Albumin adducts in fetal dried blood spots as measures of in utero exposures

Introduction.   Neonatal dried blood spots (DBS) offer a convenient matrix for estimating in utero exposures to toxic chemicals.  Since some US states and other countries have archived neonatal DBS with parental consent, such exposures may be particularly relevant in case/control studies of childhood cancers.  Here, we present a method for measuring adducts bound to human serum albumin (HSA) in neonatal DBS. We focus upon adducts of benzoquinone (BQ) and naphthoquinone (NQ) bound to HSA-Cys34.  These adducts result from exposures to precursor molecules (i.e. benzene, phenol, hydroquinone and naphthalene) which are present in cigarette smoke, and may elucidate the relationship between parental smoking and childhood leukemia.  Adduct levels reflect exposures received during the last month of gestation.    

Methods.   Adducts were measured in DBS collected, with parental consent, from 54 control children from the Northern California Childhood Leukemia Study.  One half (n = 27) of these control children had mothers who smoked during pregnancy while the other half of the children (n = 27) had nonsmoking mothers.  Children from smoking and nonsmoking mothers were matched by date of birth between 8/6/1985 and 4/29/2003.    Proteins were extracted from 3-mm punches of DBS and simultaneously digested with trypsin.  After adding an isotopically labeled internal standard, the 21-mer peptides containing HSA-Cys34 adducts were measured by liquid chromatography-triple-quadrupole mass spectrometry (LC-MS/MS) in selected reaction monitoring mode (SRM), using three adduct-specific SRM transitions and LC retention times for identification.      

Results.    Analyses of blinded samples are currently underway at the time of abstract submission and cannot be related to sample identities until all assays are completed.  Two-way ANOVA will be used to determine whether geometric mean adduct levels are significantly different between smoking and nonsmoking mothers and regarding time of storage.  If significant differences are detected by time of storage, linear models will be applied to identify trends.