Abstract

The fact that acrylamide, a proven rodent carcinogen, is present in significant quantities (up to several mg/kg of foodstuff) in a wide range of commonly consumed human foods is alarming. Attempts to determine a possible involvement of dietary acrylamide in human cancers have not been conclusive, however. To resolve the carcinogenicity of acrylamide to humans, the as yet unknown mechanism of action of acrylamide needs to be unraveled. The present review is a synopsis of research on the known and hypothetical modes of action of acrylamide of relevance for carcinogenesis. Both genotoxic and non-genotoxic modes of action of acrylamide are discussed with special emphasis on DNA adduct-targeted mutagenesis. Mechanistic data are presented from various experimental systems including in vitro experiments and in vivo rodent and human studies with special focus on mouse models. Human exposure data, including estimates of daily intake of dietary acrylamide in different populations and the corresponding cancer risk assessments are provided. The significant gaps in knowledge, which currently preclude a more definitive evaluation of human cancer risk due to exposure to dietary acrylamide, are highlighted. Future directions for research on acrylamide and cancer are outlined, and potential challenges are underscored.

Introduction

Acrylamide is a low molecular weight vinylic compound ( Figure 1 ). This colorless and odorless crystalline substance is highly water soluble, easily reactive in air and rapidly polymerizable, i.e. single molecules of acrylamide (monomers) can bind together and form a larger molecule (polymer) with new properties. Polymers of acrylamide (polyacrylamide) are extensively used in modern chemical technology for a variety of purposes. These include their use as flocculants for sewage and wastewater treatment, as coagulants for clarifying drinking water, as sealants for construction of dams, tunnels and water reservoirs, as soil stabilizers in roadways construction, as binders in paper and pulp industry and as additives/adhesives/fixatives for manufacturing various industrial and cosmetic products. In analytical biochemistry, polyacrylamides are widely used for chromatography and electrophoresis, e.g. for separation and purification of proteins ( 1 ).

Fig. 1.

Chemical structures of acrylamide and its epoxide glycidamide and the respective oxidative biotransformation by CYP2E1. The schematic metabolic pathway through which acrylamide is converted to glycidamide is shown.

Fig. 1.

Chemical structures of acrylamide and its epoxide glycidamide and the respective oxidative biotransformation by CYP2E1. The schematic metabolic pathway through which acrylamide is converted to glycidamide is shown.

The vinyl group of acrylamide is electron deficient and can be easily attacked by nucleophiles. This characteristic electrophilicity of acrylamide enables it to interact with nucleophilic biological macromolecules ( 2 ). The biological consequences of acrylamide exposure have chiefly centered on neurotoxicity ever since this effect was observed in humans occupationally exposed to this compound ( 3 ). Subsequently, experimental exposure of rodents to acrylamide has also revealed a carcinogenic mode of action for this chemical [reviewed in ( 1 , 4 )]. However, epidemiologic studies in work settings have not established an unambiguous link between acrylamide exposure and human cancers ( 5–7 ).

Research on carcinogenicity of acrylamide to humans regained momentum consequent to a tunnel construction accident in Sweden ( 8 , 9 ). Floating dead fish in an aquaculture pond downstream of a brook, wherein tunnel drainage water was pumped out, paralyzed cows that had grazed in the vicinity of the brook and tunnel construction workers with symptoms of neurotoxicity led the local authorities to declare a state of emergency ( 9 ). ‘Rhoca Gil’, an acrylamide-based grouting agent, which had been used to seal the tunnel leakage, was found in high concentrations in the water from the brook and the wells nearby ( 9 ). Elevation of acrylamide–hemoglobin (Hb) adducts in the blood of the affected cows and construction workers suggested the involvement of acrylamide in this disastrous chain of events ( 10 , 11 ). Surprisingly, considerable levels of Hb-bound acrylamide (adducts) were detected in the blood of non-exposed control individuals. The latter finding confirmed an unexplained detection of acrylamide–Hb adducts previously reported in a control population of non-smokers ( 12 ). The background levels of acrylamide in the control populations were suspected to have arisen from dietary sources ( 8 , 9 ). This theory was substantiated by demonstration that rats fed with fried chow had significantly higher levels of acrylamide–Hb adducts relative to controls. In confirmation, chemical analysis showed substantial levels of acrylamide in the fried chow, whereas negligible amounts of acrylamide were detectable in the non-fried chow ( 13 ).

Public health concern about acrylamide escalated after the provocative announcement of the presence of acrylamide in a host of commonly consumed foods particularly in fried, roasted and baked starch-based goods ( 14 ). Subsequent biochemical analysis identified the origin of dietary acrylamide by demonstration that the heating of free amino acids, in particular asparagine, and sugars during food processing (120–180°C) results in the formation of acrylamide ( 15 , 16 ). Following these findings, a few epidemiologic studies have attempted to link dietary acrylamide to human cancers ( 17 , 18 ). However, such questionnaire/interview-based studies have been inconclusive because of the ubiquity of acrylamide in the human diet and its variable generation even within the same brands of food (dependent on preparation method and ingredients used) ( 4 ). A comprehensive and updated summary of acrylamide content of various food products is shown in Table I .

Table I.

Acrylamide content of various food products

Food category Food item  Number of samples a Mean concentration (μg/kg) Maximum concentration (μg/kg)  CV (%) b 
Cereals and cereal-based products  3304 (12 346) 343 7834 156 
Cereals and pasta (raw and boiled) 113 (372) 15 47 71 
Cereals and pasta (processed: toasted, fried, grilled) 200 (634) 123 820 110 
Cereal-based processed products (all) 2991 (11 327) 366 7834 151 
Breads and rolls 1294 (5145) 446 3436 130 
Pastry and biscuits (US = cookies) 1270 (4980) 350 7834 162 
Breakfast cereals 369 (1130) 96 1346 131 
Pizza 58 (85) 33 763 270 
Fish and seafood (breaded, fried, baked)  52 (107) 25 233 180 
Meat and offal (coated, cooked, fried)  138 (325) 19 313 174 
Milk and dairy products  62 (147) 5.8 36 119 
Nuts and oilseeds  81 (203) 84 1925 233 
Pulses  44 (93) 51 320 137 
Root and tubers  2068 (10 077) 477 5312 108 
Potato purees/mashed/boiled 33 (66) 16 69 92 
Potato baked 22 (99) 169 1270 150 
Potato crisps (US = chips) 874 (3555) 752 4080 73 
Potato chips (US = french fries) 1097 (6309) 334 5312 128 
Potato chips, croquettes (frozen, not ready-to-serve) 42 (48) 110 750 145 
Stimulants and analogues  469 (1455) 509 7300 120 
Coffee (brewed), ready-to-drink 93 (101) 13 116 100 
Coffee (ground, instant or roasted, not brewed) 205 (709) 288 1291 51 
Coffee extracts 20 (119) 1100 4948 93 
Coffee decaffeinate 26 (34) 668 5399 169 
Coffee substitutes 73 (368) 845 7300 90 
Cocoa products 23 (23) 220 909 111 
Green tea (roasted) 29 (101) 306 660 67 
Sugars and honey (mainly chocolate)  58 (133) 24 112 87 
Vegetables  84 (193) 17 202 206 
Raw, boiled and canned 45 (146) 4.2 25 103 
Processed (toasted, baked, fried, grilled) 39 (47) 59 202 109 
Fruits (fresh) 11 (57) <1 10 188 
Fruits (processed: dried, fried) 37 (49) 131 770 125 
Miscellaneous Alcoholic beverages (beer, gin, wine) 66 (99) 6.6 46 143 
Condiments and sauces 19 (22) 71 1168 345 
Infant formula 82 (117) <5 15 82 
Baby food (canned, jarred) 96 (226) 22 121 82 
Baby food (dry powder) 24 (34) 16 73 125 
Baby food (biscuits, rusks, etc.) 32 (58) 181 1217 106 
Dried food 13 (13) 121 1184 266 
Food category Food item  Number of samples a Mean concentration (μg/kg) Maximum concentration (μg/kg)  CV (%) b 
Cereals and cereal-based products  3304 (12 346) 343 7834 156 
Cereals and pasta (raw and boiled) 113 (372) 15 47 71 
Cereals and pasta (processed: toasted, fried, grilled) 200 (634) 123 820 110 
Cereal-based processed products (all) 2991 (11 327) 366 7834 151 
Breads and rolls 1294 (5145) 446 3436 130 
Pastry and biscuits (US = cookies) 1270 (4980) 350 7834 162 
Breakfast cereals 369 (1130) 96 1346 131 
Pizza 58 (85) 33 763 270 
Fish and seafood (breaded, fried, baked)  52 (107) 25 233 180 
Meat and offal (coated, cooked, fried)  138 (325) 19 313 174 
Milk and dairy products  62 (147) 5.8 36 119 
Nuts and oilseeds  81 (203) 84 1925 233 
Pulses  44 (93) 51 320 137 
Root and tubers  2068 (10 077) 477 5312 108 
Potato purees/mashed/boiled 33 (66) 16 69 92 
Potato baked 22 (99) 169 1270 150 
Potato crisps (US = chips) 874 (3555) 752 4080 73 
Potato chips (US = french fries) 1097 (6309) 334 5312 128 
Potato chips, croquettes (frozen, not ready-to-serve) 42 (48) 110 750 145 
Stimulants and analogues  469 (1455) 509 7300 120 
Coffee (brewed), ready-to-drink 93 (101) 13 116 100 
Coffee (ground, instant or roasted, not brewed) 205 (709) 288 1291 51 
Coffee extracts 20 (119) 1100 4948 93 
Coffee decaffeinate 26 (34) 668 5399 169 
Coffee substitutes 73 (368) 845 7300 90 
Cocoa products 23 (23) 220 909 111 
Green tea (roasted) 29 (101) 306 660 67 
Sugars and honey (mainly chocolate)  58 (133) 24 112 87 
Vegetables  84 (193) 17 202 206 
Raw, boiled and canned 45 (146) 4.2 25 103 
Processed (toasted, baked, fried, grilled) 39 (47) 59 202 109 
Fruits (fresh) 11 (57) <1 10 188 
Fruits (processed: dried, fried) 37 (49) 131 770 125 
Miscellaneous Alcoholic beverages (beer, gin, wine) 66 (99) 6.6 46 143 
Condiments and sauces 19 (22) 71 1168 345 
Infant formula 82 (117) <5 15 82 
Baby food (canned, jarred) 96 (226) 22 121 82 
Baby food (dry powder) 24 (34) 16 73 125 
Baby food (biscuits, rusks, etc.) 32 (58) 181 1217 106 
Dried food 13 (13) 121 1184 266 

Data were obtained from the Summary Report of the 64th Meeting of the Joint Food and Agriculture Organization of the United Nations/World Health Organization Expert Committee on Food Additives ( http://www.who.int/ ). Acrylamide occurrence data for different food items analyzed from 2002 to 2004 were provided from 24 countries (see ref. 74 ). The total number of analytical results (single or composite samples) was 6752 with 67.6% from Europe, 21.9% from North America, 8.9% from Asia and 1.6% from Pacific. No data from Latin America and Africa were submitted.

a

Number of analytical results for individual plus composite samples. In parentheses, the total numbers of individual samples are shown.

b

CV, coefficient of variation.

To address the involvement of dietary acrylamide in human cancers, it would be more appropriate to conduct dietary intervention studies in which human volunteers are supplemented with foods rich in acrylamide and monitored for relevant biomarkers of exposure and effects pre- and post-supplementation. A prerequisite for such studies is the availability of biomarkers both specific for acrylamide and pertinent to cancer. Identifying these biomarkers can be achieved by unraveling the as yet unknown mechanism of acrylamide carcinogenicity ( 19 ). This review article summarizes the current knowledge of acrylamide-induced carcinogenesis, and discusses the uncertainties surrounding this field of research. Emphasis is placed on acrylamide genotoxicity and tumorigenicity with special focus on DNA adduct-targeted mutagenesis, an important area of research for which a growing body of evidence is available ( 19 ). Mechanistic data are recapitulated from experimental systems, including in vitro experiments and in vivo rodent and human studies. Highlights are provided from mouse models, for which the available data are most comprehensive for many aspects of DNA adduction, mutagenesis and tumorigenesis.

Biological properties of acrylamide

The low molecular weight and high water solubility of acrylamide enable this compound to easily pass through various biological membranes ( 1 , 2 , 4 ). Subsequently, the characteristic chemical structure of acrylamide and its ability to undergo metabolic transformation ( 20 ) make it react with different (sub)cellular targets ( 1 , 2 , 4 ). The chemical structure of acrylamide distinctive by an α,β-unsaturated carbonyl group ( Figure 1 ) constitutes its ‘Michael-type’ reactivity toward thiol, hydroxyl or amino groups, and to a lesser extent the nucleophilic centers in DNA. The Michael-type addition of acrylamide to thiols of glutathione (GSH) occurs most efficiently, and results in urinary excreted mercapturic acid conjugates, thereby representing a detoxification pathway ( 1 , 2 , 4 ). However, the addition of acrylamide to thiols of proteins such as protamines can cause alkylation, thus potentially causing epigenetic effects ( 4 ). Acrylamide can also bind to plasma proteins, primarily Hb, with an as yet undefined biological consequence. To date, Hb-bound acrylamide is essentially considered as the internal dose marker of exposure to acrylamide. Furthermore, acrylamide can undergo oxidative biotransformation by cytochrome P450 2E1 (CYP2E1) ( 20 ). The resulting metabolite is an epoxide derivative, i.e. glycidamide ( Figure 1 ), which is more reactive toward DNA and proteins than the parent compound, acrylamide ( 2 ).

Acrylamide genotoxic versus non-genotoxic modes of action

Binding of acrylamide to plasma proteins and its conjugation with GSH compete against acrylamide genotoxicity, which arises from the reaction of this compound or its metabolite, glycidamide, with DNA ( 4 ). The conjugation of acrylamide with GSH can also result in depletion of cellular GSH stores, thereby changing the redox status of the cell ( 21 ). The latter can potentially affect gene expression directly or through regulating various transcription factors, which are redox dependent ( 22–25 ). Consequently, cell transformation or proliferation and apoptosis might occur independently of acrylamide-induced genotoxicity ( 22–25 ). A hormonal mode of action for acrylamide has also been hypothesized ( 26–31 ) in view of tumorigenicity of acrylamide in rat endocrine (thyroid) and mammary glands ( 1 , 2 , 4 ). This hypothesis, however, remains to be substantiated by mechanistic studies ( 32 ). Also, critics have argued that many genotoxins induce tumorigenesis at the same organ sites without causing any hormonal imbalance/dysregulation ( 32 ). Admittedly, however, despite the large body of evidence supporting a genotoxic mode of action for acrylamide, the observed tissue-specific tumorigenicity of acrylamide cannot solely be explained by its DNA damaging and mutagenic effects (see Acrylamide-induced tumorigenesis).

Synopsis of research on acrylamide carcinogenicity

To date, scant data are available on the involvement of a non-genotoxic mode of action in acrylamide carcinogenicity. However, a large body of evidence exists on acrylamide genotoxicity in relation to its carcinogenicity. In the following sections, we will summarize the existing data on metabolism and bioavailability, DNA adduct formation, mutagenicity and tumorigenicity of acrylamide, and also discuss the interdependence of these events insomuch as the available information permits. The compiled data are from in vitro and/or in vivo studies using various experimental model systems, including rodents and humans. Because the data obtained from mouse model studies are most exhaustive, we will highlight the findings of these studies at the end of each section.

Toxicokinetics of acrylamide

Toxicokinetic studies in humans have shown a terminal elimination half-life of 2.4–7.0 h for acrylamide ( 33 , 34 ). Over 60% of acrylamide taken up by the body can be recovered from the urine ( 34 , 35 ), of which ∼86% are GSH conjugates ( 36 ). These include mercapturic acid metabolites of acrylamide, i.e. N -acetyl- S -(2-carbamoylethyl)cysteine, and glycidamide, i.e. N -acetyl- S -(2-hydroxy-2-carbamoylethyl)cysteine, and minor amounts of other derivatives ( 36 ). The ratio of glycidamide- to acrylamide–GSH conjugates excreted in human urine is ∼0.1 ( 34 , 35 ). Unchanged acrylamide in the urinary excrete accounts for 4.4% of the uptaken acrylamide, whereas unchanged glycidamide can only be found in marginal quantities ( 34 , 35 ). Both acrylamide and glycidamide have a significant affinity for binding to plasma proteins, in particular, to Hb ( 4 ). The Hb-bound acrylamide and glycidamide are N-terminal valine adducts, which are widely recognized as the internal dose markers of ‘exposure’ to acrylamide ( 4 ). Acrylamide can be found in human placenta ( 37 ) and breast milk ( 33 ), thus easily transferable to fetus or newborn infants. Transplacental exposure to acrylamide in humans has been shown by the detection of acrylamide–Hb adducts in the blood of pregnant women as well as the umbilical cord blood of their respective neonates ( 37 ). Given the shorter lifespan of neonatal erythrocytes and the lower body weight of the infants relative to adults, the internal dose of acrylamide in neonates is estimated to be at least equal to that in the mothers ( 37 ).

Toxicokinetic studies of acrylamide in rodents have shown a lower detoxification of acrylamide in mice than in rats by demonstration that per administered dose of acrylamide, mice had considerably lower levels of urinary excreted acrylamide–GSH conjugates relative to rats. However, the metabolic activation of acrylamide to glycidamide was more efficient in mice than in rats because the levels of free and GSH-conjugated glycidamide were significantly higher in the mice than in the rats ( 38 ). Also, the extent of glycidamide–Hb adduct formation per administered dose of acrylamide was 3–10 times greater in mice than in rats ( 39 ). The ratio of glycidamide- to acrylamide–GSH conjugates in rats and mice was 0.2 and 0.5, respectively ( 35 ). Thus, it can be inferred that rodents metabolize acrylamide to glycidamide more efficiently than humans. It is estimated that at equimolar dose of acrylamide, humans have ∼2- and 4-fold, respectively, lower internal exposure to glycidamide than rats and mice ( 34 , 35 ).

Acrylamide metabolism and bioavailability in mice

Toxicokinetic studies of acrylamide in male and female B6C3F1 mice following acute dosing by intravenous, gavage and dietary routes at 0.1 mg/kg/body wt showed that acrylamide was rapidly absorbed after oral dosing, widely distributed throughout various tissues and efficiently converted to glycidamide ( 40 ). Increased levels of glycidamide–DNA adducts were observed in liver after complete elimination of acrylamide from serum. Intravenous and gavage dosing with an equimolar amount of glycidamide also resulted in rapid absorption and wide distribution to tissues, and produced liver DNA adduct levels that were ∼40% higher than those from an equimolar dose of acrylamide ( 40 ). While oral administration was found to attenuate acrylamide bioavailability to 23% from the diet and 32–52% from aqueous gavage, a first-pass effect or other kinetic change resulted in higher relative internal exposure to glycidamide when compared with the intravenous route ( 40 ). A similar effect on relative glycidamide exposure was also evident as the administered dose was reduced, which suggests that as dosing rate decreases, the conversion of acrylamide to glycidamide is more efficient ( 40 ).

Kinetic studies by Sumner et al. ( 20 , 38 ) have established a major role of CYP2E1 in biotransformation of acrylamide to glycidamide. These investigators have shown that administration of acrylamide (50 mg/kg/body wt single oral gavage) to wild-type mice pre-treated with an inhibitor of P450, 1-aminobenzotriazole, as well as to mice deficient of CYP2E1 resulted in the urinary excretion of only acrylamide–GSH conjugates ( 20 , 38 ). However, similar administration of acrylamide to wild-type mice without any 1-aminobenzotriazole pre-treatment yielded both glycidamide- and acrylamide-conjugated urinary metabolites ( 20 , 38 ). Gamboa da Costa et al. ( 41 ) have demonstrated that consistent with the known deficiency of P450 activity in neonatal mice ( 42 ), formation of DNA adducts induced by glycidamide [50 mg/kg/body wt, single intra-peritoneal (i.p.) injection] in mouse neonates was 5- to 7-fold higher than that in their counterparts treated with an equimolar dose of acrylamide. Adult mice, however, showed only 1.2- to 1.5-fold higher formation of DNA adducts after treatment with glycidamide relative to acrylamide ( 41 ).

Acrylamide-induced DNA adduction and mutagenesis

Direct interaction of acrylamide with DNA through Michael-type addition reaction is considerably slow ( 4 ). In vitro treatment of calf thymus DNA with acrylamide at physiological pH and temperature for prolonged periods of time (up to 40 days) revealed a slow reactivity of acrylamide toward DNA. The reaction products were alkylated 2′-deoxynucleoside adducts ( 43 ) of adenine, cytosine and guanine ( 43 ). Addition of S9 liver extracts to the reaction mix of acrylamide and DNA, however, resulted predominantly in the formation of 7-(2-carbamoyl-2-hydroxyethyl)-guanine (N7-GA-Gua), a glycidamide-derived DNA adduct ( Figure 2a ) ( 41 , 44 ). Similarly, in vivo oral or topical administration of acrylamide to mice and rats induced primarily N7-GA-Gua adducts, which were evenly distributed throughout the animals' organs ( 41 , 44 ). The latter finding was consistent with the uniform and ubiquitous presence of the labeled acrylamide administered to the animals ( 44 ). The omnipresence of acrylamide and glycidamide and the respective DNA adducts ( 41 , 44 ) is consistent with the chemical properties of acrylamide, i.e. being a small and highly water-soluble molecule that can easily pass through biological membranes and reach various organs ( 1 , 2 , 4 ). Further characterization of the induced DNA adducts in acrylamide- and glycidamide-treated mice identified a minor glycidamide–DNA adduct, i.e. 3-(2-carbamoyl-2-hydroxyethyl)-adenine ( Figure 2b ). The formation of another minor glycidamide-based DNA adduct, i.e. 1-(2-carboxy-2-hydroxyethyl)-2′-deoxyadenosine, was also deduced ( Figure 2c ) ( 41 ).

Fig. 2.

Chemical structures of acrylamide-induced DNA adducts. The major adduct ( a ) is identified as N7-GA-Gua, and two minor adducts are identified as ( b ) 3-(2-carbamoyl-2-hydroxyethyl)-adenine and ( c ) 1-(2-carboxy-2-hydroxyethyl)-2′-deoxyadenosine. The predominant N7-GA-Gua adduct was formed ∼100-fold more extensively than 3-(2-carbamoyl-2-hydroxyethyl)-adenine adduct in all tested organs, i.e. liver, lung and kidney. The formation of 1-(2-carboxy-2-hydroxyethyl)-2′-deoxyadenosine adduct in vitro was estimated to be at least 11-fold higher than that of 3-(2-carbamoyl-2-hydroxyethyl)-adenine adduct ( 41 ).

Fig. 2.

Chemical structures of acrylamide-induced DNA adducts. The major adduct ( a ) is identified as N7-GA-Gua, and two minor adducts are identified as ( b ) 3-(2-carbamoyl-2-hydroxyethyl)-adenine and ( c ) 1-(2-carboxy-2-hydroxyethyl)-2′-deoxyadenosine. The predominant N7-GA-Gua adduct was formed ∼100-fold more extensively than 3-(2-carbamoyl-2-hydroxyethyl)-adenine adduct in all tested organs, i.e. liver, lung and kidney. The formation of 1-(2-carboxy-2-hydroxyethyl)-2′-deoxyadenosine adduct in vitro was estimated to be at least 11-fold higher than that of 3-(2-carbamoyl-2-hydroxyethyl)-adenine adduct ( 41 ).

Acrylamide was non-mutagenic in bacterial Ames assays with or without an exogenous activation system ( 1 , 45 ). Acrylamide was tested negative in assays using different strains of Salmonella typhimurium , including TA1535, TA1537, TA1538, TA97, TA98, TA100 and TA102 in the plate incorporation and/or liquid pre-incubation procedures in the presence or the absence of an exogenous activating system ( 1 , 45 ). Glycidamide was, however, mutagenic in the same assays even in the absence of an exogenous activating system ( 45 ). The observed non-mutagenicity of acrylamide has been ascribed to a suboptimal conversion of acrylamide to glycidamide by the utilized activating system ( 1 , 2 , 4 ). The widely used Aroclor-induced S9 mix is known to be competent in the epoxidation of a variety of chemical carcinogens, including polycyclic aromatic hydrocarbons, e.g. dimethylbenzo[ a ]anthracene and benzo[ a ]pyrene, as well as olefinic compounds, e.g. vinyl chloride and butadiene ( 2 ). Specifically, however, the presence of CYP2E1, the enzyme required for epoxidation of acrylamide to glycidamide ( 20 ), in this activating system remains to be determined.

Acrylamide was shown to be mutagenic at the hypoxanthine–guanine phosphoribosyl transferase (hprt) locus in cultured Chinese hamster ovary cells in the presence or the absence of exogenous activating systems. Acrylamide was also mutagenic and/or clastogenic at the thymidine kinase locus in mouse lymphoma cells even in the absence of an exogenous activation system ( 1 ). Acrylamide clastogenicity has been verified independently by other studies ( 46–49 ). Acrylamide caused meiotic mutations in mouse germ cells assayed by the morphological specific-locus test ( 50 ). Acrylamide also induced spots of genetic relevance, indicative of point mutations/chromosomal loss/somatic recombination, quantified by a mammalian spot test in mouse offspring borne from embryos exposed to acrylamide at 10–12 days post-gestation ( 51 ). Acrylamide was mutagenic in the lacZ transgene in Muta mice treated i.p. with daily doses of acrylamide (50 mg/kg/body wt) for five consecutive days ( 52 ). A recent mutagenicity experiment in Big Blue mice has also shown that animals receiving acrylamide or glycidamide at equimolar doses of up to 500 mg/l in drinking water for 3–4 weeks had significant increases in mutant frequencies of the hprt and cII genes 21 days following the last treatments ( 49 ).

Acrylamide genotoxicity in mice

A single i.p. injection of acrylamide (50–53 mg/kg/body wt) to male adult mice of two different strains, i.e. BALB/c and C3H/HeNMTV, induced significantly acrylamide–DNA adduction in various organs 6 h post-treatment ( 41 , 44 ). The induced DNA adducts detected by liquid chromatography with tandem mass spectrometry were all of glycidamide nature, including a predominant adduct, N7-GA-Gua, and a minor adduct, 3-(2-carbamoyl-2-hydroxyethyl)-adenine, with another minor adduct, 1-(2-carboxy-2-hydroxyethyl)-2′-deoxyadenosine, being also implicated ( Figure 2 ). The two minor adducts comprised roughly ∼1% of the overall detected adducts. The predominant adduct was elevated on average a few 100-fold over background in all investigated organs, e.g. ∼380-, 240- and 110-fold increases in lung, liver and kidney, respectively ( 41 , 44 ). The background level of N7-GA-Gua adducts varied in the range of 5–11 × 10 −8 nucleotides in different organs, and was ascribed to the acrylamide content of standard rodent chow and drinking water ( 13 ). The administered amounts of acrylamide yielded doses of 606–642 μM/kg/body wt per mouse, and were well tolerated without causing any severe adverse health effects ( 41 ).

Adler et al. ( 53 ) have shown that pre-treatment of mice (102/E1 × C3H/E1: F1) with 1-aminobenzotriazole for three consecutive days (50 mg/kg/body wt daily) inhibited or significantly reduced the dominant lethal mutagenic effects in germ cells induced by a single i.p. injection of acrylamide (125 mg/kg/body wt) at day 4. To explore the role of CYP2E1 metabolism in the germ cell mutagenicity of acrylamide, CYP2E1 -null and wild-type male mice were treated daily with acrylamide by i.p. injection (0, 12.5, 25 or 50 mg/kg/body wt) for five consecutive days ( 46 ). At defined times after exposure, males were mated to untreated B6C3F1 females. Females were killed in late gestation and uterine contents were examined. Dose-dependent increases in dominant lethal mutations were detected in uterine contents of female mice mated to acrylamide-treated wild-type males but not CYP2E1 -null males ( 46 ). A follow-up study examined the induction of somatic cell damage assessed in leukocytes, liver and lung of female wild-type and CYP2E1 -null mice administered acrylamide i.p. (0, 25, 50 mg/kg/body wt) once daily for five consecutive days using the alkaline single-cell gel electrophoresis assay ( 54 ). The frequency of micronucleus formation was also determined in erythrocytes of the mice 24 h after the last treatments using the flow cytometry assay. Significant dose-related increases in induction of DNA damage in somatic cells and formation of micronucleated erythrocytes were observed in acrylamide-treated wild-type but not in the CYP2E1 -null mice ( 54 ). The overall findings support the notion that genetic damage in somatic and germ cells of mice treated with acrylamide is dependent upon metabolism of the parent acrylamide to glycidamide by CYP2E1 ( 46 , 54 ).

Recently, in vivo genotoxicity of acrylamide was investigated in groups of male and female Big Blue mice administered 0, 100 or 500 mg/l of acrylamide or equimolar doses of glycidamide (via drinking water) for 3–4 weeks ( 49 ). The frequency of micronucleated reticulocytes was significantly increased in males treated with the high doses of acrylamide and glycidamide within 24 h of the last treatment (1.7- to 3.3-fold higher than control). Both doses of acrylamide and glycidamide significantly elevated lymphocyte hprt mutant frequencies in both male and female mice 21 days following the last treatments, with the high doses producing responses 16- to 25-fold higher than those of the respective controls. The high doses of acrylamide and glycidamide significantly increased liver cII mutant frequency in both males and females 21 days post-treatment (2- to 2.5-fold higher than the respective controls) ( 49 ). The more pronounced mutagenic response induced by both acrylamide and glycidamide in the hprt gene relative to the cII transgene can be ascribed to the greater number of mutable sites as well as a lower spontaneous mutant frequency rate (∼10-fold) in the former gene ( 55 ). Unlike the hprt gene whose mutation analysis can only be performed on limited organs, i.e. bone marrow and peripheral blood, the cII mutation analysis can, however, be done in various target and non-target organs without any restriction ( 55 ).

Acrylamide-induced tumorigenesis

Acrylamide has shown a characteristic organ-specific tumorigenicity in both mice and rats, although with distinct site specificity in each species ( 1 , 2 , 4 ). Mid/long-term carcinogenicity experiments in rodents have documented that administration of acrylamide through various routes, e.g. orally, topically and systemically, can increase the incidence of lung and skin adenomas and carcinomas in mice, as well as induce scrotal mesotheliomas (malignant), thyroid adenomas and/or adenocarcinoma (benign and malignant), mammary gland tumors (benign and malignant), uterine adenocarcinomas (malignant), central nervous system tumors (malignant), clitoral gland adenomas (benign) and oral papillomas (benign) in rats ( 1 , 2 , 4 ). However, the site-specific tumorigenicity of acrylamide in both species has not been accounted for by any underlying mechanisms, and remains to be elucidated.

Acrylamide tumorigenicity in mice

Acrylamide initiated skin tumorigenesis in two different strains of mice, i.e. Sencar and the ICR–Swiss mice ( 56 , 57 ). Using an ‘initiation–promotion protocol’, acrylamide administered via various routes, i.e. oral gavage, i.p. and topically (up to 50 mg/kg/body wt triweekly for 2 weeks) followed by a ‘promotion’ regimen of 12- O -tetradecanoyl-phorbol-13-acetate for 20 weeks, induced significantly papillomas and squamous cell carcinomas of the skin in a dose-dependent fashion in both strains ( 56 , 57 ). The greatest initiating effect of acrylamide in both the Sencar and ICR–Swiss mice was observed when it was administered orally ( 56 , 57 ). In the ICR–Swiss mice, oral administration of acrylamide, even in the absence of 12- O -tetradecanoyl-phorbol-13-acetate as the ‘promoting’ agent, also led to a significant formation of alveolar bronchiolar adenomas and carcinomas ( 57 ). Similar complete lung carcinogenicity of acrylamide was observed in the A/J mice treated with acrylamide through both oral and i.p. routes of exposure at concentrations up to 30 mg/kg/body wt thrice a week for a total period of 8 weeks ( 56 ). The A/J and ICR–Swiss mice killed at the age of 1 year and 8–9 months, respectively, had survival rates of 94 and 90%. In the respective strains, 30 out of 32 animals and 36 out of 40 animals, which received acrylamide orally, were available for autopsy without unacceptable levels of autolysis at the end of each experiment. The percentages of tumor-bearing animals in the A/J and ICR–Swiss mice after oral administration of acrylamide were significantly increased relative to solvent-treated controls, i.e. 93 versus 10% and 39 versus 11%, in the respective strains (detailed information on acrylamide tumorigenicity in the Sencar, ICR–Swiss, and A/J mice is available in Table II , Table III and Table IV , respectively) ( 56 , 57 ). Historically, the A/J mice develop spontaneously 1–2 lung tumors per animal around 1.5 years of age ( 58–60 ), whereas the ICR–Swiss mice have a much lower incidence of the respective tumors, which occur at 19–24 months of age ( 61–63 ).

Table II.

Tumor-initiating effects of acrylamide in the skin of female Sencar mice

Dose (mg/kg) Route of administration TPA  Cumulative number of tumor-bearing animals a /number of animals initiated   Number of squamous cell papillomas/number of animals examined b  Number of squamous cell carcinomas/number of animals examined b Percentage of animals bearing squamous cell carcinomas 
p.o. 2/40 0/34 0/34 
75 p.o. 12/40 3/35 2/35 5.7 
150 p.o. 23/40 8/33 7/33 21.2 
300 p.o. 30/40 11/38 6/38 15.8 
300 p.o. − 0/20 0/17 0/17 
0 c i.p. 0/40 0/35 0/35 
75 i.p. 10/40 2/38 2/38 5.2 
150 i.p. 13/40 3/36 4/36 11.1 
300 i.p. 21/40 6/35 4/35 11.4 
300 i.p. − 0/20 0/17 0/17 
0 d Topical 7/40 5/36 0/36 
75 Topical 4/40 3/38 1/38 2.6 
150 Topical 11/40 3/35 2/35 5.7 
300 Topical 18/40 2/34 3/34 8.8 
300 Topical − 0/20 0/20 0/20 
Dose (mg/kg) Route of administration TPA  Cumulative number of tumor-bearing animals a /number of animals initiated   Number of squamous cell papillomas/number of animals examined b  Number of squamous cell carcinomas/number of animals examined b Percentage of animals bearing squamous cell carcinomas 
p.o. 2/40 0/34 0/34 
75 p.o. 12/40 3/35 2/35 5.7 
150 p.o. 23/40 8/33 7/33 21.2 
300 p.o. 30/40 11/38 6/38 15.8 
300 p.o. − 0/20 0/17 0/17 
0 c i.p. 0/40 0/35 0/35 
75 i.p. 10/40 2/38 2/38 5.2 
150 i.p. 13/40 3/36 4/36 11.1 
300 i.p. 21/40 6/35 4/35 11.4 
300 i.p. − 0/20 0/17 0/17 
0 d Topical 7/40 5/36 0/36 
75 Topical 4/40 3/38 1/38 2.6 
150 Topical 11/40 3/35 2/35 5.7 
300 Topical 18/40 2/34 3/34 8.8 
300 Topical − 0/20 0/20 0/20 

Acrylamide was dissolved in distilled deionized water for the p.o. and i.p. routes and in ethanol for topical application. Female Sencar mice (6–8 weeks of age) were treated with acrylamide at doses of 12.5, 25.0 or 50.0 mg/kg/body wt for six applications over a 2 week period for total doses of 75, 150 or 300 mg/kg/body wt, respectively, via different routes, i.e. gastric intubation (p.o.), i.p. injection and topical administration to the shaved back of the mice. Two weeks following the tumor-initiating doses of acrylamide, a tumor promotion regimen was begun, wherein 1.0 μg 12- O -tetradecanoyl-phorbol-13-acetate (TPA) dissolved in 0.2 ml acetone was applied to the shaved back of each animal three times per week for 20 weeks. Tumor incidences were charted from weekly observations. All animals were killed at 52 weeks in the study, and histopathological evaluations were performed on all gross lesions. Data were obtained from ref. ( 56 ).

a

To be included in the cumulative count, an animal must have had a tumor of >1 mm diameter at the same location for three consecutive observations.

b

Number of animals available for histopathological examination following death or at termination of the experiment at 1 year (excluding heavily autolyzed and cannibalized animals).

c

Distilled deionized water was administered in the same volume (0.2 ml/mouse) and at the same frequency as administered to experimental groups.

d

Ethanol was administered topically in the same volume (0.2 ml/mouse) and at the same frequency as administered to experimental groups.

Table III.

Tumorigenicity of acrylamide in the skin and lung of female ICR–Swiss mice

Dose (mg/kg) TPA Skin Lung 
Cumulative number of tumor-bearing animals a /total number of treated animals  Cumulative number of tumors/animal Histological classification of squamous cell tumors at autopsy Number of animals available for autopsy Histological classification of alveolar bronchiolar tumors 
Papilloma Carcinoma Adenoma Carcinoma Total 
0 b 0/40 (35) 36 
75 4/40 (34) 0.10 34 
150 4/40 (32) 0.13 36 
300 13/40 (32) 0.43 34 10 11 
300 − 10/40 (33) 0.03 36 10 14 
Dose (mg/kg) TPA Skin Lung 
Cumulative number of tumor-bearing animals a /total number of treated animals  Cumulative number of tumors/animal Histological classification of squamous cell tumors at autopsy Number of animals available for autopsy Histological classification of alveolar bronchiolar tumors 
Papilloma Carcinoma Adenoma Carcinoma Total 
0 b 0/40 (35) 36 
75 4/40 (34) 0.10 34 
150 4/40 (32) 0.13 36 
300 13/40 (32) 0.43 34 10 11 
300 − 10/40 (33) 0.03 36 10 14 

Female ICR–Swiss mice were administered acrylamide orally by gavage at doses of 12.5, 25.0 or 50.0 mg/kg/body wt for six applications over a 2 week period for total doses of 75, 150 or 300 mg/kg/body wt, respectively. Two weeks following the tumor-initiating doses of acrylamide, a tumor promotion regimen was begun, wherein 2.5 μg 12- O -tetradecanoyl-phorbol-13-acetate (TPA) dissolved in 0.2 ml acetone was applied to the shaved back of each animal three times per week for 20 weeks. Tumor incidences were charted from weekly observations. All animals were killed at 52 weeks in the study, and histopathological evaluations were performed on all gross lesions. Data were obtained from ref. ( 57 ).

a

To be included in the cumulative count, an animal must have had a tumor of >1 mm diameter at the same location for three consecutive observations.

b

Distilled deionized water was administered in the same volume (0.2 ml/mouse) and at the same frequency as administered to experimental groups.

Table IV.

Acrylamide-induced lung adenomas in A/J mice

Dose (mg/kg) Sex Number of surviving mice/initial number of mice  Percentage of mice with lung adenomas b  Average number of lung adenomas/mouse c 
NT a 16/16 31 0.31 ± 0.48 
14/16 50 0.50 ± 0.52 
M + F 30/32 40 0.40 ± 0.50 
0 b 16/16 13 0.06 ± 0.25 
15/16 0.13 ± 0.35 
M + F 31/32 10 0.10 ± 0.30 
24 16/16 50 0.75 ± 0.93 
17/17 35 0.35 ± 0.60 
M + F 33/33 42 0.55 ± 0.78 
72 16/16 38 0.69 ± 1.03 
17/17 53 0.88 ± 1.11 
M + F 33/33 46 0.79 ± 1.05 
240 17/17 59 0.88 ± 0.99 
14/15 79 1.57 ± 1.79 
M + F 31/32 68 1.19 ± 1.40 
720 15/16 93 1.87 ± 1.55 
15/16 93 2.53 ± 1.46 
M + F 30/32 93 2.20 ± 1.52 
Dose (mg/kg) Sex Number of surviving mice/initial number of mice  Percentage of mice with lung adenomas b  Average number of lung adenomas/mouse c 
NT a 16/16 31 0.31 ± 0.48 
14/16 50 0.50 ± 0.52 
M + F 30/32 40 0.40 ± 0.50 
0 b 16/16 13 0.06 ± 0.25 
15/16 0.13 ± 0.35 
M + F 31/32 10 0.10 ± 0.30 
24 16/16 50 0.75 ± 0.93 
17/17 35 0.35 ± 0.60 
M + F 33/33 42 0.55 ± 0.78 
72 16/16 38 0.69 ± 1.03 
17/17 53 0.88 ± 1.11 
M + F 33/33 46 0.79 ± 1.05 
240 17/17 59 0.88 ± 0.99 
14/15 79 1.57 ± 1.79 
M + F 31/32 68 1.19 ± 1.40 
720 15/16 93 1.87 ± 1.55 
15/16 93 2.53 ± 1.46 
M + F 30/32 93 2.20 ± 1.52 

Male (M) and female (F) A/J mice (8 weeks old) were treated i.p. with acrylamide at doses of 1, 3, 10 or 30 mg/kg/body wt three times per week for 8 weeks for total doses of 24, 72, 240 or 720 mg/kg/body wt, respectively. Two base-line control groups were included: an untreated group (NT a ), and a vehicle control group receiving solvent distilled deionized water (0 b ). Animals were killed at 8 months of age, their lungs were fixed in Tellyesniczky's solution and the surface adenomas were counted after 24 h. Data were obtained from ref. ( 56 ).

c

Mean ± SD.

On the basis of neurotoxicity and carcinogenicity data, it is now established that acrylamide causes appreciable toxicity when it is administered at doses exceeding 50 mg/kg/body wt ( 1 , 2 , 4 ). Most recently, a mutagenicity experiment in Big Blue mice ( 49 ) has shown that animals receiving acrylamide in drinking water at a concentration of 100 mg/l, which yielded an average daily dose of 19–25 mg/kg/body wt, exhibited no signs of toxicity after 4 weeks of continuous dosing followed by an additional post-treatment period of 3 weeks. Acrylamide at a concentration of 500 mg/l, which resulted in an average daily dose of 98–107 mg/kg/body wt, caused neurotoxicity after 3 weeks of continuous dosing. However, cessation of the treatment ameliorated the neurological symptoms and the mice recovered thereafter ( 49 ).

DNA adduct-targeted mutagenicity of acrylamide

Repair resistance and promutagenicity of a DNA adduct determine its biological significance ( 55 ). The predominant DNA adduct induced by acrylamide, i.e. N7-GA-Gua, and one of its two minor adducts, i.e. 3-(2-carbamoyl-2-hydroxyethyl)-adenine, are both depurinating adducts, capable of generating apurinic/apyrimidinic (abasic) sites ( 64 , 65 ). The produced apurinic/apyrimidinic (abasic) sites are likely to give rise to 2′-deoxyadenosine incorporation during DNA replication, thereby leading to G → T transversions ( 65–68 ). Other minor adduct of acrylamide, i.e. N1-GA-Ade, has also significant miscoding potential, thus being highly promutagenic ( 64 ). We have documented a persistence of induced DNA adducts at specific locations along the cII transgene in Big Blue mouse embryonic fibroblasts treated in vitro with acrylamide or glycidamide ( 69 ). We have also demonstrated a characteristic G → T transversion mutations in the cII transgene induced by glycidamide in the same model system ( Figure 3a and b ) ( 69 ). The signature mutation of G → T transversions was not, however, as pronounced in the spectrum of mutations induced by acrylamide, although its frequency was still elevated relative to control ( 70 ).

Fig. 3.

Mutation spectrometry data of glycidamide in vitro and in vivo in the cII transgene in Big Blue mouse system [data are adopted from ( 69 ) and ( 49 ), respectively]. ( a , b ) Early stage Big Blue mouse embryonic fibroblasts were treated in vitro with 500 μM glycidamide or control solvent (double-distilled water). Mutations were quantified 8 days after treatment using the lambda select- cII mutation detection system for Big Blue rodents (Stratagene, La Jolla, CA). Verified mutant plaques were subsequently subjected to DNA sequence analysis. All jackpot mutations found in control-treated cells, i.e. G insertion/deletion at nucleotide positions 179–184, G → A transition at nucleotide position 196, G → C transversion at nucleotide position 211 and T → G transversion at nucleotide position 221, were excluded from the analyses. Total numbers of sequenced plaques were 134 and 173 from glycidamide-treated and control samples, respectively. Ins, insertion; Del, deletion. ( cf ) Male and female adult Big Blue mice were treated in vivo with glycidamide in drinking water at a concentration of 600 mg/l, which yielded an average daily dose of 88–111 mg/kg/body wt, for up to 4 weeks. Mutations in liver cII were determined 21 days post-treatment. The cII mutant plaques were verified and subjected to DNA sequence analysis. Total numbers of sequenced plaques were 75 and 58 from glycidamide-treated and control samples, respectively. Because there was no significant difference in the spectra of mutations between males and females within each group ( P ≥ 0.67), the data were combined in the respective groups ( 49 ). For comparability purposes, the in vivo data are presented both with (c–d) and without (e–f) the exclusion of the above-mentioned jackpot mutations.

Fig. 3.

Mutation spectrometry data of glycidamide in vitro and in vivo in the cII transgene in Big Blue mouse system [data are adopted from ( 69 ) and ( 49 ), respectively]. ( a , b ) Early stage Big Blue mouse embryonic fibroblasts were treated in vitro with 500 μM glycidamide or control solvent (double-distilled water). Mutations were quantified 8 days after treatment using the lambda select- cII mutation detection system for Big Blue rodents (Stratagene, La Jolla, CA). Verified mutant plaques were subsequently subjected to DNA sequence analysis. All jackpot mutations found in control-treated cells, i.e. G insertion/deletion at nucleotide positions 179–184, G → A transition at nucleotide position 196, G → C transversion at nucleotide position 211 and T → G transversion at nucleotide position 221, were excluded from the analyses. Total numbers of sequenced plaques were 134 and 173 from glycidamide-treated and control samples, respectively. Ins, insertion; Del, deletion. ( cf ) Male and female adult Big Blue mice were treated in vivo with glycidamide in drinking water at a concentration of 600 mg/l, which yielded an average daily dose of 88–111 mg/kg/body wt, for up to 4 weeks. Mutations in liver cII were determined 21 days post-treatment. The cII mutant plaques were verified and subjected to DNA sequence analysis. Total numbers of sequenced plaques were 75 and 58 from glycidamide-treated and control samples, respectively. Because there was no significant difference in the spectra of mutations between males and females within each group ( P ≥ 0.67), the data were combined in the respective groups ( 49 ). For comparability purposes, the in vivo data are presented both with (c–d) and without (e–f) the exclusion of the above-mentioned jackpot mutations.

Manjanatha et al. ( 49 ) have recently shown that Big Blue mice treated in vivo with acrylamide or equimolar doses of glycidamide through drinking water have similar induced mutation spectra in the cII transgene, both spectra being significantly different from that of the control ( 49 ). The predominant types of cII mutations induced by both acrylamide and glycidamide were G:C → T:A transversions and −1/+1 frameshifts ( Figure 3c and d ). The majority of these frameshift mutations were ‘jackpot’ mutations, particularly, G insertions/deletions at nucleotide positions 179–184. These jackpot mutations are common phenomena in transgenic rodents, and are thought to occur during early stages of development and undergo clonal expansion. Consequently, many cells from various organs of (un) treated animals may harbor the exact same mutation ( 71 ). To rule out the over-representation of jackpot mutations, it is methodologically appropriate to exclude these mutations from all spectra analyses. Excluding the jackpot mutations, the spectra of mutation induced by both acrylamide and glycidamide in vivo ( 49 ) resembled closely that produced by glycidamide in vitro ( 69 ) (see Figure 3a and e ). The similarity in mutation spectra induced in vivo by acrylamide and glycidamide, respectively, can explain our in vitro finding of the less-explicit signature mutation of G → T transversions being induced by acrylamide relative to glycidamide in early stage mouse embryonic fibroblasts with immature metabolic machinery ( 69 , 70 ). Our findings could also be explained, at least partially, by the known deficiency of P450 activity in neonatal mice ( 42 ).

Dietary acrylamide and cancer

The intentional introduction of genotoxic and carcinogenic compounds into food during manufacturing is generally prohibited by regulation ( 72 ). However, unintentional generation of process-induced contaminants can occur both at industrial and at domestic settings. Complete elimination of all these contaminants may not be feasible, especially because of the ever-increasing sensitivity of the analytical methods, which enables detection of trace amounts of specific contaminants in food matrices ( 72 ). Acrylamide like other classes of carcinogens, including heterocyclic amines, polycyclic aromatic hydrocarbons and N -nitroso compounds, can be formed during cooking or other thermal processing in a wide range of food products ( 73 ). High formation of acrylamide can occur in plant commodities rich in carbohydrates and low in proteins, as a result of heating at temperatures exceeding 120°C ( 15 , 16 ). Besides the composition of food items, i.e. content of free amino acid, e.g. asparagine, and reducing sugars, e.g. fructose and glucose, formation of acrylamide is dependent on the modality of cooking, particularly temperature and duration ( 14–16 ). As such, large variations exist in acrylamide content of different brands of the same food product as well as among various batches of the exact same brand (see Table I ).

The World Health Organization estimates a daily intake of dietary acrylamide in the range of 0.3–2.0 μg/kg/body wt for the general population ( 74 ). For high-percentile consumers (90th to 97.5th), daily intakes of dietary acrylamide vary in the range of 0.6–3.5 μg/kg/body wt, and as high as 5.1 μg/kg/body wt for the 99th-percentile consumers. The daily intakes of dietary acrylamide in children are estimated to be 2–3 times those of adults based on average body weight ratios. The daily intakes of dietary acrylamide for the general population and high consumers (including children) are estimated to be on average 1 and 4 μg/kg/body wt, respectively. The main sources of dietary acrylamide are potato chips (16–30%), potato crisps (6–46%), coffee (13–39%), pastry and sweet biscuits (10–20%) and bread and rolls/toasts (10–30%). Other food products can account for <10% of the total intake of dietary acrylamide (see Table I ) ( 74 ).

So far, several epidemiologic studies have attempted to find a link between dietary acrylamide exposure and human cancers ( 17 , 18 , 75–78 ). However, no association has been found between the intake of dietary acrylamide and the risk for development of any type of cancer ( 75 , 79–81 ). The absence of positive results in these observational studies, however, cannot be interpreted as proof of no carcinogenicity of acrylamide to humans ( 75 , 79–81 ). Obviously, the conducted studies have potential limitations, including inadequate statistical power due to the small size of study populations and the narrow range of exposure between cases and controls. By virtue of design, case–control studies require a reasonable difference in exposure levels between cases and controls; yet, acrylamide is abundantly present in foods eaten nearly universally, e.g. bread products. Also, the observational studies are prone to misclassification of acrylamide exposure as a result of reliance on crude estimates of dietary intake using food frequency questionnaires. Notwithstanding is that the administered questionnaires of some of these studies have been originally tailored to elicit information regarding exposure to carcinogens other than acrylamide, e.g. heterocyclic amines ( 17 ); at the original design of these studies, the presence of acrylamide in foods was not even discovered. Other drawbacks of these studies include selection and recall biases, especially in hospital-based case–control studies in which cases tend to modify their dietary habits soon after the appearance of early symptoms of the disease, as well as report their dietary intakes more accurately than controls ( 75 , 79–81 ).

As a default assumption, genotoxic carcinogens are often considered to have no threshold limit of exposure, i.e. a single exposure to one molecule of carcinogen can trigger the biological process leading to cancer ( 82 ). In the food safety area, such compounds are mainly dealt with by the principle of ‘as low as reasonably achievable’, as is the case for acrylamide ( 72 ). Regulatory agencies also perform (semi) quantitative risk assessment for acrylamide exposure using tumorigenicity data, which are largely available from a few mid/long-term bioassays of acrylamide in rodents ( 56 , 57 , 83 , 84 ). Inherent in the latter approach, however, are major uncertainties caused by the extrapolation of animal data at exposure levels that are 3–5 orders of magnitude higher than those to which humans are generally exposed ( 74 ). The growing body of evidence on inter-species differences in acrylamide bioavailability and metabolism and sensitivity to tumor induction at specific organ sites also casts shadow on extrapolating data from rodents to humans ( 74 ). Thus far, several cancer risk assessments have been performed for dietary acrylamide using various dose–response computer modelings, which have led to different carcinogenic potency estimates ( 82 ). For example, utilizing a linearized multi-stage model, which does not account for inter-species variabilities, a lifelong daily intake of 1 μg acrylamide has been estimated to lead to a lifetime cancer risk of 1 × 10 −5 ( 85 ). Another model based on the scaling of dose per unit of body surface area has estimated a corresponding risk of 6 × 10 −5 ( 86 ).

Comparing the average daily intake of 1 μg/kg/body wt of dietary acrylamide by the general population and a dose of 0.30 mg/kg body wt/day for induction of mammary tumors in rats, the margin of exposure for dietary acrylamide in humans is 300 ( 74 ). The respective margin of exposure for high-consumer groups with a daily intake of 4 μg/kg/body wt of dietary acrylamide is 75 ( 74 ). The estimated margin of exposures are considered to be low for a genotoxic and carcinogenic substance that is so omnipresent in human food chain ( 74 ). Admittedly, there are still significant knowledge gaps, which currently preclude a more reliable estimate of the cancer risk due to acrylamide intake via foods. Nevertheless, the significant presence of a known rodent carcinogen in commonly consumed human foods is a legitimate health concern. It is recommended that appropriate efforts be continued to minimize human exposure to acrylamide, especially through reducing acrylamide concentrations in frequently consumed food products ( 74 ). For example, efficient reduction in acrylamide formation has been achieved by using the enzyme asparaginase to selectively remove asparagine prior to heating of cereal and potato products ( 74 ). However, this approach is only applicable to certain foods prepared from liquidized or slurried materials. Formulation changes such as increasing the concentrations of other amino acids or proteins, which compete with asparagine in the Maillard reaction or react with formed acrylamide, may also lower the content of acrylamide in foods ( 74 ). Genetic or chemical manipulation of enzymes involved in acrylamide biotransformation, e.g. CYP2E1 induction or inhibition (by flavonoids), is also an important area of research, which awaits further exploration.

Concluding remarks

The provocative announcement that acrylamide, a proven rodent carcinogen ( 87 ), is found in a variety of commonly consumed human foods has raised public health concerns ( 14–16 ). Thus far, the involvement of dietary acrylamide in human cancers has been investigated using conventional exposure assessment in relation to cancer incidence rating ( 19 ). However, such questionnaire/interview-based investigations are prone to inconclusiveness because of the ubiquity of acrylamide in the human diet and its variable generation even within the same individual food ( 4 ). In addition, the presence of acrylamide in the environment, i.e. in occupational settings or in the ambient air, e.g. as a constituent of environmental tobacco smoke ( 88 ), confounds the profile of human exposure to acrylamide.

A more reliable approach is to monitor well-defined human populations using biological markers, which are both specific for acrylamide and pertinent to cancer ( 19 ). This requires an understanding of the as yet unknown mechanism of acrylamide carcinogenicity. The underlying mechanism of acrylamide carcinogenicity can be resolved by investigating a multi-stage continuum that starts with exposure to acrylamide and ends with tumorigenesis ( 19 ). Of significance in this continuum are acrylamide-induced DNA adduction and mutagenesis, as well as possible epigenetic alterations ( 19 ). Many chemical carcinogens display a characteristic DNA adduct-targeted mutagenicity in cancer-related genes, i.e. oncogenes and tumor suppressor genes ( 55 ). Also, a range of environmental carcinogens can cause heritable epigenetic modifications to the DNA, thus changing gene expression while the primary DNA sequence being conserved ( 89–96 ). To date, the engagement of an epigenetic pathway in acrylamide-induced carcinogenesis is unclear, and remains to be determined. However, substantial data are available on the genotoxic mode of action of acrylamide of relevance for carcinogenesis ( 4 , 32 ).

So far, all available data point to an involvement of glycidamide in acrylamide-induced DNA adduction and mutagenesis ( 4 , 32 ). However, the interrelated chain of events starting from glycidamide–DNA adduct formation and leading to mutation has not been directly investigated. The correlative nature of events contributing to acrylamide-induced mutagenesis and tumorigenesis is also another area of research, which remains to be investigated. These investigations should help unravel the underlying mechanism of action of acrylamide carcinogenicity. Increasing the mechanistic knowledge of acrylamide-induced carcinogenesis will help identify biological markers that can be used for monitoring of human exposure to acrylamide, in particular, to determine a possible link between dietary acrylamide and human cancers.

Abbreviations

    Abbreviations
  • CYP2E1

    cytochrome P450 2E1

  • GSH

    glutathione

  • Hb

    hemoglobin

  • hprt

    hypoxanthine-guanine phosphoribosyl transferase

  • i.p.

    intra-peritoneal

  • N7-GA-Gua

    7-(2-carbamoyl-2-hydroxyethyl)-guanine

Work of the authors is supported by National Institutes of Health grant CA84469.

Conflict of Interest Statement: None declared.

References

1.
Dearfield
KL
, et al.  . 
Acrylamide: its metabolism, developmental and reproductive effects, genotoxicity, and carcinogenicity
Mutat. Res.
 , 
1988
, vol. 
195
 (pg. 
45
-
77
)
2.
Dearfield
KL
, et al.  . 
Acrylamide: a review of its genotoxicity and an assessment of heritable genetic risk
Mutat. Res.
 , 
1995
, vol. 
330
 (pg. 
71
-
99
)
3.
McCollister
DD
, et al.  . 
Toxicology of acrylamide
Toxicol. Appl. Pharmacol.
 , 
1964
, vol. 
103
 (pg. 
172
-
181
)
4.
Friedman
M
Chemistry, biochemistry, and safety of acrylamide. A review
J. Agr. Food Chem.
 , 
2003
, vol. 
51
 (pg. 
4504
-
4526
)
5.
Collins
JJ
, et al.  . 
Mortality patterns among workers exposed to acrylamide
J. Occup. Med.
 , 
1989
, vol. 
31
 (pg. 
614
-
617
)
6.
Granath
F
, et al.  . 
Cancer risk from exposure to occupational acrylamide
Occup. Environ. Med.
 , 
2001
, vol. 
58
 (pg. 
608
-
609
)
7.
Marsh
GM
, et al.  . 
Mortality patterns among workers exposed to acrylamide: 1994 follow up
Occup. Environ. Med.
 , 
1999
, vol. 
56
 (pg. 
181
-
190
)
8.
Reynolds
T
Acrylamide and cancer: tunnel leak in Sweden prompted studies
J. Natl Cancer Inst.
 , 
2002
, vol. 
94
 (pg. 
876
-
878
)
9.
Tornqvist
M
, et al.  . 
Leakage of Acrylamides from a Tunnel Construction Work: Exposure Monitoring and Health Effects to Humans and Animals
 , 
2000
Belgium
Antwerp
(pg. 
1
-
8
European ALARA Network
10.
Godin
AC
, et al.  . 
Acrylamide and N-methylolacrylamide poisoning in a herd of Charolais crossbreed cattle
Vet. Rec.
 , 
2002
, vol. 
151
 (pg. 
724
-
728
)
11.
Hagmar
L
, et al.  . 
Health effects of occupational exposure to acrylamide using hemoglobin adducts as biomarkers of internal dose
Scand. J. Work Environ. Health
 , 
2001
, vol. 
27
 (pg. 
219
-
226
)
12.
Bergmark
E
Hemoglobin adducts of acrylamide and acrylonitrile in laboratory workers, smokers and nonsmokers
Chem. Res. Toxicol.
 , 
1997
, vol. 
10
 (pg. 
78
-
84
)
13.
Tareke
E
, et al.  . 
Acrylamide: a cooking carcinogen?
Chem. Res. Toxicol.
 , 
2000
, vol. 
13
 (pg. 
517
-
522
)
14.
Tareke
E
, et al.  . 
Analysis of acrylamide, a carcinogen formed in heated foodstuffs
J. Agric. Food Chem.
 , 
2002
, vol. 
50
 (pg. 
4998
-
5006
)
15.
Stadler
RH
, et al.  . 
Acrylamide from Maillard reaction products
Nature
 , 
2002
, vol. 
419
 (pg. 
449
-
450
)
16.
Mottram
DS
, et al.  . 
Acrylamide is formed in the Maillard reaction
Nature
 , 
2002
, vol. 
419
 (pg. 
448
-
449
)
17.
Mucci
LA
, et al.  . 
Dietary acrylamide and cancer of the large bowel, kidney, and bladder: absence of an association in a population-based study in Sweden
Br. J. Cancer
 , 
2003
, vol. 
88
 (pg. 
84
-
89
)
18.
Pelucchi
C
, et al.  . 
Fried potatoes and human cancer
Int. J. Cancer
 , 
2003
, vol. 
105
 (pg. 
558
-
560
)
19.
Besaratinia
A
, et al.  . 
DNA adduction and mutagenic properties of acrylamide
Mutat. Res.
 , 
2005
, vol. 
580
 (pg. 
31
-
40
)
20.
Sumner
SC
, et al.  . 
Role of cytochrome P450 2E1 in the metabolism of acrylamide and acrylonitrile in mice
Chem. Res. Toxicol.
 , 
1999
, vol. 
12
 (pg. 
1110
-
1116
)
21.
Srivastava
S
, et al.  . 
Effect of single and repeated doses of acrylamide and bis-acrylamide on glutathione-S-transferase and dopamine receptors in rat brain
Brain Res.
 , 
1986
, vol. 
371
 (pg. 
319
-
323
)
22.
Abate
C
, et al.  . 
Redox regulation of fos and jun DNA-binding activity in vitro
Science
 , 
1990
, vol. 
249
 (pg. 
1157
-
1161
)
23.
Banerjee
S
, et al.  . 
In vitro transformation of C3H/10T1/2 and NIH/3T3 cells by acrylonitrile and acrylamide
Cancer Lett.
 , 
1986
, vol. 
32
 (pg. 
293
-
304
)
24.
Schulze-Osthoff
K
, et al.  . 
Redox signalling by transcription factors NF-kappa B and AP-1 in lymphocytes
Biochem. Pharmacol.
 , 
1995
, vol. 
50
 (pg. 
735
-
741
)
25.
Tsuda
H
, et al.  . 
Acrylamide; induction of DNA damage, chromosomal aberrations and cell transformation without gene mutations
Mutagenesis
 , 
1993
, vol. 
8
 (pg. 
23
-
29
)
26.
Agrawal
AK
, et al.  . 
Neurotransmitter receptors in brain regions of acrylamide-treated rats. I: effects of a single exposure to acrylamide
Pharmacol. Biochem. Behav.
 , 
1981
, vol. 
14
 (pg. 
527
-
731
)
27.
Agrawal
AK
, et al.  . 
The effects of acrylamide treatment upon the dopamine receptor
Toxicol. Appl. Pharmacol.
 , 
1981
, vol. 
58
 (pg. 
89
-
99
)
28.
Ali
SF
, et al.  . 
Effect of acrylamide on neurotransmitter metabolism and neuropeptide levels in several brain regions and upon circulating hormones
Arch. Toxicol.
 , 
1983
, vol. 
52
 (pg. 
35
-
43
)
29.
Bolt
HM
Genotoxicity—threshold or not? Introduction of cases of industrial chemicals
Toxicol. Lett.
 , 
2003
, vol. 
140–141
 (pg. 
43
-
51
)
30.
European Commission, European Chemicals Bureau Institute for Health and Consumer Protection
European Union Risk Assessment Report Acrylmaide
 , 
2002
Luxembourg, UK
Office for Official Publications of the European Communities
31.
Tyl
RW
, et al.  . 
Effects of acrylamide on rodent reproductive performance
Reprod. Toxicol.
 , 
2003
, vol. 
17
 (pg. 
1
-
13
)
32.
Rice
JM
The carcinogenicity of acrylamide
Mutat Res.
 , 
2005
, vol. 
580
 (pg. 
3
-
20
)
33.
Sorgel
F
, et al.  . 
Acrylamide: increased concentrations in homemade food and first evidence of its variable absorption from food, variable metabolism and placental and breast milk transfer in humans
Chemotherapy
 , 
2002
, vol. 
48
 (pg. 
267
-
274
)
34.
Fuhr
U
, et al.  . 
Toxicokinetics of acrylamide in humans after ingestion of a defined dose in a test meal to improve risk assessment for acrylamide carcinogenicity
Cancer Epidemiol. Biomarkers Prev.
 , 
2006
, vol. 
15
 (pg. 
266
-
271
)
35.
Boettcher
MI
, et al.  . 
Excretion of mercapturic acids of acrylamide and glycidamide in human urine after single oral administration of deuterium-labelled acrylamide
Arch. Toxicol.
 , 
2006
, vol. 
80
 (pg. 
55
-
61
)
36.
Fennell
TR
, et al.  . 
Metabolism and hemoglobin adduct formation of acrylamide in humans
Toxicol. Sci.
 , 
2005
, vol. 
85
 (pg. 
447
-
459
)
37.
Schettgen
T
, et al.  . 
Trans-placental exposure of neonates to acrylamide—a pilot study
Int. Arch. Occup. Environ. Health
 , 
2004
, vol. 
77
 (pg. 
213
-
216
)
38.
Sumner
SC
, et al.  . 
Urinary metabolites from F344 rats and B6C3F1 mice coadministered acrylamide and acrylonitrile for 1 or 5 days
Chem. Res. Toxicol.
 , 
1997
, vol. 
10
 (pg. 
1152
-
1160
)
39.
Paulsson
B
, et al.  . 
The multiplicative model for cancer risk assessment: applicability to acrylamide
Carcinogenesis
 , 
2001
, vol. 
22
 (pg. 
817
-
819
)
40.
Doerge
DR
, et al.  . 
Toxicokinetics of acrylamide and glycidamide in B6C3F1 mice
Toxicol. Appl. Pharmacol.
 , 
2005
, vol. 
202
 (pg. 
258
-
267
)
41.
Gamboa da Costa
G
, et al.  . 
DNA adduct formation from acrylamide via conversion to glycidamide in adult and neonatal mice
Chem. Res. Toxicol.
 , 
2003
, vol. 
16
 (pg. 
1328
-
1337
)
42.
Fu
PP
, et al.  . 
Metabolic activation capacity of neonatal mice in relation to the neonatal mouse tumorigenicity bioassay
Drug Metab. Rev.
 , 
2000
, vol. 
32
 (pg. 
241
-
266
)
43.
Solomon
JJ
, et al.  . 
Direct alkylation of 2′-deoxynucleosides and DNA following in vitro reaction with acrylamide
Cancer Res.
 , 
1985
, vol. 
45
 (pg. 
3465
-
3470
)
44.
Segerback
D
, et al.  . 
Formation of N-7-(2-carbamoyl-2-hydroxyethyl)guanine in DNA of the mouse and the rat following intraperitoneal administration of [14C]acrylamide
Carcinogenesis
 , 
1995
, vol. 
16
 (pg. 
1161
-
1165
)
45.
Hashimoto
K
, et al.  . 
Mutagenicity of acrylamide and its analogues in Salmonella typhimurium
Mutat. Res.
 , 
1985
, vol. 
158
 (pg. 
129
-
133
)
46.
Ghanayem
BI
, et al.  . 
Comparison of germ cell mutagenicity in male CYP2E1-null and wild-type mice treated with acrylamide: evidence supporting a glycidamide-mediated effect
Biol. Reprod.
 , 
2005
, vol. 
72
 (pg. 
157
-
163
)
47.
Paulsson
B
, et al.  . 
Hemoglobin adducts and micronucleus frequencies in mouse and rat after acrylamide or N-methylolacrylamide treatment
Mutat. Res.
 , 
2002
, vol. 
516
 (pg. 
101
-
111
)
48.
Schmid
TE
, et al.  . 
Detection of aneuploidy by multicolor FISH in mouse sperm after in vivo treatment with acrylamide, colchicine, diazepam or thiabendazole
Mutagenesis
 , 
1999
, vol. 
14
 (pg. 
173
-
179
)
49.
Manjanatha
MG
, et al.  . 
Genotoxicity of acrylamide and its metabolite glycidamide administered in drinking water to male and female Big Blue mice
Environ. Mol. Mutagen.
 , 
2006
, vol. 
47
 (pg. 
6
-
17
)
50.
Russell
LB
, et al.  . 
Induction of specific-locus mutations in male germ cells of the mouse by acrylamide monomer
Mutat. Res.
 , 
1991
, vol. 
262
 (pg. 
101
-
107
)
51.
Neuhauser-Klaus
A
, et al.  . 
Mutagenic and teratogenic effects of acrylamide in the mammalian spot test
Mutat. Res.
 , 
1989
, vol. 
226
 (pg. 
157
-
162
)
52.
Hoorn
AJ
, et al.  . 
Detection of chemical mutagens using Muta mouse: a transgenic mouse model
Mutagenesis
 , 
1993
, vol. 
8
 (pg. 
7
-
10
)
53.
Adler
ID
, et al.  . 
1-Aminobenzotriazole inhibits acrylamide-induced dominant lethal effects in spermatids of male mice
Mutagenesis
 , 
2000
, vol. 
15
 (pg. 
133
-
136
)
54.
Ghanayem
BI
, et al.  . 
Absence of acrylamide-induced genotoxicity in CYP2E1-null mice: evidence consistent with a glycidamide-mediated effect
Mutat. Res.
 , 
2005
, vol. 
578
 (pg. 
284
-
297
)
55.
Besaratinia
A
, et al.  . 
Investigating human cancer etiology by DNA lesion footprinting and mutagenicity analysis
Carcinogenesis
 , 
2006
, vol. 
27
 (pg. 
1526
-
1537
)
56.
Bull
RJ
, et al.  . 
Carcinogenic effects of acrylamide in Sencar and A/J mice
Cancer Res.
 , 
1984
, vol. 
44
 (pg. 
107
-
111
)
57.
Bull
RJ
, et al.  . 
Carcinogenic activity of acrylamide in the skin and lung of Swiss-ICR mice
Cancer Lett.
 , 
1984
, vol. 
24
 (pg. 
209
-
212
)
58.
Koizumi
A
, et al.  . 
Energy restriction that inhibits cellular proliferation by torpor can decrease susceptibility to spontaneous and asbestos-induced lung tumors in A/J mice
Lab. Invest.
 , 
1993
, vol. 
68
 (pg. 
728
-
739
)
59.
Shimkin
MB
, et al.  . 
Lung tumors in mice: application to carcinogenesis bioassay
Adv. Cancer Res.
 , 
1975
, vol. 
21
 (pg. 
1
-
58
)
60.
Stoner
GD
, et al.  . 
Strain A mouse lung tumor bioassay
J. Am. Coll. Toxicol.
 , 
1982
, vol. 
1
 (pg. 
145
-
169
)
61.
Derelanko
MJ
, et al.  . 
Derelanko
MJ
Hollinger
MA
Carcinogenesis
Handbook of Toxicology
 , 
2001
New York, NY
CRC Press
(pg. 
627
-
646
)
62.
Li
EE
, et al.  . 
Relationships of DNA adduct formation, K-ras activating mutations and tumorigenic activities of 6-nitrochrysene and its metabolites in the lungs of CD-1 mice
Carcinogenesis
 , 
1994
, vol. 
15
 (pg. 
1377
-
1385
)
63.
Manenti
G
, et al.  . 
Outbred CD-1 mice carry the susceptibility allele at the pulmonary adenoma susceptibility 1 (Pas1) locus
Carcinogenesis
 , 
2003
, vol. 
24
 (pg. 
1143
-
1148
)
64.
Koskinen
M
, et al.  . 
Specific DNA adducts induced by some mono-substituted epoxides in vitro and in vivo
Chem. Biol. Interact.
 , 
2000
, vol. 
129
 (pg. 
209
-
229
)
65.
Loeb
LA
, et al.  . 
Mutagenesis by apurinic/apyrimidinic sites
Annu. Rev. Genet.
 , 
1986
, vol. 
20
 (pg. 
201
-
230
)
66.
Glaab
WE
, et al.  . 
Specificity of mutations induced by methyl methanesulfonate in mismatch repair-deficient human cancer cell lines
Mutat. Res.
 , 
1999
, vol. 
427
 (pg. 
67
-
78
)
67.
Haracska
L
, et al.  . 
Roles of yeast DNA polymerases delta and zeta and of Rev1 in the bypass of abasic sites
Genes Dev.
 , 
2001
, vol. 
15
 (pg. 
945
-
954
)
68.
Zhao
B
, et al.  . 
Role of DNA polymerase eta in the bypass of abasic sites in yeast cells
Nucleic Acids Res.
 , 
2004
, vol. 
32
 (pg. 
3984
-
3994
)
69.
Besaratinia
A
, et al.  . 
Genotoxicity of acrylamide and glycidamide
J. Natl Cancer Inst.
 , 
2004
, vol. 
96
 (pg. 
1023
-
1029
)
70.
Besaratinia
A
, et al.  . 
Weak yet distinct mutagenicity of acrylamide in mammalian cells
J. Natl Cancer Inst.
 , 
2003
, vol. 
95
 (pg. 
889
-
896
)
71.
Lambert
IB
, et al.  . 
Detailed review of transgenic rodent mutation assays
Mutat. Res.
 , 
2005
, vol. 
590
 (pg. 
1
-
280
)
72.
Tritscher
AM
Human health risk assessment of processing-related compounds in food
Toxicol. Lett.
 , 
2004
, vol. 
149
 (pg. 
177
-
186
)
73.
Jagerstad
M
, et al.  . 
Genotoxicity of heat-processed foods
Mutat. Res.
 , 
2005
, vol. 
574
 (pg. 
156
-
172
)
74.
World Health Organization
Summary Report of the Sixty-Fourth Meeting of the Joint FAO/WHO Expert Committee on Food Additive (JECFA)
 , 
2005
(pg. 
1
-
47
Rome, Italy. The ILSI Press International Life Sciences Institute, Washington DC.
75.
Hagmar
L
, et al.  . 
Inconclusive results from an epidemiological study on dietary acrylamide and cancer
Br. J. Cancer
 , 
2003
, vol. 
89
 (pg. 
774
-
776
)
76.
Mucci
LA
, et al.  . 
Prospective study of dietary acrylamide and risk of colorectal cancer among women
Int. J. Cancer
 , 
2006
, vol. 
118
 (pg. 
169
-
173
)
77.
Mucci
LA
, et al.  . 
Dietary acrylamide and risk of renal cell cancer
Int. J. Cancer
 , 
2004
, vol. 
109
 (pg. 
774
-
776
)
78.
Pelucchi
C
, et al.  . 
Dietary acrylamide and human cancer
Int. J. Cancer
 , 
2006
, vol. 
118
 (pg. 
467
-
471
)
79.
Erdreich
LS
, et al.  . 
Epidemiologic evidence for assessing the carcinogenicity of acrylamide
Regul. Toxicol. Pharmacol.
 , 
2004
, vol. 
39
 (pg. 
150
-
157
)
80.
Mucci
LA
, et al.  . 
The role of epidemiology in understanding the relationship between dietary acrylamide and cancer risk in humans
Adv. Exp. Med. Biol.
 , 
2005
, vol. 
561
 (pg. 
39
-
47
)
81.
Petersen
BJ
, et al.  . 
Exposure to acrylamide: placing exposure in context
Adv. Exp. Med. Biol.
 , 
2005
, vol. 
561
 (pg. 
63
-
76
)
82.
Felsot
AS
Acrylamide Angst: Another Annoying Distraction about Food Safety Agrichemical and Environmental News: A Monthly Report on Environmental and Pesticide Related Issues
 , 
2002
Washington, DC
Washington State University
83.
Johnson
KA
, et al.  . 
Chronic toxicity and oncogenicity study on acrylamide incorporated in the drinking water of Fischer 344 rats
Toxicol. Appl. Pharmacol.
 , 
1986
, vol. 
85
 (pg. 
154
-
168
)
84.
Robinson
M
, et al.  . 
A combined carcinogen bioassay utilizing both the lung adenoma and skin papilloma protocols
Environ. Health Perspect.
 , 
1986
, vol. 
68
 (pg. 
141
-
145
)
85.
World Health Organization
Guidelines for Drinking-Water Quality
 , 
1996
Geneva, Switzerland
WHO
86.
US Environmental Protection Agency
Assessment of Health Risks from Exposure to Acrylamide
 , 
1990
Washington, DC
US EPA Office of Toxic Substances
87.
International Agency for Research on Cancer
Acrylamide
 , 
1995
Lyon
IRAC Press
88.
Smith
CJ
, et al.  . 
"IARC group 2B carcinogens" reported in cigarette mainstream smoke
Food Chem. Toxicol.
 , 
2001
, vol. 
39
 (pg. 
183
-
205
)
89.
Anway
MD
, et al.  . 
Epigenetic transgenerational actions of endocrine disruptors and male fertility
Science
 , 
2005
, vol. 
308
 (pg. 
1466
-
1469
)
90.
Feinberg
AP
, et al.  . 
The history of cancer epigenetics
Nat. Rev. Cancer
 , 
2004
, vol. 
4
 (pg. 
143
-
153
)
91.
Hutt
JA
, et al.  . 
Life-span inhalation exposure to mainstream cigarette smoke induces lung cancer in B6C3F1 mice through genetic and epigenetic pathways
Carcinogenesis
 , 
2005
, vol. 
26
 (pg. 
1999
-
2009
)
92.
Lee
YW
, et al.  . 
Carcinogenic nickel silences gene expression by chromatin condensation and DNA methylation: a new model for epigenetic carcinogens
Mol. Cell Biol.
 , 
1995
, vol. 
15
 (pg. 
2547
-
2557
)
93.
Mass
MJ
, et al.  . 
Arsenic alters cytosine methylation patterns of the promoter of the tumor suppressor gene p53 in human lung cells: a model for a mechanism of carcinogenesis
Mutat. Res.
 , 
1997
, vol. 
386
 (pg. 
263
-
277
)
94.
Ray
SS
, et al.  . 
Dioxin-induced immortalization of normal human keratinocytes and silencing of p53 and p16INK4a
J. Biol. Chem.
 , 
2004
, vol. 
279
 (pg. 
27187
-
27193
)
95.
Song
S
, et al.  . 
Induction of cyclooxygenase-2 by benzo[a]pyrene diol epoxide through inhibition of retinoic acid receptor-beta 2 expression
Oncogene
 , 
2005
, vol. 
24
 (pg. 
8268
-
8276
)
96.
Ushijima
T
, et al.  . 
Aberrant methylations in cancer cells: where do they come from?
Cancer Sci.
 , 
2005
, vol. 
96
 (pg. 
206
-
211
)