Detection of polycyclic aromatic hydrocarbon metabolites by high-pressure liquid chromatography after purification on immunoaffinity columns in urine from occupationally exposed workers.

Abstract

The objective in our study was to quantitate benzo[a]pyrene (B[a]P) metabolites by a combination of immunoaffinity chromatography and high-pressure liquid chromatography (HPLC) with fluorescence detection in urine from workers exposed to high levels of polycyclic aromatic hydrocarbons (PAH). Furthermore, by the simultaneous quantitation of 1-hydroxypyrene, the correlation between the B[a]P-tetrol and 1-hydroxypyrene would provide a means of evaluating the validity of 1-hydroxypyrene as a surrogate biomarker for occupational exposure to the potent carcinogen B[a]P in an electrode paste plant. The study was carried out at an electrode paste plant that produces electrode paste for Söderberg electrodes. A total of 34 pre- and post-shift urine samples and 17 personal air samples were collected from 17 workers during a normal work week. The concentration of 1-hydroxypyrene was measured in all urine samples. A recent method of quantitating B[a]P-r-7, t-8, t-9, c-10-tetrol in urine of humans exposed to low levels of PAH has been described. A modified version of this method involving purification of urine samples on immunoaffinity columns and HPLC analysis with fluorescence detection was used on urine samples from workers exposed to high levels of PAH. A monoclonal antibody (8E11) with binding affinity to B[a]P-tetrols was used. This antibody also binds several PAH-DNA adducts and metabolites, including 1-hydroxypyrene. Gas chromatography/mass spectroscopy (GC/MS) was also used for identification of metabolites isolated by HPLC fractionation. From personal air sampling the mean exposure to particulate PAHs was 38 microg/m3. The mean concentration of urinary 1-hydroxypyrene was 3.9 micromol/mol creatinine in preshift samples and 10.2 micromol/mol creatinine in postshift samples. We could not identify detectable amounts of urinary B[a]P-tetrol by HPLC or fluorescence spectroscopy after purification on immunoaffinity columns. However, in the HPLC analysis we identified several hydroxyphenantrene metabolites that were detected at relatively high concentrations in all of the workers' urine samples. We could not separate 2- and 3-hydroxyphenanthrene (2 + 3-OH-Phe) in peak 1, and peak 2 contained both 1- and 9-hydroxyphenanthrene (1 + 9-OH-Phe). The phenanthrene metabolites were mainly conjugated to glucuronic acid and sulfate. There was a significant correlation between the 1-hydroxypyrene concentration and 2 + 3-OH-Phe (r = 0.73) and 1 + 9-OH-Phe (r = 0.64) in the urine samples. 1-Hydroxypyrene was measured in all post-shift urine samples but was not significantly correlated with workplace pyrene exposure, indicating that skin exposure is an important route of pyrene exposure in this factory. As with 1-hydroxypyrene, dermal PAH uptake may also account for the poor correlation between 2 + 3- and 1 + 9-OH-Phe and ambient phenanthrene. Since dermal uptake is likely to be important in occupational PAH exposure in addition to inhalation, estimation of total PAH exposure is best achieved by quantitation of PAHs excreted into body fluids. However, it remains unclear whether there might be a difference in uptake and urinary excretion of 3-ring, 4-ring, or 5-ring PAHs and in the correlation between these metabolites and ambient-air PAH measurements. In summary, using immunaffinity chromatography, we did not find detectable amounts of B[a]P-tetrol in urine from workers occupationally exposed to PAH. However, by an HPLC/immunoaffinity method, relatively high amounts of 1-hydroxypyrene as well as 2 + 3- and 1 + 9-OH-Phe were quantitated in the urine samples, both of which are relevant as biomarkers of PAH exposure.