Once systemically available in the body, toxic substances or their reactive inter- mediates may interact with macromolecules such as deoxyribonucleic acid (DNA), ribonucleic acid (RNA), and proteins. Some of these substances or their electrophilic metabolites are capable of covalent binding to nucleophilic sites in endogenous macromolecules. Others may be noncovalently associated with plasma proteins and be released in target organs. This may lead to tissue injury.
In some cases, the binding to DNA causes mutations (different expression of (onco)genes) or chromosomal changes (micronuclei, chromosomal aberrations, sister chromatid exchanges). This may be a first change in a multistep process eventually leading to the conversion of a normal cell into a tumor cell. Most adducts of carcinogens have not been measured in target tissue but in sites where no tumors will occur, such as in lymphocytes. These non–target-site adducts
216Scheepersetal.
Table 6 Haemoglobin Adducts Positively Identified by GC–MS in Humans
Agent Metabolite/substance Amino acid Cleavage reagent Analyte Detector Ref.
Acetaldehyde vala
4-Aminobiphenyl N-hydroxy-4- cys 4-aminobiphenyl
aminobiphenyl
Benzene Benzene oxide cys tfaa/msa ? i
Benzo[a]pyrene benzo[a]pyrenediolepoxide asp/glu Benzo[a]pyrene-tetrols NCI 53, ii
1,2- and 1,4- cysb tfaa/msa O,O′,S-tris-trifluoro ace- NCI iii
Benzoquinone tylhydroquinone
1,3-Butadiene vala iv, v
Ethylene oxide ethylene oxide vala pentafluorophynyl NCI vi, vii
thiohydantoin
Hexahydrophthalic an- lys acid/Pronase E pfbb NCI viii, ix
hydride
Isoprene 2-ethenyl-2-methoxirane vala mod. Edman degr. (95%)
2-(1′-methylethenyl)oxirane vala (5%) x
4-(methylnirosamino- 4-Hydroxy-1-(3-pyridyl)-1- alkaline hydrolysis ECMS xi
1-(3-pyridyl)- butanone butanone
N′-nitrosonornicotine 4-Hydroxy-1-(3-pyridyl)-1- alkaline hydrolysis ECMS v
butanone
N,N- N-methyl-carbamoylated vala xii
dimethylformamide
1-Nitropyrene 1-nitrosopyrene cys alkaline hydrolysis hfb-1-aminopyrene MS-MS xiii
alkaline hydrolysis NCI xiv
2-Nitrofluorene 2-nitrosofluorene cys alkaline hydrolysis hfb-2-aminofluorene NCI xiv
HealthRiskAssessment217
3-nitrofluoranthene cys alkaline hydrolysis hfb-3-aminofluoranthene NCI xiv
9-nitrophenanthrene cys alkaline hydrolysis hfb-9-aminophenanthrene NCI xiv
6-nitrochrysene cys alkaline hydrolysis hfb-6-aminochrysene NCI xiv
Propanil 3,4-dichloroaniline alkaline hydrolysis 3,4-dichloroaniline NCI xv
2,4-Toluenediisocyanate vala/lys xvi
Styrene styrene-7,8-oxide cys Raney nickel 1- and 2-phenylethanol NCI xvii
Abbreviations: tfaa⫽trifluoroacetic anhydride, msa⫽methanesulfonic acid; hfb⫽hexafluorobutyric; ECMS⫽electron capture MS, pfbb⫽pentafluorben- zyl bromide.
aN-terminal valine.
bCommerically available human Hb.
i. K. Yeowell-O’Connell, N. Rothman, M.T. Smith, R.B. Hayes, G. Li. S. Waidyanatha, M. Dosemeci, L. Zhang, S. Yin, N. Titenko-Holland and S.M.
Rappaport, Carcinogenesis, 19 (1998) 1565–1571.
ii. S. Tas, J.P. Buchet, R. Lauwerys, Int. Arch. Occup. Environ. Health, 66 (1994) 343–138.
iii. S. Waidyanatha, K. Yeowell-O’Connell, S.M. Rappaport, Chem. Biol. Interact., 115 (1998) 117–139.
iv. K.P. Braun, J.G. Pavlovich, D.R. Jones, C.M. Peterson, Alcohol Clin. Exp. Res., 21 (1997) 40–43.
v. S. Osterman-Golkar and J.A. Bond, Environ. Health Perspect. 104 (1996) 907–915.
vi. P.B. Farmer, E. Bailey, S.M. Gorf, M. To¨rnqvist, G. Osterman-Golkar, A. Kautiainen, D.P. Lewis-Enright, Carcinogenesis, 7, 637–640.
vii. E. Bailey, A.G.F. Brooks, C.T. Dollery, P.B. Farmer, B.J. Passingham, MA. Sleightholm, D.W. Yates, Arch. Toxicol. 62, 247–253.
viii. C.H. Lindh and B.A. Jonsson Toxicol. Appl. Pharmacol., 153 (1998) 152–160.
ix. C.H. Lindh and B.A. Jonsson, J. Chromatogr. Biomed. Sci. Appl., 710 (1998) 81–90.
x. E. Tareke, B.T. Golding, R.D. Small, M. To¨rnqvist, Xenobiotica, 28 (1998) 663–672.
xi. S.E. Atawodi, S. Lea, F. Nyberg, A. Mukeria, V. Constatinescu, W. Ahrens, I. Bureske-Hohfeld, C. Fortes, P. Boffetta, M.D. Frisen, Cancer Epidemiol.
Biomarkers Prev., 7 (1998) 817–821.
xii. J. Angerer, T. Goen, A. Kramer, H.U. Kafferlein, Arch. Toxicol., 72 (1998) 309–313.
xiii. Y.M. van Bekkum, P.T.J. Scheepers, P.H.H. van den Broek, D.D. Velders, J. Noordhoek, R.P. Bos, J. Chromatogr. B, 701 (1997) 19–26.
xiv. Zwirner-Bayer and H.-G. Neumann, Mutat. Res. 441 (1999) 135–144.
xv. R. Pastorelli, G. Catenacci, M. Guianci, R. Fanelli, E. Valoti, C. Minoia, L. Airoldi, Biomarkers 3 (1998) 227–233.
xvi. D. Schu¨tze, O. Sepai, J. Lewalter, L. Miksche, D. Henschler, G. Sabbioni, Carcinogenesis, 16 (1995) 572–582.
xvii. S. Fustinoni, C. Colosio, A. Colombi, L. Lastrucci, K. Yeowell-O’Conell, S.M. Rappaport, Int. Arch. Occup. Environ. Health 71 (1998) 35–41.
218 Scheepers et al.
Figure 6 Proposed mechanism of Hb adduct formation by 1-NP, and decomposition of the adduct by acid or basic hydrolysis during pretreatment of the sample prior to GC–
MS analysis.
Figure 7 GC–MS–MS mass chromatogram of the product ions of heptafluorobutyryl- 1-aminopyrene (HFB-1-AP) at m/z 216 (top) and HFB-D9-AP at m/z 225 (middle) with mass spectrum, and total ion chromatogram (bottom) of hydrolyzed pooled Hb of rats ex- posed to a single dose of 1 mg or 10 mg of 1-nitropyrene in trioctanoin per kg body weight by gavage. Description of system and conditions in Table 2. (From Ref. xiii in Table 6.)
Health Risk Assessment 219
are good measures of the internal dose of the active genotoxic compound. In experimental animal models, target-site DNA adduct levels have been found to correlate with the appearance of tumors for four different classes of carcinogens:
N-nitrosoamines, aflatoxines, aromatic amines, and PAHs [58,59]. However, a relationship between the DNA adduct level at a non–target site with the adduct level in target tissue must be established before DNA adduct levels can be used as a basis of establishing risk estimates [60]. Deoxyribunucleic acid adducts may be removed and return the nucleic acids to their original state by cell turnover or by enzymatic repair.
In contrast to DNA adducts, adducted proteins are not known to be subject to enzymatic repair. Their life span may approach the life span of the unadducted protein (in humans: 120 days for hemoglobin and 20 days for albumin). Long- term exposure to low concentrations of an adduct-forming substance may be re- flected in an accumulated protein adduct level. Nucleophilic sites such as amine, thiol, and carboxylic functional group are available in different amino acids such as N-terminal valine, cysteine, histidine, aspartic acid, and glutamine, for which adducts have been identified in vivo. For PAHs and aromatic amines, linear dose–
response relationships have been established between the dose administered in animal models and hemoglobin (Hb) or albumin adduct levels [61]. On the other hand, a high correlation was observed for some aromatic amines between levels of Hb adducts and DNA adducts in animal models [62]. Bladder cancer patients have been found to a have an elevated 4-aminobiphenyl (4-ABP) adduct level [63], and the 4-ABP–Hb adduct level correlated with the 4-ABP–DNA adduct level in the target tissue (bladder epithelium) [64].
Adducts of DNA, RNA, and proteins such as Hb and albumin have been characterized with LC–MS and GC–MS. Table 6 presents an overview of Hb adducts from humans mostly analyzed by LC–MS or GC–NCI-MS. Usually these determinations help to understand the mechanism of toxicity and may be used for the assessment of uptake and systemic availability of toxic substances in exposure monitoring using biomarkers of exposure. Efforts have been made to use these biomarkers also in risk calculations [65].
The formation of protein and DNA adducts may be studied in vitro by incubation of the macromolecule with a reactive intermediate or with the parent compound in the presence of a metabolizing system, or in vivo in experimental animals. In most studies, radiolabels are used that may be analyzed by liquid scintillation counting but more recently also by a specific MS-based technique (see section 5 of this chapter).
For structure characterization and quantification, adducts to Hb may be cleaved by alkaline or acid hydrolysis (Fig. 6), by enzymatic hydrolysis, e.g., Pronase E, or by Raney nickel. The cleavage products may be separated from the debris by liquid–liquid chromatography, solid-phase extraction, gel filtration, ion-exchange chromatography, or affinity chromatography. In most cases, the
Figure 8 Characterization by GC–MS in CI mode (using methane as a reaction gas) of cleavage product from plasma proteins recovered from Sprague-Dawley rats following administration of 1 mg 1-NP per kg of body weight by gavage (a) and of synthesized and subsequently reduced and methylated 1-NP-4,5-dihydrodiol (b). The samples for MS characterization were obtained from HPLC separation monitored by fluorescence detec- tion. The appearance of the peak of 1-AP is indicated. Description of system and condi- tions in Table 2.
Figure 9 EI-MS (upper panel) and EI-MS–MS (lower panel) chromatograms of a plasma protein hydrolysate from a nonsmoking worker exposed to diesel exhaust in an indoor facility with running truck engines. The blood sample was collected on a Thursday morning. The preceding exposure to 1-NP was 510 pg/m3on Monday and 1906 pg/m3on Tuesday (time- weighted average exposure during 8 h working period). The subject was classified as having a slow acetylator and slow CYP2A1 phenotype and being GSTM1 deficient. Shown are the total ion chromatogram (top), product ion trace of the HFB derivatized 1-aminopyrene (middle) and of the HFB derivatized internal standard (bottom). In the MS–MS mode, it was possible to quantify the adduct content as 421 fg 1-NP per mg of plasma protein.
222 Scheepers et al.
cleavage product is then derivatized in a separate step following cleanup by frac- tionation. Figure 7 shows the GC–MS–MS analysis of hydrolyzed Hb of rats exposed to 1-NP. In order to assess the performance of the MS–MS system com- pared with MS in EI mode, a blood sample from a worker exposed to diesel exhaust was compared in both detectors. The EI-MS chromatogram shows the appearance of numerous peaks emerging from the sample matrix (the most promi- nent peak tentatively identified as a cholesterol compound) and from the use of plastics in the pretreatment or analytical system (presumably a phthalate analog), whereas the MS–MS chromatogram shows a much cleaner total ion current and product ion traces with peaks that can be readily used for quantification purposes.
Some investigators report the use of a reagent during the cleavage process in a combined procedure such as in (modified)N-alkyl Edman degradation for analysis of adducts to the N-terminal valine. The reagent pentafluorophenyl iso- thiocyanate cleaves the adducted N-terminal amino acid from the protein chain as a substituted pentafluorophenyl thiohydantoin, which is subjected to MS analy- sis [66].
Figure 8 shows a plasma (presumable albumin) adduct of a metabolite of 1-NP, tentatively identified as 1-AP-4,5-dihydrodiol after methylation and the GC–MS analysis of a product isolated from rats exposed to 1-NP. It is compared with an identically treated synthetic standard (1-NP-4,5-dihydrodiol).
In the analysis of traces of protein adducts from human blood samples, MS–MS offers great sensitivity over EI-MS, equaling the sensitivity of NCI-MS for the analysis of Hb adducts of arylamines (Fig. 9). In the next section, future developments in the use of MS for adduct analysis in the reconstruction of histori- cal exposure are further described.