simic and Hunter 1983~. They readily donate electrons of hydrogen atoms, thereby generating phenoxyl radicals. OH H9C4_~h'C4Hg  R. + ROO O H9C4_~4H9 + ROOH R. (15) Phenoxyl radicals are much more stable than aliphatic alkoxyl radicals because of conjugation with the aromatic ring. Phe- nols with alkyl substitutents ortho to the hydroxyl group are widely used as chain- breaking antioxidants (Howard 1973). These compounds donate a hydrogen atom to per- oxyl radicals that are formed during the propagation step of autoxidation. The stabil- ity and steric hindrance of phenoxyI radicals prevent them from abstracting hydrogen at- oms from donors that will react with chain- carrying peroxyl radicals. Phenoxyl radicals couple to second molecules of peroxyl radical to form peroxycyclohexadienones. o HgC4_:C4H9 + ROO R1 R o HgC4_~ O R. -o ~C4Hg (16)

Lawrence I. Marnett 595 As a result, every phenol molecule removes two peroxyl radicals from autoxidation JO mixtures, which interrupts the radical chain ~ - , and inhibits the autoxidation process. ~ I' + Nu H Phenols that lack alkyl groups ortho to ~~< the hydroxyl group react with one-electron t' oxidants to form phenoxyl radicals that are quite reactive. An important reaction of phenoxyl radicals is coupling to 02, which forms peroxyl radicals. O b+o~ o to 11 rH \' (17) As discussed above, peroxyl radicals ab- stract hydrogen atoms from reactive mol- ecules such as unsaturated fatty acids to initiate and propagate radical-chain oxida- tions. This leads to the paradox that phenols, which are generally thought to be antioxi- dants, can actually stimulate free-radical au- toxidation. The reactions of phenoxyl radi- cals are especially important because phenols can be oxidized to phenoxyl radicals by high- valence metals as well as alkoxyl and peroxyl free radicals. This provides a mechanism for metal-catalyzed initiation of free-radical oxi- dations. Free radicals are believed to enhance carcinogenesis, particularly the promotion phase, and free-radical initiators are promot- ers in the two-stage assay in mouse skin (Slaga et al. 1981~. This may contribute to the reported carcinogenicity of certain phenols in mouse forestomach (Ito et al. 1982) and to the cocarcinogenicity of catechols on mouse skin (Van Duuren and Goldschmidt 1976; Hecht et al. 1981~. Unhindered phenols are oxidized to hydro- quinones that are extremely air-sensitive and oxidize to quinones (Irons and Sawahata 1985~. OH OH ~ ,., o OH ° (18) Quinones are strong electrophiles that un- dergo addition of nucleophiles to the dou- ble bonds of the ring (Irons and Sawahata 1985~. o o AH Nu  o OH I;>, Nu OH (1 9) This can result in the formation of protein and nucleic acid adducts and to depletion of glutathione. Polycyclic quinones are readily reduced to semiquinones that are either further reduced to hydroquinones or are oxidized by O2 to regenerate the quinone and form O2 (Smith et al. 1985~. The con- tinuous reduction and oxidation of quinones is called redox cycling and can lead to copi- ous superoxide formation (Smith et al. 1985~. This causes DNA strand scission, mutage- nicity, and toxicity, all of which probably require metals. Redox cycling may well ac- count for some of the pathophysiological effects of phenols and catechols. Quantification of Exposure and Estimation of Human Risk One of the fundamental unsolved problems of toxicology is how to extrapolate dose/ response data obtained in animal testing (usually in rodents) to risk assessment in humans. An approach to crossing this spe- cies barrier is to quantitate the dose that reaches the target organ at a series of expo- sure levels. This is called molecular dosim- etry. By knowing the amount of com- pound that must reach a target cell to exert an effect in, for example, rats, it should be possible to more intelligently estimate the risk of a given amount of the same com- pound reaching the same cell type in hu- mans. The effective dose that reaches target tissues is more relevant to risk assessment

596 Aldehydes and Alcohols in Mobile Source Emissions than are the atmospheric levels. For exam- ple, at an ambient concentration of 10 ppb, the amount of formaldehyde in- spired in 24 hr by an average human is 7 ,umole. This does not seem a significant amount until one realizes that it is almost exclusively localized in the mucus and epi- thelial cells lining the upper respiratory tract. What is the starting point to be if one is to quantitate binding of aldehydes, alco- hols, and phenols to critical intracellular targets for their toxic and carcinogenic ef- fects? Most of the toxicity exhibited by these agents is probably due to covalent binding to protein in which the aldehyde or quinone reacts as an electrophile. There is no doubt that protein binding to these compounds or their metabolites occurs and that it can cause toxicity. Saturated and a,,l3 unsaturated aldehydes as well as qui- nones bind rapidly to sulfhydryl proteins and inactivate them. Stable adducts also form to lysine residues. Aldehyde/lysine conjugates have been isolated from rat urine that most likely arise from proteolysis of aldehyde/protein conjugates (McGirr et al. 1985~. This provides direct evidence for covalent binding of aldehydes to proteins in viva. A considerable amount of informa- tion suggests that covalent binding to sulf- hydryl groups of DNA polymerases is responsible for the toxic effects of a,,l3 unsaturated aldehydes (Schauenstein et al. 1977). Evidence also suggests that the cv- tostatic effects of quinones derives from their ability to bind specifically and cova- lently to tubulin (Irons et al. 1981~. A similar reaction of tubulin or another com- ponent of the flagellar system may also account for the ciliatoxic activity of al- dehydes. Molecular dosimetry offers an approach to the quantitation of physiologically rele- vant damage resulting from exposure to aldehydes and quinone metabolites of phe- nols. Methods could be developed for the analysis of covalent adducts to proteins that are involved in pathological responses in target tissues. Of course, this requires that the key protein targets are known. Under- standing the role of individual proteins in . . . tOXlCo. OglC responses represents a mad or gap in our knowledge. This is why mech- anistic toxicology studies in cell culture are so important. Until an adequate knowledge of key protein targets is available, methods should be designed to quantitate adducts to proteins that are not necessarily important in the observed response but are abundant in the cell or tissue in which the response is observed. This approach is similar to the  use of hemoglobin for estimation of expo- sure to methylating agents and carcino- gens. Abundant proteins that contain reac- tive amine groups would be especially useful because amine/aldehyde conjugates are more stable than thiol/aldehyde con- jugates. It is likely that the target cells for the effects of airborne aldehydes are in the epithelial tissue of the nasal and respiratory tracts. This is based on the types of effects observed and the extreme reactivity of the molecules. It is very unlikely that aldehydes or quinones escape pulmonary epithelia to reach peripheral tissues. If most of the covalently bound material is localized in the nasal or respiratory t'ract, it should be pos- sible to sample these matrices in individual human subjects by ravage techniques. Molecular dosimetry of nucleic acid ad- ducts in target tissue will be important when attempting to relate inhaled dose to carcinogenic response. Thus, method de- velopment is recommended. However, quantitation of nucleic acid adducts in viva is of less value than quantitation of pro- tein adducts for estimation of inhaled dose. The levels of DNA adducts produced in cells are several orders of magnitude lower than protein adducts and are subject to varying degrees of removal by repair en- zymes, which further lowers the steady- state concentration of the nucleic acid ad- ducts. ~ Recommendation 9. Methods should be developed for quantitation of the amounts of aldehydes that reach target or- gans or potential target organs in humans and rodents. Despite the long history of investigation of the reaction of formaldehyde with nu

Lawrence I. Marnett 597 cleic acids and the knowledge that formal- dehyde is a nasal carcinogen, the identity of the abducts that it forms on reaction with DNA in vivo is unknown. This may be due to the hydrolytic instability of imine and hydroxymethyl derivatives of purines and pyrimidines, and it underscores the need for development of novel methods of iso- lation and analysis. Less is known of the reaction of acetaldehyde and acrolein with DNA although acrolein/deoxyguanosine adducts have been recently identified and detected following the reaction of acrolein with DNA in vitro (Chung et al. 1984~. Acrolein is metabolized by microsomal cy- tochrome P-450 preparations to glycidal- dehyde, which binds to DNA and is a carcinogen (International Agency for Re- search on Cancer 1976; Patel et al. 1983~. The adduct that glycidaldehyde forms on reaction-with deoxyguanosine is different than the acrolein/deoxyguanosine adducts, so isolation and quantitation of acro- lein/DNA adducts following inhalation of acrolein would determine the extent to which metabolism plays a role in acro- lein's genotoxic effects in viva. The avail- ability of methods for detection and quan- titation of aldehyde/DNA adducts formed in viva might be important as part of a molecular dosimetry approach to risk as- sessment. · Recommendation 10. Methods should be developed to detect and quantitate DNA adducts derived from formaldehyde, acet- aldehyde, and acrolein. The techniques should then be applied to the detection of DNA adducts formed in target tissues after administration of carcinogenic and subcarcingenic doses of the inhaled com- pounds. Detection of these adducts in cultured target cells would be a helpful intermediate step in adaption of the analyt- ical methods to detection of in viva adducts. There is ample precedent for the im- portance of electrophilic additions to pro- teins and nucleic acids in aldehyde and phenol biochemistry and toxicology, but the possibility that free-radical reactions contribute to their health effects has not been rigorously established. Therefore, it would be useful to conduct experiments to probe for the involvement of free radicals as mediators of aldehyde and phenol pa- thology. This is not a trivial undertaking, because free radicals are species with rela- tively short half-lives that make them nearly impossible to detect and quantitate directly (usually <1 see). Nevertheless, it is now possible to trap certain types of free  radicals that might be produced from al- dehydes and phenols (Packer 1984~. In ad- dition, products of in viva free-radical re- actions can be detected and quantitated as indirect evidence for free-radical formation (Packer 1984~. · Recommendation 11. Experiments should be performed to determine if al- dehydes exert toxicologic effects by gener- ation of free radicals. Summary A review of the literature indicates that aldehydes are the most potent biologically active substances of the compounds under consideration in mobile source emissions. They exert toxicologic effects at concentra- tions approximately 10-100 times their ambient atmospheric levels. Variations in ambient levels have been reported with occasional toxicologically relevant concen- trations reported in heavily polluted met- ropolitan areas. Inhaled aldehydes exert their toxicologic effects in the upper respi- ratory tract, and there is no reason to believe that they trigger systemic re- sponses. This may be due to their high reactivity or to the fact that they are rapidly metabolized. Metabolism can result in de- toxification or metabolic activation. The formation of substantial amounts of meth- anol and formaldehyde during normal human metabolism precludes development of analytical methods for their quantitation based on "systemic" approaches, such as plasma or urine analysis. These observa- tions mandate novel approaches to quanti

598 Aldehydes and Alcohols in Mobile Source Emissions ration of exposure and estimation of risk to the human population. Aldehydes and quinones are reactive electrophiles that form adducts with DNA and proteins. The structures of several aldehyde/nucleic acid adducts have been elucidated, but the ~ . . . . . extent ot t heir formation In veto Is un known. Phenols are oxidized to free radi- cals, which may play a role in tumor pro- motion. Aldehydes are oxidized to very reactive free radicals in chemical systems but it is uncertain if they form free radi- cals in biochemical systems in vitro or in Volvo. Summary of Research Recommendations Exposure HIGH PRIORITY Recommendation 1 Routine monitoring of atmospheric alcohol and aldehyde levels should be performed in regions where alcohol-based fuels are or will be in heavy use. LOW PRIORITY Recommendation 2 Methods should be developed to routinely analyze phenols and catechols in urban air. Health Effects HIGH PRIORITY Recommendation4 A chronic inhalation toxicology study of acrolein should be undertaken in rats, with emphasis on carcinogenicity. Recommendation 5 A chronic inhalation toxicology study of mixtures of formalde hyde and acrolein should be undertaken in rats and hamsters, with emphasis on carcinogenicity. MEDIUM PRIORITY Recommendation 3 Chronic low-dose inhalation toxicology studies should be un dertaken to determine if tissue damage occurs in response to exposure to levels of formaldehyde, acetaldehyde, and acrolein that are 10-100 times lower than their RD50s. Recommendation 8 Attempts should be made to develop an initiation-promotion protocol for carcinogenesis testing of aldehydes and other compo nents of mobile source emissions. LOW PRIORITY Recommendation 7 A chronic inhalation toxicology study of methanol should be undertaken in rats and hamsters, with emphasis on carcinogenic ity.

Lawrence I. Marnett 599 Cellular Effects HIGH PRIORITY Recommendation 6 Experiments should be undertaken in cells cultured from various segments of the upper respiratory tract to determine the mecha nisms by which aldehydes exert pathological changes such as toxicity, hyperplasia, ciliatoxicity, and so on. Molecular Dosimetry HIGH PRIORITY Recommendation 9 Methods should be developed for quantitation of the amounts of aldehydes that reach target organs or potential target organs in humans and rodents. Recommendation 10 Methods should be developed to detect and quantitate DNA adducts derived from formaldehyde, acetaldehyde, and acrolein. The techniques should then be applied to the detection of DNA adducts formed in target tissues after administration of carcino genic and subcarcingenic doses of the inhaled compounds. Detec tion of these adducts in cultured target cells would be a helpful intermediate step in adaption of the analytical methods to detection of in viva adducts. MEDIUM PRIORITY Recommendation 11 Experiments should be performed to determine if aldehydes exert toxicologic effects by generation of free radicals. References Alarie, Y. 1981. Bioassay for evaluating the potency of airborne sensory irritants and predicting accept- able levels of exposure in man, Food Cosmet. Toxi- col. 19:623-626. Albert, R. E., Sellakumar, A. R., Laskin, S., Kuschner, M., Nelson, N., and Snyder, C. A. 1982. Gaseous formaldehyde and hydrogen chloride induction of nasal cancer in the rat, J. Nat. Cancer Inst. 68:597-603. Auerbach, C., Moutschen-Dahman, M., and Mouts- chen, J. 1977. Genetic and cytogenetical effects of formaldehyde and related compounds, Mutat. Res. 39:317-362. Babink, C., Steinhagen, W. H., and Barrow, C. S. 1985. Sensory irritation response to inhaled al- dehydes after formaldehyde pretreatment, Toxicol. Appl. Pharmacol. 79:143-149. Beauchamp, R. O., Jr., Andjelkovich, D. A., Kliger Correspondence should be addressed to Lawrence J. Marnett, 435 Chemistry, Wayne State University, Detroit, MI 48202. man, A. D., Morgan, K. T., and d'A Heck, H. 1985. A critical review of the literature on acrolein toxicity, CRC Grit. Rev. Toxicol. 14:309-380. Bedford, P., and Fox, B. W. 1981. The role of formaldehyde in methylene dimethylsulphonate- induced DNA cross-links and its relevance to cyto- toxicity, Chem . -Biol. Interact. 38:119-126. Beland, F. A., Fullerton, N. F., and Heflich, R. H. 1984. Rapid isolation, hydrolysis, and chromatog- raphy of formaldehyde-modified DNA, J. Chroma- togr. 308:121-131.  Boreiko, C. J., Couch, D. B., and Swenberg, J. A. 1982. Mutagenic and carcinogenic effects of form- aldehyde, Environ. Sci. Res. 25:353-367. Buckley, L. A., Jiang, X. Z., James R. A., Morgan, K. T., and Barrow, C. S. 1984. Respiratory tract lesions induced 'by sensory irritants at the RD50 concentration, Toxicol. Appl. Pharmacol . 74:417- 429. Carson, B. L., Beall, C. M., Ellis, H. V., Baker, L. H., and Herndon, B. L. 1981. Acrolein Health Effects, NTIS PB82-161282; EPA-68-03-2928; EPA-460/3-81-034, Gov. Rep. Announce. l-'.S. In- dex 12. Chanet, R., and von Borstel, R. C. 1979. Genetic

600 effects of formaldehyde in yeast. III. Nuclear and cytoplasmic mutagenic effects, Mutat. Res. 62: 239-253. Chang, J. C. F., Steinhagen, W. H., and Barrow, C. S. 1981. Effects of single or repeated formalde- hyde exposure on minute volume of B6C3F1 mice and F-344 rats, Toxicol. Appl. Pharmacol. 61:451- 459. Chaw, Y. F. M., Crane, L. E., Lange, P., and Shapiro, R. 1980. Isolation and identification of cross-links from formaldehyde-treated nucleic ac- ids, Biochemistry 19:552~5531. Chio, K. S., and Tappel, A. L. 1969. Synthesis and characterization of the fluorescent products derived from malonaldehyde and amino acids, Biochemistry 8:2821-2826. Chung, F.-L., and Hecht, S. S. 1983. Formation of cyclic 1,N2-adducts by reaction of deoxyguanosine with c'-acetoxy-N-nitrosopyrrolidine, 4-(carb- ethoxynitrosamino)butanal, orcrotonaldehyde, Can- cer Res. 43: 123() 1235. Chung, F.-L., Young, R., and Hecht, S. S. 1984. Formation of cyclic 1,N2-propanodeoxyguanosine adducts in DNA upon reaction with acrolein or crotonaldehyde, Cancer Res. 44:990-995. Clay, K. L., Murphy, R. C., and Watkins, W. D. 1975. Experimental methanol toxicity in the pri- mate: analysis of metabolic acidosis in the monkey, Toxicol. Appl. Pharmacol. 34:49-61. Cleveland, W. S., Graedel, T. E., and Kleiner, B. 1977. Urban formaldehyde: observed correlation with source emissions and photochemistry, Atmos. Environ. 11 :357-360. Consensus Workshop on Formaldehyde. 1984. Re- port on the Consensus Workshop on Formalde- hyde, Environ. Health Perspect. 58:323-381. Dalbey, W. E. 1982. Formaldehyde and tumors in hamster respiratory tract, Toxicology 24:9-14. Dalhamn, T., and Rosengren, A. 1971. Effect of different aldehydes on tracheal mucosa, Arch. Oto- laryngol. 93:496-500. Doll, R., and Peto, R. 1981. The Causes of Cancer, pp. 124~1312, Oxford University Press, Oxford. Draminski, W., Eder, E., and Henschler, D. 1983. A new pathway of acrolein metabolism in rats, Arch. Toxicol. 52:243-247. Eells, J. T., Makar, A. B., Noker, P. E., and Tephly, T. R. 1981. Methanol poisoning and formate oxi- dation in nitrous-oxide treated rats, J. Pharmacol. Exp. Ther. 217:5701. Eells, J. T., Black, K. A., Makar, A. B., Tedford, C. E., and Tephly, T. R. 1982. The regulation of one-carbon oxidation in the rat by nitrous oxide and methionine, Arch. Biochem. Biophys. 219:316-326. Egle, J. L., Jr. 1972a. Retention of inhaled acetalde- hyde in the dog, Arch. Environ. Health 24:353- 357. Egle, J. L., Jr. 1972b. Retention of inhaled formalde- hyde, propionaldehyde, and acrolein in the dog, Arch. Environ. Health 25:119-124. Elmenhorst, H., and Schultz, C. H. 1968. Fluchtige inhaltssotffe des Tabaksrauches. Die chemischen Bestandteile der Gas-Dampf-Phase, Beitr. Tabak- forsch. 4:9~123.  Aldehydes and Alcohols in Mobile Source Emissions Fassett, D. W. 1963. Aldehydes and acetals, In: Indus- trial Hygiene and Toxicology (F. A. Patty, ed.), Vol. II, 2nd rev. ea., pp. 1959-1989, Interscience, New York. Feldman, M. Y. A. 1967. Reaction of formaldehyde with nucleotides and ribonucleic acid, Biochim. Biophys. Acta 149:20-34. Feron, V. J., and Kruysse, A. 1977. Effects of expo- sure to acrolein vapor in hamsters simultaneously treated with benzotaipyrene or dimethylnitrosa- mine, J. Toxicol. Environ. Health 3:379-394. Feron, V. J., Kruysse, A., and Woutersen, R. A. 1982. Respiratory tract tumors in hamsters exposed to acetaldehyde vapour alone or simultaneously to benzo[a]pyrene or dimethylnitrosamine, Eur. J. Cancer Cl in. Oncol . 18:13-31. Fornace, A. J., Jr. 1982. Detection of DNA single- strand breaks produced during the repair of damage by DNA-protein cross-linking agents, Cancer Res. 42:14~149. Fornace, A. J., Jr., Lechner, J. F., Grafstrom, R. C., and Harris, C. C. 1982. DNA repair in human bronchial epithelial cells, Carcinogenesis 3:137 1377. Fraenkel-Conrat, H. 1954. Reaction of nucleic acid with formaldehyde, Biochim. Biophys. Acta 15:307- 309. Frazelle, J. H., Abernethy, D. J., and Boreiko, C. J. 1983. Weak promotion of C3HlOT1/2 cell transfor- mation by repeated treatments with formaldehyde, Cancer Res. 43:3236-3239. Goldmacher, V. S., and Thilly, W. G. 1983. Formal- dehyde is mutagenic for cultured human cells, Mu- tat. Res. 116:417-422. Goldsmith, J. R., and Friberg, L. T. 1977. Effects of air pollution on human health. In: Air Pollution, Vol. II (A. C. Stern, ed.), pp. 457-610, Academic Press, New York. Goodman, J. I., and Tephly, T. R. 1971. A compar- ison of rat and human liver formaldehyde dehydro- genase, Biochim. Biophys. Acta 252:489-505. Grafstrom, R. C., Fornace, A. J., Jr., Autrup, H., Lechner, J. F., and Harris, C. C. 1983. Formalde- hyde damage to DNA and inhibition of DNA repair in human cells, Science 220:216-218. Grafstrom, R. C., Fornace, A., Jr., and Harris, C. C. 1984. Repair of DNA damage caused by formalde- hyde in human cells, Cancer Res. 44:4323-4327. Grosjean, D. 1982. Formaldehyde and other carbonyls in Los Angeles ambient air, Environ. Sci. Technol. 16:25~262. Haggard, H. W., and Greenberg, L. A. 1939. Studies on the absorption, distribution, and elimination of alcohol. The elimination of methyl alcohol, J. Phar- macol. Exp. Ther. 66:479 496. Hatch, G. G., Conklin, P. M., Chrostensen, C. C., Casto, B. C., and Nesnow, S. 1983. Synergism in the transformation of hamster embryo cells treated with formaldehyde and adenovirus, Environ. Muta- gen. 5:49-57. Hecht, S. S., Carmella, S., Mori, H., and Hoffmann, D. J. 1981. A study of tobacco carcinogenesis. 20. Role of catechol as a major cocarcinogen in the

Lawrence I. Marnett weakly acidic fraction of smoke condensate, J. Nat. Cancer Inst. 66:163-169. Heinrich, R., and Angerer, J. 1982. Occupational chronic exposure to organic solvents. X. Biological monitoring parameters for methanol exposure, Int. Arch. Occup. Environ. Health 50:341-349. Hendrick, D. J., and Lane, D. J. 1975. Formalin asthma in hospital staff, Br. Med.J. 1:607008. Hoard, D. E. 1960. The applicability of formol titra- tion to the problem of end-group determination in polynucleotides. A preliminary investigation, Bio- chim. Biophys. Acta 40:62-70. Hoshika, Y. 1977. Simple and rapid gas-liquid-solid chromatography analysis of trace concentrations of acetaldehyde in urban air, J. Chromatogr. 137:455- 460. Howard,J. A.1973. Homogeneousliquidphaseautox- idations, In: Free Radicals a K. Kochi, ed.), Vol. II, pp. 3~2, Wiley-Interscience, New York. International Agency for Research on Cancer. 1976. Glycidaldehyde, In: IARC Monographs on the Eval- uation of Carcinogenic Risk of Chemicals to Man, Vol. 11, p. 175, IARC, Lyon, France. Irons, R. D., and Sawahata, T., 1985. Phenols, cate- chols, and quinones, In: Bioactivation of Foreign Compounds (M. W. Anders, ed.), pp. 259-281, Academic Press, New York. Irons, R. D., Neptun, D. A., and Pfeifer, R. W. 1981. Inhibition of lymphocyte transformation and mi- crotubule assembly by quinone metabolites of ben- zene: evidence for a common mechanism, J. Reticu- loendothel. Soc. 30:359-372. Ishikawa, T., Esterbauer, H., and Sies, H. 1986. Role of cardiac glutathione transferase and of the gluta- thione S-conjugate export system in biotransforma- tion of 4-hydroxynonenal in the heart, J. Biol. Chem. 261:157~1581. Ito, N., Hagiwara, A., Shibata, M., Ogiso, T., and Fukushima, S. 1982. Induction of squamous-cell carcinoma in the forestomach of F344 rats treated with butylated hydroxyanisole, Gann 73:332-334. Jakab, G. J. 1977. Adverse effect of a cigarette smoke component, acrolein, on pulmonary antibacterial defenses and on viral-bacterial interactions in the lung, Am. Rev. Respir. Dis. 115:3~38. Jakoby, W. B., and Habig, W. H. 1980. Glutathione transferases, In: Enzymatic Basis of Detoxification (W. B. Jakoby, ed.), Vol. II, pp. 6~94, Academic Press, New York. Jakoby, W. B., Sekura, R. D., Lyon, E. S., Marcus, C. J., and Wang, J.-L. 1980. Sulfotransferases, In: Enzymatic Basis of DetoxiJ7cation (W. B. Jakoby, ed.), Vol. II, pp. 199-228, Academic Press, New York. Kane, L. E., Barrow, C. S., and Alarie, Y. 1979. A short-term test to predict acceptable levels of expo- sure to airborne sensory irritants, Am. Ind. Hyg. Assoc. J. 40:207-229. Kasper, C. B., and Henton, D. 1980. Glucuronida- tion, In: Enzymatic Basis of Detoxi.fication (W. B. Jakoby, ed.), Vol. II, pp. ~36, Academic Press, New York. Kerns, W. D., Pavkov, K. L., Donofrio, D. J., Gralla, E. J., and Swedberg, J. A. 1983. Carcinogenicity of  601 formaldebyde in rats and mice after long-term inhalation exposure, Cancer Res. 43:4382-4392. Kilburn, K. H., and McKenzie, W. N. 1978. Leu- kocyte recruitment to airways by aldehyde-carbon combinations that mimic cigarette smoke, Lab. Invest. 38:134-142. Kornberg, A. 1980. DNA Replication, W. H. Free- man, San Francisco. Krokan, H., Grafstrom, R. C., Sundqvist, K., Ester- bauer, H., and Harris, C. C. 1985. Cytotoxicity, thiol depletion, and inhibition of O6-methyl-gua- nine-DNA methyltransferase by various aldehydes in cultured human bronchial fibroblasts, Carcinogen- esis 6:1755-1759. Kutzman, R. S., Popenoe, E. A., Schmaeler, M., and Drew, R. T. 1985. Changes in rat lung structure and composition as a result of subchronic exposure to acrolein, Toxicology 34:139-151. Kuwata, K., Uebori, M., and Yamazaki, Y. 1981. Reversed-phase liquid chromatographic determina- tion of phenols in auto exhaust and tobacco smoke as p-nitrobenzeneazophenol derivatives, Anal. Chem. 53:1531-1534. Lee, H. K., Alarie, Y., and Karol, M. H. 1984. Induction of formaldehyde sensitivity in guinea pigs, Toxicol. Appl. Pharmacol. 75:147-155. Levy, S., Nocentini, S., and Billardon, C. 1983. Induction of cytogenetic effects in human fibroblast cultures after exposure to formaldehyde or X-rays, Mutat. Res. 119:309-317. Lloyd, W. G. 1973. Autoxidations, In: Methods in Free Radical Chemistry (E. S. Huyser, ed.), Vol. 4, pp. 1-131, Marcel Dekker, New York. Maibach, H. 1983. Formaldehyde: effects on animal and human skin, In: Formaldehyde Toxicity a E. Gibson, ed.), pp. 163-174, Hemisphere Publ. Corp., Washington, D.C. Makar, A. B., and Tephly, T. R. 1975. Inhibition of monkey liver alcohol dehydrogenase by 4- methylpyrazole, Biochem. Med. 13:330342. Mannering, G. J., and Parks, R. E., Jr. 1957. Inhibi- tion of methanol metabolism with 3-amino-1,2,4- triazole, Science 126:1241-1242. Martin-Amat, G., McMartin, K. E., Hayreh, S. S., Hayreh, M. S., and Tephly, T. R. 1978. Methanol poisoning: ocular toxicity produced by formate, Toxicol. Appl. Pharmacol. 45:201-208. Mashford, P. M., end Jones, A. R. 1982. Formalde- hyde metabolism by the rat: a reappraisal, Xenobiot- ica 12:119-124. McGhee, J. D., and von Hippel, P. H. 1975. Formal- dehyde as a probe of DNA structure. I. Reaction with exocyclic amino groups of DNA bases, Bio- chemistry 14:1281-1296. McGirr, L. G., Hadley, M., and Draper, H. H. 1985. Identification of N"-acetyl-~-2-propenal lysine as a urinary metabolite of malondialdehyde, J. Biol. Chem. 260:15427-15431. McMartin, K. E., Makar, A. B., Martin-Amat, G., Palese, G., and Tephly, T. R. 1975. Methanol poisoning. I. The role of formic acid in the devel- opment of metabolic acidosis in the monkey and the reversal by 4-methylpyrazole, Biochem. Med. 13: 319-333.

602 McMartin, K. E., Martin-Amat, G., Makar, A. B., and Tephly, T. R. 1977. Methanol poisoning. V. Role of formate metabolism in the monkey, J. Pharmacol. Exp. Ther. 201:564-572. McMartin, K. E., Martin-Amat, G., Noker, P. E., and Tephly, T. R. 1979. Lack of a role for formal- dehyde in methanol poisoning in the monkey, Biochem. Pharmacol. 28:645-649. Morgan, K. T. 1983. Localization of areas of inhibi- tion of nasal mucociliary function in rats following in vivo exposure to formaldehyde, Am. Rev. Respir. Dis. 127:166. Morgan, K. T., Patterson, D. L., and Gross, E. A. 1983. Formaldehyde and the mucociliary apparatus, In: Formaldehyde: Toxicology, Epidemiology and Mech- anisms a J. Clary, J. E. Gibson, and A. S. Waritx, eds.), Marcel Dekker, New York. Neely, W. B. 1964. The metabolic fate of formalde- hyde-~4C intraperitoneally administered to the rat, Biochem. Pharmacol. 13:1137-1142. Newsome,J. R., Norman, V., and Keith, C. H. 1965. Vapour phase analysis of tobacco smoke, Tob. Sci. 9:102-110. Noker, P. E., Fells, J. T., and Tephly, T. R. 1980. Methanol toxicity: treatment with colic acid and 5-formyltetrahydrofolic acid, Alcohol. Clin. Exp. Res. 4:378-383. Packer, L. (ed.) 1984. Methods in Enzymology. Oxygen Radicals in Biological Systems. Academic Press, New York. Patel, J. M., Gordon, W. P., Nelson, S. D., and Leibman, K. C. 1983. Comparison of hepatic bio- transformation and toxicity of allyl alcohol and [1,1-2H2]allyl alcohol in rats, Drug Metab. Dispos. 11 :164-166. Pryor, W. A. 1984. Free radicals in autoxidation and in aging, In: Free Radicals in Molecular Biology, Aging, and Disease (D. Armstrong, R. S. Sohal, R. G. Cutler, and T. F. Slater, eds.), pp. 13-41, Raven Press, New York. Ragan, D. L., and Boreiko, C. J. 1981. Initiation of C3H/lOT1/2 cell transformation by formaldehyde, Cancer Lett. 13:325-331. Rietbrock, V. 1969. Kinetik und wege des Methanol- umsatzes, Naunyn-Schmiedebergs Arch. Pharmakol. Exp. Pathol. 263:88-105. Roe, O. 1955. The metabolism and toxicity of meth- anol, Pharmacol. Rev. 7:399 ~112. Roe, O. 1982. Species differences in methanol poison- ing, CRC Crit. Rev. Toxicol. 10:275-286. Ross, W. E., and Shipley, N. 1980. Relationship between DNA damage and survival in formalde- hyde-treated mouse cells, Mutat. Res. 79:277-283. Ross, W. E., McMillan, D. R., and Ross, C. F. 1981. Comparison of DNA damage by methylmelamines and formaldehyde, J. Nat. Cancer Inst. 67:217-221. Schauenstein, E., Esterbauer, H., and Zollner, H. 1977. Aldebydes in Biological Systems, pp. 35-38, Pion Limited, London. Sedivec, V., Mraz, M., and Flek, J. 1981. Biological monitoring of persons exposed to methanol va- pours, Int. Arch. Occup. Environ. Health 48:257-271. Sellakumar, A. R., Snyder, C. A., Solomon,J.J., and Albert, R. E. 1985. Carcinogenicity of formalde  Aldehydes and Alcohols in Mobile Source Emissions hyde and hydrogen chloride in rats, Toxicol. Appl. Pharmacol. 81:401-406. Simic, M. G., and Hunter, E. P. L. 1983. Interactions of free radicals and antioxidants, In: Radioprotectors and Anticarcinogens (O. F. Nygaard and M. G. Simic, eds.), pp. 449-460, Academic Press, New York. Slaga, T. J., Klein-Szanto, A. J., Triplett, L. L., Yotti, L. P., and Trosko, J. E. 1981. Skin tumor-pro- moting activity of benzoyl peroxide, a widely used free radical-generating compound, Science 213: 1023-1025. Slaga, T. J., Solanki, V., and Logani, M. 1983. Studies on the mechanism of action of antitumor promoting agents: suggestive evidence for the in- volvement of free radicals in promotion, In: Radio- protectors and Anticarcinogens (O. F. Nygaard and M. G. Simic, eds.), pp. 471-485, Academic Press, New York. Slott, V. L., and Hales, B. F. 1985. Teratogenicity and embryolethality of acrolein and structurally related compounds in rats, Teratology 323:6~72. Smith, M. T., Evans, C. G., Thor, H., and Orrenius, S. 1985. Quinone-induced oxidative injury to cells and tissues, In: Oxidative Stress (H. Sies, ed.), pp. 91-114, Academic Press, New York. Snider, J. R., and Dawson, G. A. 1985. Tropospheric light alcohols, carbonyls, and acetonitrile: concen- trations in the southwestern United States and Henry's law data, J. Geophys. Res. 90:3797-3805. Strittmatter, P., and Ball, E. G. 1955. Formaldehyde dehydrogenase, a glutathione-dependent enzyme system, J. Biol. Chem. 213:445-461. Swarin, S. J., and Lipari, F. 1983. Determination of formaldehyde and other aldehydes by high per- formance liquid chromatography with fluorescence detection, J. Liq. Chromatogr. 6:425-444. Swenberg, J. A., Barrow, C. S., Boreiko, C. J., d'A Heck, H., Levine, R. J., Morgan, K. T., and Starr, T. B. 1983. Non-linear biological responses to formaldehyde and their implications for carcino- genic risk assessment, Carcinogenesis 4:945-952. Szabad, J., Soos, I., Polgar, G., and HeiJa, G. 1983. Testing the mutagenicity of malondialdehyde and formaldehyde by the Drosophila mosaic and the sex-linked recessive lethal tests, Mutat. Res. 113: 117-133. Tanner, R. L., and Meng, Z. 1984. Seasonal varia- tions in ambient atmospheric levels of formalde- hyde and acetaldehyde, Environ. Sci. Technol. 18: 723-726. Tephly, T. 1985. The Toxicity of Methanol and Its Metabolites in Biological Systems, Health Effects Institute Report, Cambridge, Mass. Thelander, L., and Reichard, P. 1979. Reduction of ribonucleotides, Ann. Rev. Biochem. 48:133-158. Travis, J., and Salvesen, G. S. 1983. Human plasma proteinase inhibitors, Ann. Rev. Biochem. 52:655- 709. Treitman, R. D., Burgess, W. A., and Gold, A. 1980. Air contaminants encountered by firefighters, Am. Ind. Hyg. Assoc.J. 41:79~802. Van Dunren, B. L., and Goldschmidt, B. M. 1976. Carcinogenic and tumor-promoting agents in to

Lawrence I. Marnett 603 bacco carcinogenesis, J. Nat. Cancer Inst. 56: 1237- 1242. Van Duuren, B. L., and Loevrengart, G. 1977. Reac- tion of DNA with glycidaldehyde. Isolation and identification of a deoxyguanosine reaction prod- uct, J. Biol. Chem. 252:537~5371. Van Dunren, B. L., Orris, L., and Nelson, N. 1965. Carcinogenicity of epoxides, lactones, and peroxy compounds. II. J. Nat. Cancer Inst. 35:707-717. Van Duuren, B. L., Langseth, L., Orris, L., Teebor G., Nelson, N., and Kuschner, M. 1966. Carcino- genicity of epoxides, lactones, and peroxy com- pounds. IV. Tumor response in epithelial and con- nective tissue in mice and rats, J. Nat. Cancer Inst. 37:825-838. Van Duuren, B. L., Langseth, L., Goldschmidt, B. M., and Orris, L. 1967. Carcinogenicity of epoxides, lactones, and peroxy compounds. VI. Structure and carcinogenic activity, J. Nat. Cancer Inst. 39:1217-1228. Watkins, W. D., Goodman, J. I., and Tephly, T. R. 1970. Inhibition of methanol and ethanol oxidation by pyrazole in the rat and monkey in viva, Mol. Pharmacol. 6:567-572. Willson, R. L. 1985. Organic peroxy free radicals as ultimate agents in oxygen toxicity, In: Oxidative Stress (H. Sies, ed.), pp. 41-72, Academic Press, New York. Witz, G., Lawrie, N. J., Amoruso, M. A., and Goldstein, B. D. 1985. Inhibition by reactive al- dehydes of superoxide anion radical production in stimulated human neutrophils, Chem.-Biol. Interact. 53:13-23. Woutersen, R. A., Appleman, L. M., Feron, V. J., and Van Der Heijden, C. A. 1984. Inhalation tox- icity of acetaldehyde in rats. II. Carcinogenicity study: interim results after 15 months, Toxicology 31:12~133. Yant, W. P., and Schrenck, H. H. 1937. Distribution of methanol in dogs after inhalation of methanol and administration by stomach tube and subcutane- ously,J. Ind. Hyg; Toxicol. 19:337-345.