Antagonistic activity towards the androgen receptor independent from natural sex hormones in human milk samples from the Norwegian HUMIS cohort

Bérénice Collet,Barbara M.A. van Vugt-Lussenburg,Kees Swart,Rick Helmus,Matthijs Naderman,Eva de Rijke,Merete Eggesbø,Abraham Brouwer,Bart van der Burg

doi:10.1016/j.envint.2020.105948

Abstract

In this paper, we investigated the possible presence of endocrine disrupting chemicals (EDCs) based on measuring the total estrogenic and androgenic activity in human milk samples. We used specific bioassays for analysis of the endocrine activity of estrogens and estrogen-like EDCs and androgens and androgen-like EDCs and developed a separation method to evaluate the contribution from natural hormones in comparison to that of EDCs to total endocrine activities. We extracted ten random samples originating from the Norwegian HUMIS biobank of human milk and analyzed their agonistic or antagonistic activity using the ERα- and AR CALUX® bioassays. The study showed antagonistic activity towards the androgen receptor in 8 out of 10 of the assessed human milk samples, while 2 out of 10 samples showed agonistic activity for the ERα. Further investigations demonstrated anti-androgenic activity in the polar fraction of 9 out of 10 samples while no apolar extracts scored positive. The culprit chemicals causing the measured antagonistic activity in AR CALUX was investigated through liquid chromatography fractionation coupled to bioanalysis and non-target screening involving UHPLC-Q-TOF-MS/MS, using a pooled polar extract. The analysis revealed that the measured anti-androgenic biological activity could not be explained by the presence of endogenous hormones nor their metabolites. We have demonstrated that human milk of Norwegian mothers contained anti-androgenic activity which is most likely associated with the presence of anthropogenic polar EDCs without direct interferences from natural sex hormones. These findings warrant a larger scale investigation into endocrine biological activity in human milk, as well as exploring the chemical sources of the activity and their potential effects on health of the developing infant.

Highlights

Breast milk contains a complex mixture of proteins, lipids, carbohydrates as well as a high concentration of bioactive components, and is acknowledged to be important to the infant’s post-natal growth and development
At relatively high dosage levels, several endocrine active chemicals (EACs) have been shown to lead to adverse effects, and these chemicals are referred to as endocrine disrupting chemicals (EDCs) (Bergman et al, 2012)
As a first step we demonstrate the performance of the Androgen Receptor (AR) CALUX bioassay, both in agonistic and antagonistic mode, by comparing our internal database to the in vivo Hershberger database recently established by Browne et al (2018), as has been done to evaluate the ToxCast/Tox 21 AR model conducted by Kleinstreuer et al (2017)

Summary

Introduction

Breast milk contains a complex mixture of proteins, lipids, carbohydrates as well as a high concentration of bioactive components, and is acknowledged to be important to the infant’s post-natal growth and development. It has been shown that single specific in vitro tests covering only a limited set of conserved pathways can provide very good predictions with respect to developmental and reproductive toxicity landmarks such as sex organ deformities (Van der Burg et al, 2014) These highly selective reporter gene methods to assess (anti-)estrogenic and androgenic activity, were shown to be excellent alternatives to the traditional in vivo techniques in rodents and extremely suitable for measurements in complex mixtures such as body fluids (Pedersen et al, 2010; Sonneveld et al, 2006). Estrogen Receptor alpha (ERα) and Androgen Receptor (AR) mediated bioassays are based on two human U2-OS osteoblastic osteosarcoma cell lines stably transfected to endogenously express the ERα or AR, respectively The activation of these specific receptors upon ligand stimulation triggers their binding to hormone-responsive elements (HRE), which is linked to a luciferase (“reporter”) gene, leading to luciferase expression. Over the years CALUX bioassays have been proven to be quick, specific methods able to measure the total effect of ligands on a receptor of interest using a limited volume of samples (Houtman et al, 2009; Kraus et al, 1995; Sonneveld et al, 2005)

Results

Discussion

Conclusion