Distribution of perfluorooctanesulfonate and perfluorooctanoate into human plasma lipoprotein fractions

Abstract

Some cross-sectional epidemiological studies have reported positive associations of serum concentrations of non-high density lipoprotein cholesterol with serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA). However, the strength of the reported associations is inconsistent for exposure–response across three orders of magnitude of serum PFOS and/or PFOA concentrations. These positive associations are unexpected based on toxicological/mechanistic studies, suggesting that the associations may have a biological, rather than a causal, basis. This study tested the hypothesis that PFOS and PFOA distribute into serum lipoprotein fractions such that increases in serum lipoproteins would result in corresponding increases in serum concentrations of PFOS and PFOA. Based on observed binding of PFOS and PFOA to isolated β-lipoproteins in physiological saline (96% and 40% bound, respectively) in preliminary experiments using ultrafiltration and LC–MS/MS methods, binding to human donor plasma lipoprotein fractions was investigated by two density gradient methods. The majority of PFOS and PFOA recovered masses were found in lipoprotein-depleted plasma. Plasma density gradient fractionation data suggested that maximally 9% of PFOS distributes to lipoprotein-containing fractions, yet only 1% or less of PFOA is so distributed. These data do not support a strong role for plasma lipoprotein fractions in explaining the inconsistent dose–response associations reported in cross-sectional epidemiological studies.