On July 7, 2014 the scientific peer-reviewed journal PLOS ONE published the study “Characterization of estrogen and androgen activity of food contact materials by different in vitro bioassays (YES, YAS, ERalpha and AR CALUX) and chromatographic analysis (GC-MS, HPLC-MS)” by Austrian researchers (Mertl et al. 2014). The scientists from the OFI Austrian Research Institute for Chemistry and Technology in Vienna, Austria tested chemical migrants of various different food contact materials (FCMs) for their ability to mimic the natural hormones estrogen and testosterone in cell-based, so called in vitro bioassays.


The aim of the study was to compare two different in vitro bioassay systems for their suitability to detect hormonal activity in FCM overall migrate (of unknown chemical composition). The two systems of interest were: (1) yeast based assays, genetically engineered to contain human estrogen and androgen (testosterone) receptor, respectively, the YES (yeast estrogen screen) and YAS (yeast androgen screen); and (2) human bone cancer cell line assays, genetically engineered to contain a reporter gene system responding to estrogen (ERα CALUX) and androgen (AR CALUX) mimicking chemicals. Since all bioassays measure receptor activation they can also be used to detect chemicals blocking receptor activation, i.e. antiestrogenic and antiandrogenic activity. All types of receptor interaction are considered potential endocrine disruption, and would require further investigation for each individual chemical.

Samples and migration experiments

Johannes Mertl and colleagues prepared overall migrates from 18 pristine FCM samples received directly from packaging manufacturers or fillers/retailers; none of the samples had been in contact with food. Samples included actual food packaging (e.g. foil, film, tray, beverage bottle) or resins used in the manufacture of FCM articles, ranging from polystyrene (PS), polypropylene (PP), polyethylene (PE) and polyethylene terephthalate (PET) to composite films (CF; unknown polymers) and food cartons (FC; multimaterial). Migration experiments were performed according to protocols laid down in the EU Plastic FCM regulation 10/2011 at 60°C during 10 days, using official food simulants according to the intended FCM use (i.e. 95% ethanol solution for FCM intended for fatty foods); food simulants ranged from desalted (pure) water, 20%, 50% and 95% ethanol in water solutions and were considered by the authors to resemble respective worst case conditions.

Sample preparation

In a next step, the overall migration solutions were concentrated to prepare them for analysis in the bioassays. This step is necessary because living cells will not survive highly concentrated ethanol solutions. Therefore, a so called solid phase extraction (SPE) was performed (using Oasis HLB cartridges) where the respective food simulant is flushed but migrants dissolved in the food simulant are bound to the SPE material; according to the authors this procedure had been optimized based on several known endocrine disrupting chemicals (EDCs) present in FCMs. Migrants are then released from the SPE material by rinsing with an appropriate, bioassay compatible solvent. In some cases the solution may additionally be evaporated under air to further concentrate it.

Bioassays and chemical analysis

Bioassays were then performed for estrogenic, androgenic, antiestrogenic and antiandrogenic effects, whereby the anti-hormonal activity tests required addition of actual estrogen/androgen. For the samples where hormonal activity was found in the bioassays, the authors performed chemical analysis with the aim of identifying substances likely responsible for the observed endocrine activity. Thereby, they focused on a selection of 29 FCM authorized substances using gas chromatography-mass spectrometry (GC-MS) and high performance liquid chromatography-mass spectrometry (HPLC-MS).


Of the 18 samples tested, 6 were found to be estrogenic (ERα CALUX); four of these were also estrogenic in YES which had a higher limit of detection and therefore did not detect low estrogenicity for the other two samples. The six estrogenic samples were: two composite films, three PS samples and one PE sample. Androgenic effects were not observed, one PS sample was found to be antiandrogenic in both AR CALUX and YAS assays. In the yeast based assay systems, several antiestrogenic and antiandrogenic samples were found, however these results could, with exception of the PS antiandrogenic sample, not be confirmed in the respective CALUX assays. The authors conclude that the findings may likely be false positives related to unknown yeast-specific factors.

Chemicals confirmed in the hormone active samples included several estrogenic styrene dimers, by-products of polystyrene manufacture (1,3 diphenylpropane CAS 1081-75-0; 1,2 diphenylcyclobutane CAS 20071-09-4); 1,3 diphenylpropane was also found to be antiandrogenic. Other suspected estrogen active chemicals found to be present in the active samples were the plasticizer diethylhexyladipate (DEHA; CAS 103-23-1) and the antioxidant BHT (CAS 128-37-0), both found in different CF samples. However, Mertl and colleagues point out that no estrogenic activity was detected in their respective bioassays, whilst others have found DEHA and BHT to be estrogenic in different test systems. The estrogenic break down product 2,4 cumylphenol (CAS 2772-45-4) was found in a PE resin sample, likely stemming from the antioxidant Alkanox 28.

Conclusions and evaluation

The authors conclude that the work intensive human cell-based bioassays (ERα/AR CALUX) are more suitable for evaluation of overall migrate mixtures, because (1) lower detection limits can be achieved and (2) false positives due to yeast-specific effects are avoided. However, they concede that it is difficult to elucidate chemicals responsible for observed hormonal activity due to the presence of non-intentionally added substances (NIAS) for which not always analytical standards exist, and because essentially the chemical composition of FCMs remain unknown.

Strengths of the present study are (1) the use of pristine FCMs which have not been in contact with food, (2) adherence to regulatory migration testing protocols and (3) integration of biological analysis with chemical analysis, as far as possible. On the other hand, the specific SPE used in the present study for sample preparation has been shown by others to miss unknown estrogenic compounds (Wagner and Oehlmann 2011), and therefore may lead to an increased number of false negative results. Therefore, the present study cannot rule out the presence of hormonal activity in the samples showing no (anti)estrogenic/(anti)androgenic effects. In addition, the use of food simulants in the present study may further select specifically for chemicals with less hormonal activity and other results may be obtained if 3% acetic acid, or the non-official food simulant standardized tap water, would be used. In conclusion, identification of hormonal activity in FCMs remains a challenge, especially for mixtures of unknown chemicals present at low concentrations. Further research is necessary to close knowledge gaps and reduce uncertainty relating to the health effects of chronic exposure to low levels of EDCs from FCMs.


Mertl, J. et al. (2014) “Characterization of estrogen and androgen activity of food contact materials by different in vitro bioassays (YES, YAS, ERalpha and AR CALUX) and chromatographic analysis (GC-MS, HPLC-MS)”. PLOS ONE, 9(7) (published online July 7, 2014).

Wagner, M. and J. Oehlmann (2011). “Endocrine disruptors in bottled mineral water: Estrogenic activity in the E-Screen.” The Journal of Steroid Biochemistry and Molecular Biology 127(1–2): 128-135.