Chapter 4: Biomarkers of Exposure: Hemoglobin Adducts

Chemical compounds that are electrophilically reactive or form reactive intermediates in the metabolism are potentially toxic owing to their ability to react with electron-dense atoms in proteins and DNA. The reactivity makes these compounds difficult to measure in vivo, as a result of their fast detoxification and short half-lives.

One approach used for measurement of exposure or the in vivo dose of electrophilic compounds is based on measurement of their stable adducts with biomacromolecules. The macromolecules used for this purpose are mostly DNA, serum albumin and hemoglobin (Hb) in blood. An adduct has been defined as a complex that forms when a chemical binds to a biological molecule. In the present context it has also been found practical to use the term adduct for a moiety covalently bound to the macromolecule as a consequence of a reaction (Figure 4.1).

The emphasis of this chapter is on the application of adducts with Hb as a biomarker of occupational, environmental and lifestyle exposures to carcinogenic compounds. Some applications are illustrated with examples from the authors’ own work. The usefulness of adduct studies is by no means restricted to mutagens and carcinogens. The chemical reactivity of compounds and/or of their metabolic intermediates may, even at very low concentrations, alter tissue constituents in such a way as to give rise to various harmful conditions or diseases.

As discussed later in more detail the Hb adduct level reflects the dose (AUC; area under the concentration vs. time curve) of an electrophilic compound in blood. The AUC is determined by the absorption (or formation) of the compound, distribution, metabolism and excretion (Figure 4.2).

By the middle of the twentieth century, adducts with proteins, including Hb, were used to measure the bioavailability of reactive compounds in experimental animals. Notably, studies with aromatic amines contributed to this field of research. In 1953, Jackson and Thompson¹  demonstrated that a radiolabeled derivative of phenylhydroxylamine bound strongly to Hb in erythrocytes and that this bound material was eliminated only in the course of erythrocyte degradation. The usefulness of monitoring Hb adducts from aromatic amines was later supported by studies of occupational exposure.^2,3 

In 1974, Ehrenberg and co-workers suggested that in vivo doses from exposure to electrophilic agents might be determined through the measurement of the adducts they form with tissue proteins,⁴  leading to the exploration of Hb as a monitor molecule. In experiments with mice the approach was tested with the directly reactive agent ethylene oxide, and a compound that requires metabolic activation, namely dimethylnitrosamine.⁵  The early development of analytical methods for the isolation of adducts, and for their analysis by gas chromatography–mass spectrometry (GC-MS), was important for the application of Hb adduct measurement to exposure to potentially genotoxic and cancer risk increasing agents. Adducts with Hb in humans were used for the first time by Calleman et al.⁶  to calculate doses in blood explicitly following occupational exposure to ethylene oxide.

Hemoglobin (Hb) is synthesized in the erythrocytes during their development in the bone marrow. The normal human adult Hb molecule, α₂β₂ (HbA), is a tetramer consisting of two α-chains and two β-chains, each containing a heme group. In “normal” adults hemoglobin HbA is the dominating component, making up ∼97% of the total Hb substance.

The major sites of adduct formation are cysteine-S, histidine-N, the N-terminal NH₂ group, and the carboxyl groups of aspartic and glutamic acid and of the C-terminal amino acid. Further, serine, threonine, tyrosine, lysine, arginine, methionine, and tryptophan residues may react with electrophilic compounds. Valine is the N-terminal amino acid of both the α- and the β-chain of adult Hb.

The pattern of binding of an electrophilic compound to the various nucleophilic sites in a protein may, at least to some extent, be predicted. The reactivity of nucleophilic atoms towards alkylating agents generally decreases in the order S > N > O. Alkylating agents have been assigned s-values⁷  that describe their ability to react selectively with these atoms. Alkylating agents with high s-values, such as ethylene oxide, have a strong propensity for reactions with cysteine-S. Agents with a low s-value, such as the reactive intermediates of alkylnitrosamines, show less selectively and give relatively high levels of adducts with reactive oxygen atoms.

The reactivity of a nucleophilic site in Hb is also dependent on its pK_a; the base in an acid–base equilibrium being far more reactive than the acid (reviewed by Törnqvist et al.⁸ ). This influences the rate of adduct formation. Thus, because of the low pK_a of the N-terminal amino groups in hemoglobin (6.8–7.8), histidines (5.6–7), and cysteines (7.9–8.5) these sites have a comparatively high reactivity towards, for example, alkylating agents. Lysines have high pK_a-values (9.5–12.5) and consequently have a relatively low reactivity towards this type of electrophilic compound. However, there is also a correlation between alkalinity (as measured by pK_a) and nucleophilic strength, counteracting the effect of protonation. Thus, certain types of electrophilic compound, such as acylating agents, may give relatively high levels of adducts with lysine. A majority of the applications are based on determination of adducts with cysteine and N-terminal valine.

For high-molecular-weight compounds, it may be difficult to predict the sites of reaction. It has been suggested that the tertiary structure of the protein may be playing a more important role in adduct formation for these compounds.⁹ 

Neonates: At birth, HbF (α₂γ₂) comprises a major part of the child's Hb. These levels decline and after 6 months adult Hb (α₂β₂) takes over as the predominant form of Hb in normal children. In HbF the N-terminal amino acid of the γ-chain is glycine. Such differences in Hb should be considered in biomonitoring of Hb adducts in neonates.

Other species: There is a high degree of homology in the amino acid sequences in Hb of different mammalian species. Valine is the N-terminal amino acid in the α- and β-chain of several species. This amino acid residue has a similar reactivity in mouse, rat, dog and human Hb as has been shown in in vitro experiments with various low-molecular-weight epoxides and with acrylonitrile and acrylamide.¹⁰  Thus, besides a correction for species differences in the life span of erythrocytes in the laboratory animals used, comparisons of the dose of reactive compounds could be based on measurements of adducts with N-terminal valine, with only minor correction for differences in reactivity. However, the Hb from some other species studied, for instance bovine Hb, has N-terminal valines in two of the four chains.

The Hb of the rat and some but not all mouse strains contains a cysteine residue with a particularly high reactivity. Segerbäck¹¹  compared the in vitro rates of reaction of ethylene oxide with Hb and found 170 and 12 times higher reactivity of cysteine in rat and mouse Hb, respectively, than in human Hb. Species differences in Hb binding in vivo that are attributed to differences in cysteine reactivity have been demonstrated in several studies, for instance for acrylonitrile in rats and mice.¹²  This difference in reactivity has to be accounted for in interspecies comparisons.

The life span of erythrocytes with their content of Hb is generally considered to be about 4 months in humans. Thus, the level of Hb adducts observed in a blood sample is the result of adduct formation and adduct removal during the 4 months prior to sampling [Scheme 4.1, Equation (4.1)].

The yield of adduct formation in Hb depends on the concentration and the persistence (AUC) of the electrophilic compound within the erythrocytes. Further, the yield depends on the chemical reactivity of the electrophile towards the nucleophilic sites involved in the adduct formation. The relationship between adduct yield and AUC for a single exposure of relatively short duration is shown in Scheme 4.1, Equation (4.2).

The turnover of erythrocytes is the main cause of adduct removal. Although most adducts studied to date have been shown to be chemically stable, some adducts, for example those with carboxyl groups, have been shown to be eliminated faster than would be predicted based on erythrocyte turnover. In special cases there might be other parameters that have to be considered in comparisons of Hb adduct levels. In humans, changes in body weight and blood volume may become relevant during pregnancy and in neonates. In experimental animals exposed to carcinogens the dilution of Hb adducts due to increased body weight during the course of the experiment has to be taken into account. Further, exposures at high levels to certain chemicals may cause hemolysis, and increased altitude does increase Hb concentrations. Approximately 20% of the Hb content is lost from the circulating erythrocyte during its life time. This process is most pronounced in old cells and has a marginal effect only on adduct elimination (reviewed by Osterman-Golkar and Vesper¹³ ). Recent studies have indicated an interindividual variation in the erythrocyte life span. Furne et al.¹⁴  found a life span of 122 ± 23 days in 40 healthy volunteers. It has been believed that fetal red blood cells have a considerably shorter life span than in adults. Later literature however strongly indicates that the fetal erythrocyte life span is about the same as in adults.^15,16 

Single exposures to reactive compounds are frequently used in experimental animals to study adduct formation and removal. Acute human exposures may occur in connection with accidental release of chemicals, cancer therapy or anaesthesia. In a few studies volunteers have been exposed to defined low concentrations (given in mg/kg body weight) of a carcinogen in order to determine the relation between administered amount (exposure) and in vivo dose as measured by Hb adduct levels.¹⁷ 

Following a single exposure, chemically stable adducts decline in a nearly linear fashion reaching zero or a background level after a period of time corresponding to the erythrocyte life span. The relatively well characterized and long life span of the erythrocyte is one of the major advantages of Hb adducts as biomarkers. In contrast to adducts with DNA, adducts with Hb are not subjected to repair. Figure 4.3 illustrates the time windows available for measurements of urinary metabolites (typically 1 day) and DNA adducts (the half-life is set to about 4 days in this example) as compared to months in the case of stable Hb adducts.

Adducts with Hb accumulate during prolonged exposure and reach a steady-state level where the rates of adduct formation and adduct removal are equal (Figure 4.4). In the case of chemically stable adducts the steady state is reached after about 122 days (t_er) and the accumulated adduct level could be estimated as about a × 61, where a is the average daily adduct increment [Scheme 4.1, Equation (4.1b)]. Unstable adducts accumulate to a less degree (see Figure 4.4). If the exposure is terminated the adduct levels would decline in a curvilinear manner.^18,19  Environmental and life-style exposures to carcinogenic compounds are generally long term and at a fairly constant level. Thus, measured adduct levels are usually assumed to represent a steady state and could be used for the estimation of the daily average adduct increment and calculation of daily dose [AUC; Scheme 4.1, Equation (4.2)].

The methodologies applied today for the measurement of Hb adducts usually involve mass spectrometry (MS) as the final analytical step. A few studies have employed other detection techniques such as fluorescence detection or immunochemical approaches. Mass spectrometry techniques can offer the high sensitivity and selectivity needed for measurement of the very low levels of adducts that often are encountered. Both gas chromatography (GC) and liquid chromatography (LC) combined with MS are used. GC-MS has been the major technique used, primarily because it has been available for a longer time. Recent developments are often focused on LC-MS techniques. A few applications using capillary electrophoresis in combination with MS have also been described.

Measurement of Hb adducts in the general population from a low background exposure of potentially toxic compounds puts high demands on the analytical performance with regard to sensitivity. N-terminal adducts from, for example, epoxides can be measured down to a few pmol per gram Hb, which corresponds to about 1 modification in 10 million non-modified globin chains. To be able to measure adducts at these low levels the pre-processing of the blood samples is crucial. Which method to choose for processing of the sample and for final detection depends on both the properties of the electrophilic compound of interest and the physical and chemical properties of the adduct formed. Electrophilic compounds differ in reactivity towards different sites in the globin, as mentioned in the Introduction.

There are two principal ways to perform adduct measurements, either by analysis after detachment of the adduct from the amino acid residue in the protein, or by cleavage of peptide bonds and analysis of the modified amino acid or modified peptide. The latter approach, where the analyte includes a moiety from the protein, is advantageous because it increases the specificity of the analyte. Below are examples of frequently used principles.

Aromatic amines, isocyanates and polycyclic aromatic hydrocarbons (PAHs) are examples of compounds that can be analyzed as Hb adducts using mild hydrolysis of the Hb sample. Adducts bound as sulfonamides to cysteine or esters to carboxyl groups can be detached from the nucleophilic atom in the amino acid through alkaline or acidic hydrolysis. The principle was demonstrated about 30 years ago and is applied frequently (reviewed by^20,21 ). Different approaches for extraction and enrichment of the free adducts have been described. Liquid–liquid extraction or solid phase extraction has been optimized for different adducts and their specific physical/chemical properties. Derivatization of detached and enriched adducts has been performed to improve chromatographic properties as well as the sensitivity of the analysis. As an example, a methodology frequently applied for biomonitoring of 4-aminobiphenyl, a known carcinogen formed in for example cigarette smoke, is illustrated in Figure 4.5. The detached 4-aminobiphenyl is derivatized with pentafluoropropionic anhydride to obtain an analyte with high response when analyzed by GC-MS.

Adducts with N-terminal valine can be measured with different methodologies, but the principle most frequently applied has been based on the “N-alkyl Edman” procedure. It was first applied for analysis of ethylene oxide Hb adducts but has since then been applied to measurements of several other alkylating agents such as acrylamide, acrylonitrile and several low-molecular-weight epoxides.⁸  In contrast to the classical Edman degradation in which an acidification step is required for detachment of N-terminal residues, the adduct-modified N-terminal amino acids detach spontaneously as thiohydantoins at pH∼7 when reacted with phenyl isothiocyanates. A fluorinated Edman reagent has been used for the detachment of the adduct-modified valines, which improves the sensitivity when analysis is carried out with GC-MS (NICI).²²  A high specificity is obtained because the final analyte consists of an intact adduct moiety as well as the amino acid site.^8,22  The properties of the Edman reagent have been exploited to improve different clean-up approaches and detection techniques; for example, fluorescein isothiocyanate has been shown to facilitate analysis with LC-MS.^23,24  The principle is illustrated in Figure 4.6.

Detachment of adduct-modified amino acid sequences by enzymatic digestion is also a technique with several applications for measurement of adducts with Hb as well as with serum albumin. The enzyme usually cleaves at specific sites in the protein chain and the adduct-modified site is left intact. An example is given in Figure 4.7, where the principle of a reference method for measurement of HbA1c is illustrated.²⁵  Measurement of HbA1c/glycohemoglobin is a widespread application for monitoring of the mean level of glucose in the blood. HbA1c is obtained when glucose forms an adduct with the N-terminal valine of the β-chain in Hb after a two-step non-enzymatic reaction. The carbonyl carbon of the glucose reacts with the amino group on the N-terminal valine and an unstable Schiff base is formed. The Schiff base may either dissociate or undergo an Amadori rearrangement to form a stable adduct. HbA1c is the most abundant fraction (75–80%) in the group of glycohemoglobins that also consists of adducts formed with other sites in Hb and with other monosaccharides. The adducts from glucose in healthy people without diabetes are present in levels as high as a few percentage points of the N-terminal valine in Hb.

With regard to human exposure to carcinogens, Hb adducts were initially applied as biomarkers for the detection and quantification of occupational exposure. Several classes of compounds that are reactive or could give reactive metabolites are produced or used as intermediates in the chemical industry. A range of studies has concerned exposure to aromatic amines, which are widely used in the production of dyes, agricultural chemicals, etc. Later nitroarenes (also including compounds used as explosives) and aromatic diisocyanates (for production of polyurethane foams and other products) have been biomonitored with the same method (see Figure 4.5).^20,26  Occupational exposure to ethylene oxide, used for instance for sterilization of medical equipment, was early studied by Hb adduct measurement.^6,27  Many studies of exposure to low-molecular-weight electrophilic compounds have followed, using measurement of adducts with N-terminal valines in Hb (c.f. Figure 4.6) (reviewed by Ogawa et al.²⁸ ). Compounds such as styrene oxide,²⁹  propylene oxide and epoxides from butadiene,^30,31  acrylonitrile and acrylamide³²  have been monitored in work environments. Isocyanate exposure could be measured specifically as carbamoylated N-termini in Hb, and this method was applied to elucidate the exposure after the Bhopal disaster with high acute exposure to methyl isocyanate (used for pesticide production) in 1984.³³  In workers exposed to dimethylformamide the same adduct (from a metabolite) has been monitored.³⁴  A class of compounds known as allergens and extensively used in the chemical industry is organic acid anhydrides. For these compounds Hb adducts (released by hydrolysis) could also serve as biomarkers.³⁵  An influence of genetic polymorphism in metabolizing enzymes on adduct levels has been found in studies of occupational exposure, for instance to acrylonitrile.³⁶ 

Biomonitoring of exposure through Hb adducts has often been of great importance for retrospective assessment of exposure. One illustrative example concerns an acute situation with high exposures to acrylamide in tunnel construction work in Sweden, where Hb adduct measurement was used to clarify the exposure situation. The construction of the railway tunnel through a mountain ridge (Hallandsås) in the south-west of Sweden had encountered difficulties caused by heavy water inflow into the tunnel, and therefore a grouting agent had to be applied to seal the tunnel walls. A product containing the monomers acrylamide and N-methylolacrylamide was used in large quantities from August 1997. Subsequently, at the end of September, dead fish in a fish culture and paralyzed cows were found downstream of the rivulet into which leakage water from the tunnel was pumped. The grouting agent was a suspected cause. Hb adducts were measured in the cattle and demonstrated that very high exposure to the acrylamides had occurred (through the drinking water supply). This was taken as proof that the leakage of acrylamides from the tunnel was the cause of the observed effects. Questions were raised immediately about exposure of the tunnel workers and the residents in the area; there could be exposure to acrylamides through inhalation, skin, and even ingestion. The tunnel work was stopped. In the acute phase a few supposedly highly exposed workers and control persons were selected for monitoring of Hb adducts in blood, which was the only possible way to detect and quantify whether any exposure had occurred. The measurements showed adduct levels up to about 4 nmol/g Hb, thus indicating high exposures and that there was a demand for a careful risk assessment. Acrylamide has high reactivity towards proteins and is neurotoxic. It is metabolized to glycidamide which compared to acrylamide has high reactivity towards DNA, and is mutagenic (Figure 4.8). The risk assessment was aided by data generated in a study of acrylamide-exposed workers in China, where Hb adducts had been measured and peripheral nervous system (PNS) symptoms investigated.^32,37  The adduct levels measured in the tunnel workers showed that there was a high risk of PNS symptoms and the assessment of cancer risk found the exposure to be unacceptably high. A complete characterization of the occupational exposure situation for all the workers was done by measurement of Hb adducts (as shown in Figure 4.9).³⁸  From the results of the examination of PNS symptoms a no-observed adverse effect level (NOAEL) for mild symptoms was estimated and found to correspond to 0.5 nmol/g Hb of the measured acrylamide adduct.³⁸ 

In this example measurement of Hb adducts was not used to monitor an ongoing exposure; it was used to demonstrate and quantify an exposure that had already occurred and to make a risk assessment. In other cases measurement of Hb adducts has been used to control or improve the work environment and/or to exclude exposure and thus reduce anxiety.³⁹ 

There is one specific application of Hb adduct measurement that is now used as a routine method. This is the measurement of glycated Hb (see Figure 4.7), which is used in the assessment of glycemic control in patients with diabetes and which can provide an indication of how well diabetes is being managed. A similar, less frequently used application is the monitoring of urea levels associated with renal failure by measurement of increased carbamoylation of N-terminal valine in Hb.⁴⁰ 

With regard to exposure to genotoxic compounds there are a limited number of clinical applications of Hb adduct measurement so far. One example concerns the widely used drug lidocaine, which has been shown to form genotoxic metabolites of dimethylaniline, measured as Hb adducts, in treated patients.⁴¹  Monitoring of Hb adducts has been tested for use in monitoring of chemotherapy by nitrosoureas.⁴²  A few cases are described in the literature in which Hb adducts have been used to identify exposure to sulfur mustard in victims of the Iran–Iraq conflict.^43,44  Another illustration of application within medicine is a case in which analysis of acrylamide adducts was used to aid in the diagnosis of severe ill-health in one person, revealing that unknown high exposure to acrylamide was the most probable cause of the observed symptoms.³⁹ 

Adducts from aromatic amines have been measured in many studies of smokers and non-smokers (reviewed by Sabbioni and Jones²⁰ ). It has been shown that adduct levels from several aromatic amines are increased in smokers compared with non-smokers. In particular, adducts from 3- and 4-aminobiphenyl (ABP) have been studied in depth. For instance, it has been shown that the levels of 4-ABP adducts increase with exposure to environmental tobacco smoke. The level of 4-ABP adducts in non-smokers is consistent with contamination in air, food and water. Individual susceptibilities with regard to metabolic enzymes have been studied, and it has been shown that “slow” acetylators have a higher adduct level from 3- and 4-ABP than “rapid” acetylators at similar exposure levels.

Urban air pollution has been one subject of study, and ethylene, which is metabolized to ethylene oxide, is one air pollutant studied in depth. Also in this case smokers have an increased level of adducts. Studies of the origin of the background levels (ca. 20 pmol/g Hb) of this adduct in non-smokers showed that the major contribution arose from endogenous ethylene/ethylene oxide (reviewed by Törnqvist and Kautianen⁴⁵ ), which thus makes it difficult to observe an increment caused by exposure from air pollution. Furthermore, animal experiments showed an influence on the adduct level from fat in the diet and intestinal flora.

The results from studies of adducts from aromatic amines and ethylene oxide demonstrate clearly that Hb adduct measurements could detect an exposure to potentially genotoxic compounds due to complex exposure from tobacco smoking. More interesting, though, is the fact that background exposure in non-smokers from environmental exposure, but also from endogenous production, has been demonstrated with measurements of Hb adducts.

The investigation of the exposure to acrylamide in the tunnel construction work (see above) initiated studies that disclosed that acrylamide is present in many staple foods. A background level of acrylamide adducts with Hb from non-smoking control persons (ca. 40 pmol/g; see Figure 4.9) had already been indicated earlier.⁴⁶  If the observation in non-smokers really originated from a background exposure to acrylamide it would correspond to a rather high continuous exposure (daily intakes of 10–100 μg). The cancer risk evaluations done in the context of the exposed tunnel workers strongly indicated that this potential background exposure had to be studied in more depth. It was hypothesized that the adduct level in non-smokers was related to heated food. This was based on the fact that acrylamide adduct levels are higher in smokers, and on the observation of very low adduct levels in different animal species studied as controls in the context of the acrylamide leakage during construction of the tunnel. Experiments showed that rats given fried laboratory animal feed had clearly much higher acrylamide adduct levels than the control rats. The identity of the adduct was proven by mass spectrometric methods and higher levels of acrylamide were also demonstrated in the fried feed.⁴⁷  In the next step researchers studied whether acrylamide could be formed in human foodstuffs, and high levels of acrylamide were demonstrated in potato products, bread, etc. during normal preparation at high temperatures.⁴⁸ 

The finding that acrylamide is formed in foods was unexpected, and led to concerns about health risks in the population associated with the exposure to acrylamide via food. Research was initiated in related fields, and biomonitoring of Hb adducts from acrylamide has since then been applied in many experimental and molecular epidemiological studies. Table 4.1 shows background levels of the acrylamide Hb adduct (to N-termini) measured in non-smokers from different populations. The relatively small variations in median levels, 25–50 pmol/g, observed in these studies reflect variations both between methods/laboratories and among populations.^49–63  Small differences have, however, been observed between populations from different countries, which are most likely to be explained by eating habits.⁵⁹  Measured Hb adduct levels also have indicated a higher intake of acrylamide in adolescents and children.^57,63  The presence of acrylamide in foods has also been verified in innumerable studies. Mean daily intakes have been estimated to be 0.5–1.0 μg/kg body weight (35–70 μg/person).

A human study that involved high intake of acrylamide-rich foods, with measurement of acrylamide concentrations, for 4 days showed clear-cut relationships between intake and the increments in adduct levels for both acrylamide and the metabolite glycidamide.⁶⁴  Figure 4.10 illustrates the daily increment in adduct level above the present background level of the adduct.⁶⁴  The lower curve illustrates the “area under the concentration curve” from the intake of acrylamide. Such data from intervention experiments could be used to calculate the mean intakes from adduct levels measured in groups.

Hb adducts from ethylene oxide, propylene oxide, methylating and ethylating agents, acrylonitrile or acrylamide have been measured in several studies of smokers and non-smokers. Higher levels have been found in smokers throughout, and background levels have been found in non-smokers (for instance by ref. 52). For some compounds a correlation of adduct level with genotypes for metabolizing enzymes has been observed, for instance for glutathione transferase T1 and adduct levels from ethylene oxide.⁶⁵ 

Monitoring of Hb adducts has been shown to be very useful for many simple alkylating agents, as illustrated above. Limitations are indicated; for instance bulky compounds may have difficulty in crossing the erythrocyte membrane. Hb adducts, however, at very low levels, have been measured from PAH (reviewed by Kafferlein et al.⁶⁶ ). Measurement of adducts to serum albumin is an alternative, and have been applied successfully, for instance for measurement of high exposures to benzene.⁶⁷  Another example is exposure to aflatoxin, which has been monitored through serum albumin adducts in several populations.⁶⁸ 

The use of protein adduct monitoring is expected to increase with regard to the studied compounds and monitor proteins used. One example is recent studies that demonstrate that certain beneficial exposures could be monitored through adducts. This concerns isothiocyanates, which have been shown to have cancer preventive effects in animal experiments.⁶⁹  Other recent developments include use of adducts with keratin to measure dermal exposure specifically.⁷⁰ 

An interesting area for further research is the exploration of the significance of background adducts as a reflection of exposure from exogenous and endogenous sources of electrophilic compounds and of metabolism. Several methods for adduct analysis are developed to semi-high throughput methods,^71,72  and it is now possible to measure adducts in large cohorts. The largest cohorts studied concern monitoring of the intake of acrylamide from food in the general population,^56,63  and mother–child cohorts (von Stedingk et al., to be published) with more than 1000 blood samples analyzed. Methodological developments also strive towards analysis of small volumes of blood.^63,73  The new developments could facilitate the recent adductomic approach, which is aimed at the identification of unknown adducts or finding adduct patterns that signify health status or exposures.

Economic support from the Swedish Cancer and Allergy Foundation, Swedish Research Council Formas and the EU Integrated Project NewGeneris, 6th Framework Programme, Priority 5: Food Quality and Safety (Contract no. FOOD-CT-2005-016320) is acknowledged.