Plastic packaging is used in the production, processing, transport, handling and storage of food and beverage and is of high importance as it protects food and beverage from physical damage, soiling and microbial spoilage (Muncke et al., 2017). The abundance and high resistance to degradation of plastic waste both on land and at sea has become a global concern regarding the issue of environmental pollution with plastic debris (Muncke, 2011). In addition, plastics are contributors towards increasing the effects of global warming and the depletion of non-renewable fossil resources such as oil (Nakajima et al., 2017). Petroleum-based plastics are known sources of hazardous chemicals, with chemical plasticisers such as bisphenol-A (BPA) and phthalates regularly cited for their potential hazards and epidemiological response (Zimmermann et al., 2019). BPA and phthalates can migrate out of petroleum-based plastic packaging into food, soil, water or oceans and consequently be ingested by animals or humans with the potential to cause adverse health effects (Yang et al., 2011; Moreira et al., 2015; Muncke et al., 2017; Guillard et al., 2018).

Some petroleum-based plastic leachates have also been found to cause endocrine dysfunction in humans by interfering at the point of production, release, transport, metabolism, binding or elimination of natural hormones in the body and, therefore, are recognised as endocrine disrupting chemicals (EDCs) (Yang et al., 2011; Moreira et al., 2015). EDCs are defined by the WHO/IPCS (2002) as “an exogenous substance or mixture that alters function(s) of the endocrine system and consequently causes adverse health effectsin an intact organism, or its progeny, or (sub) populations.” Human exposure to synthetic hormone mimicking chemicals such as BPA and phthalates is associated with developmental, reproductive and metabolic disrupting effects including infertility or reduced fertility, learning and memory difficulties, diabetes, obesity and cardiovascular disease (Casals-Casas, and Desvergne, 2011; Schug et al., 2011; WHO/UNEP, 2013). EDCs can affect our sex hormones leading to disruption of reproductive development, sexual differentiation, early embryonic development and puberty (Casals-Casas and Desvergne, 2011). Estrogen receptors (ERs) are one of the major targets of EDCs, thus affecting the normal estrogen signalling pathways (Shanle and Xu, 2011).

The production and use of bio-based plastics may be a potentially safer alternative to petroleum-based plastics, whilst also helping to reduce the dependency of plastics on non-renewable fossil fuels, reducing greenhouse gas emissions and the pressure on landfills from plastic waste, reduce food waste and food chemical contamination (Mohan, 2011; Álvarez-Chávez et al., 2012; Peelman et al., 2013; Nakajima et al., 2017; Guillard et al., 2018; Firoozi et al., 2021; McGauran et al., 2021). Although bioplastics are currently assumed to be safe, there is little research investigating their environmental, health and safety impacts throughout their life cycles (Álvarez-Chávez et al., 2012). One recent study highlighted that bioplastic and plant based materials may not be safer than conventional plastics (Zimmermann et al., 2020). In addition, a recent endocrine society report highlights that much more testing and development is required before we can fully address problems related to recyclability; land, biocide and water use in the production of starch containing plants for bio-based plastics; and toxic additives in plastics are solved (Flaws et al., 2020). Ideally, all new products should be safety tested prior to use, however, this is currently not the case as it is not mandatory for laboratory tests to be carried out for all packaging (European Commission, 2011a; Packaging Essential Requirements Regulations, 2015). Specific measures exist for some types of materials and substances, however, not all materials and articles, non-intentionally added substances or aids to polymerisation are required to be tested (European Commission, 2011a). Packaging which may come into contact with food should be tested to ensure they are safe for food contact, do not transfer their constituents to foods in quantities exceeding the specific migration limits and do not cause physical or other changes to the food (European Commission, 2011a). Toxicological study could help guide bioplastic product development at early stages to assist in decision making towards safer products. The production processes, generation of by-products and the biodegradability of the bioplastic could potentially pose risks to the environment, human or animal health and safety (Álvarez-Chávez et al., 2012; Nakajima et al., 2017; Ramakrishnan et al., 2018). It is important to understand the flow and production processes, chemicals used during production and any by-products generated, in order to identify any adverse impacts in all parts of the life cycles of the bioplastics (Álvarez-Chávez et al., 2012).

Additives or non-intentionally added manufacturing by-products such as fertilizers and pesticides may leach out of bioplastics into food, drink or the environment and consequently could be ingested by humans or animals (Álvarez-Chávez et al., 2012; Seltenrich, 2015). Leaching of monomers and additives from plastics into its food contents is often accelerated if the product is exposed to common-use stresses such as ultraviolet radiation in sunlight and/or via boiling or heating. Even when using materials that initially are deemed as safe, the stresses of manufacturing can change chemical structures or create chemical reactions converting a safe material or chemical into one of concern (Yang et al., 2011). However, there is limited information about the safety of bioplastics and their leachates on human health and the environment.

Migration of leachates from packaging into food is usually assessed using substitutes of foods, known as food simulants. The hydrophilic, lipophilic or amphiphilic character of the food determines the food simulant used. For example, migration into acidic foods (pH below 4.5) is measured using acetic acid 3% (w/v) (European Commission, 2011a). The migration testing is carried out under the worst foreseeable conditions of use of the plastic, therefore, the migration into food simulants is thought to overestimate the actual migration into foods (Muncke, 2014). Overall migration testing is performed on materials intended to come into contact with different food categories or a combination of food categories. Testing is performed under European standards using standardised testing conditions, in order to assess the transfer and migration of compounds into food, drink and/or the environment (European Commission, 2011a; Muncke, 2011).

The aim of this study was to investigate the toxicological safety of a range of bio-based polymers originating from a variety of renewable resources (Table 1) and in development for food and beverage packaging. In addition to the bio-based polymers, a traditional non-bio-based synthetic petroleum based polymer called polypropylene (PP) and some inorganic bio-based additives mixed with PP were also assessed. Assessment was achieved by applying migration test samples, prepared according to European test standards, to mammalian reporter gene assays (RGAs) incorporating natural sex steroid receptors capable of measuring (ant)agonism of estrogen, androgen and progestogen nuclear receptor transcriptional activity (Willemsen et al., 2004; Connolly et al., 2011; Frizzell et al., 2013). Parallel to the RGAs, the MTT assay was used to measure cellular metabolic activity as an indicator of cell viability and to monitor cytotoxic effects of test samples.

The potential hormonal activity of the migration test samples were investigated at the level of nuclear receptor transcriptional activity using a panel of RGAs. Parallel to the RGAs, cellular metabolic activity was also measured using the MTT assay to monitor the cytotoxic effects of the samples. All results are summarised in Table 4.

This study investigated the potential endocrine disrupting effects of bio-based polymers at the level of estrogen, androgen and progestagen nuclear receptor transcriptional activity using RGAs. Parallel to RGAs, metabolic activity was measured using the MTT assay, as an indicator control of sample gross cellular cytotoxicity. All cytotoxic and hormonal leachate activity observed in the study is summarised in Table 7.

The study revealed that eggshell/polypropylene (10:90), PHB, PBAT, TPS, PBS, PCL and PLA/PBAT blends did not induce any hormonal or cytotoxic activity at any of the standardised testing conditions. Therefore, they are potentially safer candidates for further development in the bioplastic industry, particularly for food packaging applications.

After 10 days of migration testing at 40°C, eggshell/polypropylene (0.1:99.9) in MeOH and the non-bio-based PP in MeOH and dH₂O decreased metabolic activity but only in the estrogen responsive MMV-Luc cell line. This cell line specific effect may be due to the robustness of the differing RGA parent cell lines, the MCF-7 cell line is the parental lineage for the MMV-Luc cell line while T47D is the parental lineage for the TM-Luc and TARM-Luc cell lines. The traditional synthetic polymer PP is commonly used in the manufacture of food containers and has previously been described as a safe non-EDC-related material (Bang et al., 2012; Chung et al., 2013) with cytotoxic effects (Hwang et al., 2019; Zimmermann et al., 2019). Using a CCK-8 colorimetric assay to measure both cell viability and proliferation, Hwang et al. (2019) found that direct contact of PP particles and PP dispersed in medium had cytotoxic effects on peripheral blood mononuclear and murine macrophages Raw 264.7 cells, concluding chemicals may be released from PP particles when they are dissolved in the medium or in direct contact with cells. Zimmermann et al. (2019) found that food contact materials (FCMs) containing PP, such as yoghurt cups and gummy candy packaging, had varied baseline toxicity and cytotoxicity, with some inducing toxicity in the Microtox assay by inhibiting bioluminescence in the bacterium Aliivbrio fischeri and some inducing cytotoxicity in the AREc32 assay and yeast-based RGA cell lines. The toxicity of PP based FCMs is dependent on the chemical composition of the packaging, additives such as the antioxidant butylated hydroxytoluene and the plasticiser tributyl acetylcitrate have been reported to induce cytotoxicity in ToxCast data (Zimmermann et al., 2019). In the current study, a reduction in metabolic activity was only observed at test conditions of the higher temperature (40°C) and longer period (10 days). Similarly, Alin and Hakkarainen (2011) found that heating PP based FCMs lead to an increase in degradation of antioxidants, which led to increased migration of PP into the food simulants. It would be of future interest to identify the cytotoxic PP leachate(s) responsible, but this will require further work and analytical methods capable of detecting unknowns. Overall, these results indicate that PP is not suitable for FCM development and eggshell/polypropylene (0.1:99.9) may only be suitable for use with more polar solvent like based foods stored and used at low temperatures. Alternatively, these formulation profiles could be developed for non-food contact applications.

In contrast, eggshell/polypropylene (40:60) and eggshell/polypropylene (10:90) which contain lower levels of PP (60% and 90%, respectively) did not reduce metabolic activity at any of the standardised test conditions. In addition, the cytotoxic PP leachate(s) readily leached into MeOH but only into dH₂O when eggshell was at very low concentrations of 0.1% or was not present. This suggests that eggshell may bind the toxicant at a higher affinity in dH₂O when compared to in MeOH. This specific binding could provide a beneficial clean up tool in many packaging formulations. Indeed it would be very interesting future work to add eggshell to a polymer known to leach EDCs and assess if the introduction of calcium carbonate can bind and clean up the EDCs.

Most plastic products leach estrogenic chemicals, highlighting the potential for endocrine disrupting leachates to be present in food or drink products (Yang et al., 2011; Zimmermann et al., 2020). The current study demonstrated that after 10 days of migration testing, four of the test samples presented estrogenic activity as outlined in Table 4, including poultry litter ash/polypropylene (10:90), poultry feather based polymer, eggshell/polypropylene (40:60) and PLA. Whilst it is desirable to have zero hormonal activity, detected estrogenic activity was at low levels, ranging from 2.67% to 5.47% relative response in the ER RGA with corresponding EEQ’s ranging from 0.26 to 0.50 ng/ml. Assuming that the average body weight of an adult is 70 kg, the exposure level of estrogenic compounds via food or drink packaged in the bio-based polymers can be calculated (EFSA, 2012; Plotan et al., 2013). Plotan et al. (2014) calculated the exposure levels to estrogenic compounds based on the concentration of estrogen agonist activity (ng/kg body weight (bw)/day) = equivalent E2 concentration (ng/g) x dilution factor x daily dose (g/day)/average body weight (ABW) (kg). However, the daily doses of the products are unknown as the bio-based polymers are in the developmental stage and have not yet been used in the industry. Therefore, assuming an adult consumed an average daily intake of 500 g or 500 ml of food or drink packaged in the bio-based polymers, the exposure level of estrogenic compounds ingested can be calculated, ranging from 1.86–3.57 ng/kg bw/day (Table 5). These exposure levels are lower than the typical adult eating an omnivorous diet, which was estimated as 23 ng/person/day EEQ of endogenous estrogens. However, this is an estimated value which does not take into consideration the contribution of phytoestrogens or other estrogens which have no estrogenic potency or occurrence data available (Caldwell et al., 2010; Plotan et al., 2013). JECFA (2000) established that the ADI for E2 is up to 50 ng/kg bw/day. Therefore, it may be concluded that based on the assumed exposure level of estrogenic compounds ingested, the level of estrogenic activity of the bio-based polymer leachates do not pose a safety risk for consumers. Similarly to our findings suggesting that using inorganic waste additives such as eggshell can remove cytotoxic leachates from polymers such as PP, it may also be possible in the future to manipulate biopolymer recipes to reduce or completely eliminate this low level hormonal activity.

The estrogenic activity detected in the test samples was minimal compared to the estrogenic activity of BPA which has been identified to leach out of plastic products (Kuruto-Niwa et al., 2005; Grignard et al., 2012; Bittner et al., 2014; Hu et al., 2016). The European Commission set a specific migration limit for BPA; 600 μg/kg of food before banning the use of BPA in the manufacture of polycarbonate infant feedings bottles (European Commission, 2004; 2011b). BPA was reported to migrate out of polycarbonate plastic baby bottles at a migration level of 1.83 μg/kg (0.008 µM) (Simoneau et al., 2011). Le et al. (2008) found BPA leaches out of polycarbonate drinking bottles at room temperature at a concentration of 1 ng/ml (0.004 µM), however, after exposure to 100°C the concentration of BPA (7.67 ng/ml; 0.03 µM) leaching out of plastic bottles increased. BPA at relative leaching concentrations may be associated with hormone responsive breast cancers which are likely mediated by BPA’s estrogenic activity; activating ERα-mediated signalling and subsequently alter response in gene transcription and expression in breast epithelial cells (Weng et al., 2010). In addition, an association between BPA (1.66 ng/ml) at relative leaching concentrations and the risk of recurrent miscarriages in humans was identified by Shen et al. (2015). Previous work (Harper, 2020) has demonstrated that BPA (0.002–20 µM), bisphenol-F (BPF) (0.002–200 µM) and bisphenol-S (BPS) (0.002–20 µM) exhibited estrogenic activity at concentrations relevant to the tolerable daily intake (TDI) of 4 µg BPA/kg body weight/day for a 70 kg person (dose equates to ∼0.02 µM) (EFSA, 2015). The estrogenic responses varied from 14.91% to 54.23% for BPA, 17.1%–131.2% for BPF and 16.23%–79.22% for BPS with corresponding EEQ’s ranging from 1.354 to 7.838 ng/ml for BPA, 1.58–736.47 ng/ml for BPF and 1.49–22.83 ng/ml for BPS. The EEQ calculation allows the estrogenic potency of the compounds to be determined (Vega-Morales et al., 2013). Previous studies have revealed that BPA, BPF and BPS in vitro have similar estrogenic potencies which are 100,000-fold less potent than 17β–estradiol (Grignard et al., 2012; Rochester and Bolden, 2015; Mesnage et al., 2017). In comparison to the bisphenols, the test samples which induced a weak estrogenic response in the current study, had significantly lower estrogenic potencies through receptor activity, ranging from 3 to ∼2800 times less potent. The minimal estrogenic activity and potency of the bio-based polymer leachates investigated in the current study may suggest that the bio-based polymers could be a safer alternative to plastic packaging using compounds such as BPA.

Poultry feather based bio-polymer is composed of propanol glycol (28%), a catalyst included in the recipe (sodium sulphite, 1–2%) and the protein keratin which contains high amounts of the amino acid cysteine with thiol groups (Shalini and Singh, 2009; Xu et al., 2017). The MMV-Luc cell line is specific for the detection of ER transcriptional activation and endogenously expresses both Erα and Erβ, but predominantly Erα (Willemsen et al., 2004; Eze et al., 2019). The ER is a DNA-binding protein that functions as a transcription factor in response to estrogen activation of the protein. The DNA-binding domain of the receptor contains two zinc fingers; each finger has a zinc atom coordinated to 4 cysteines (Garcia-Morales et al., 1994). Therefore, the cysteine and thiol groups in the keratin protein may disrupt the hormone-binding domain of Erα (Zhou et al., 2013). Estrogenic activity of the poultry feather based bio-polymer was only observed at the higher temperature of 40°C. Aranberri et al. (2017) found that chicken feathers used for biodegradable polymeric materials had a large low-temperature endothermic peak at 77°C; this peak ranged from 30–130°C and identified the amount of bound water in the keratin structure which can be referred as the “denaturation” temperature. Therefore, keratin protein may have started to unfold and loosen, subsequently leading to the migration of estrogen-like compounds or additives out of the bio-based polymer (Groot and Bakker, 2016; Lapidus, 2017).

PLA exhibited the strongest estrogenic activity out of all the bio-based polymers tested with an EEQ value of 0.50 ng/ml. Zimmermann et al. (2019) found that the FCMs with PLA, yoghurt cup and vegetable tray, activated the human ERα using a yeast-based estrogen reporter gene assay. In agreement with our findings, Zimmermann et al. (2019) observed low estrogenic activity for PLA, 5.47% and 0.34%, respectively. ToxCast data identified numerous compounds including fatty acids; dodecanoic acid (lauric acid), octadecanoic acid (stearic acid), n-hexadecanoic acid (Palmitic acid) and oleic acid which can leach out of PLA plastics and induce estrogenic activity (Zimmermann et al., 2019). Studies have found fatty acids to modulate ERs by alterations in E2 binding to receptors and/or by cleaving native ERs (Borras and Leclercq, 1992; Prinsloo and van Aswegen, 2002). Borras and Leclercq, 1992 found that unsaturated fatty acids can disrupt ER action through an intracellular interaction between ER and unsaturated fatty acids acting as second messengers in regulation of cellular functions or through transmembrane modulation of phosphokinases and/or phospholipases implicated in ER mechanism of action. Yang et al. (2011) reported estrogenic activity (>15% of relative maximum response to E2) of plastic packaging made from PLA leached into saline using the MCF-7 cell proliferation assay. The results from Yang et al. (2011) correlate to the results found in the current study as PLA leached into more polar solvents, dH₂O and saline, and exhibited estrogenic activity whilst no estrogenic activity was found for PLA tested in less polar solvents, MeOH and ethanol. Polar solvents have been identified to increase the degradation and release of migrant compounds in PLA packaging due to polymer hydrolysis. Lactic acid is the lone monomer in PLA and can migrate either as lactic acid or joined to dimmers and oligomers produced by PLA hydrolysis (Fortunati et al., 2012). There has been a reduction in the application of PLA in food packaging, as PLA has demonstrated low thermal stability and poor barrier properties (Fortunati et al., 2012).

Poultry litter ash/polypropylene (10:90) presented both estrogenic and anti-androgenic hormonal activity. In agreement with our findings, Bittner et al. (2014), identified that PP in the absence of any additional chemicals added during manufacturing of BPA-free polycarbonate plastics exhibited no detectable estrogenic activity in MCF-7 cells. We also found no (anti)androgenic activity for PP in this study. Consequently, the hormonal activity must originate from the poultry litter ash element. The composition of poultry litter ash/polypropylene (10:90) includes polypropylene, potassium sulphate, potassium chloride, calcium phosphate, sodium fluoride and phosphorus pentoxide. Fluoride has previously been identified as an EDC, altering normal endocrine function and response (National Research Council, 2006) and sodium fluoride has also been found to act as an EDC at low doses (Vandenberg et al., 2012), inhibiting insulin secretion plus the parathyroid and thyroid hormones. Poultry litter ash/polypropylene (10:90) also leached a substance which antagonised androgen receptor (AR) activation in the less polar solvent, MeOH, but not the more polar solvent, dH₂O, at 40°C after 10 days of migration testing. Sodium fluoride dissolves more freely in MeOH as the temperature of the solvent increases, therefore, sodium fluoride may have leached out of the poultry litter ash/polypropylene (10:90) into MeOH at 40°C (Germuth, 1931). Sodium fluoride has also been reported to inhibit AR mRNA expression in Sertoli cells in mice, reducing AR protein and gene expression in the testis, therefore, negatively affecting male fertility (Huang et al., 2008). The decrease in AR transcriptional activity observed in the current study may be due to the effect of fluoride on 90-kDa heat-shock protein (HSP90) which may reduce transforming growth factor receptor-2 (TGFR-2) levels, and suppress TGF-β signalling activity, via upregulated cyclins D1 (CCND1) (Knudsen, et al., 1999; Le et al., 2017). Previous studies have also identified that the plastic leachate BPA can directly bind to the AR acting as an antagonist of the AR via inhibiting the actions of endogenous androgens and transcription of AR target genes at concentrations above and below the total daily intake (4 μg/kg body weight/day for a 70 kg person) (Lee et al., 2003; Xu et al., 2005; Sun et al., 2006; Bonefeld-Jørgensen et al., 2007; Teng et al., 2013; Ma et al., 2019). BPA can reduce AR transcriptional activity in mouse Sertoli 15p-1 cells with an IC₅₀ 0.08 µM and can decrease AR transcriptional activity by more than 20% at relative leaching concentrations 0.01 µM (Lee et al., 2003). The antagonistic effects of BPA on the AR may lead to adverse effects on the male reproductive system such as decreasing sperm counts, interfering with sperm motility and reducing reproductive organ weights (Xu et al., 2005). The reduction in AR transcriptional activity following exposure to poultry litter ash/polypropylene (10:90) was less than that of relative leaching concentrations of BPA. Therefore, the bio-based polymers investigated in the current study may potentially be safer alternatives to plastic packaging using compounds such as BPA and potentially safer candidates for further development in the bioplastic industry.

Estrogenic activity was detected in the bio-based polymer with the highest ratio of eggshell to polypropylene (40:60). The composition of chicken eggshell consists of 95% calcium carbonate, eggshell membrane, phosphorous, magnesium, strontium and traces of sodium, zinc, iron, lead, manganese and copper (Mora, 2003; Rovenský et al., 2003; Ummartyotin and Tangnorawich, 2015; Mohammadzadeh et al., 2019; Ajayan et al., 2020). The leachate compound(s) responsible for inducing the estrogenic activity of the poultry eggshell is unknown. Eggshells may contain biologically active compounds such as heavy metals and amino acids rich proteins such as histidine, arginine and aspartic acid (Nakano et al., 2003; Nagamalli et al., 2017). Information regarding the estrogenic activity of the metal strontium is limited. However, previous studies have found that metals such as cadmium, mercury, arsenic, lead, manganese and zinc can affect hormone levels, hormone production, gene transcription activity of steroid receptors and mimic hormone action (Garcia-Morales et al., 1994; Lafuente et al., 2003; Smida et al., 2004; Srivastava et al., 2004; Pine et al., 2005; Bodwell et al., 2006; Agusa et al., 2007; Iavicoli et al., 2009). Some metals have been classified as metalloestrogens due to their ability to interfere with the action of estrogenic hormones (Safe, 2003). Metals and amino acids have the ability to disrupt the action of E2, ERs and DNA due to the participation of zinc-fingers motifs in processes such as ER dimerization and binding with DNA. These zinc-fingers motifs have cysteine and histidine with high affinity for metals. Studies have shown that metals and amino acids are capable of binding to the DNA-binding domain of the ER zinc finger motifs and disrupting the binding capacity of some transcription factors with DNA (Zawia et al., 1998; Reddy and Zawia, 2000; Razmiafshari et al., 2001; Torre et al., 2011). Metals including aluminium, antimony, arsenite, barium, cadmium, chro-mium (Cr(II)), cobalt, copper, lead, mercury, nickel, selenite, tin and vanadate are capable of binding to the ligand binding domain of the Erα, enhancing the agonist activity of E2 to their ERs, therefore, acting as estrogen agonists in test systems including affinity chromatography, MCF7 proliferation and reporter genes (Darbre, 2006). Consequently, the presence of metals in eggshells may have the potential to induce estrogenic activity in the MMV-Luc cell line via a similar action to that of metalloestrogens.

The migration testing employed in the current study was carried out under the worst foreseeable use of the plastic conditions, therefore, the migration of leachates into food simulants is thought to overestimate the actual migration into foods (European Commission, 2011a; Muncke, 2014). Consequently, it should also be noted that in the current study the cytotoxic and hormonal activity of the leachates may be overestimated.

The endocrine disrupting potential of FCMs is dependent on the chemical composition of the plastic packaging, however, manufacturers of plastic products typically do not list their chemical composition (Yang et al., 2011). Polymers and bio-polymers are produced with small quantities of various additives such as antioxidants and plasticisers which are physically but not chemically bound to the polymeric structure, therefore, these additives can leach from the plastic products at very low (e.g., nanomolar to picomolar) concentrations which individually or in combination can cause adverse health effects (Flaws et al., 2020). The manufacturing process of plastic uses many monomers and additives that may exhibit estrogenic activity or potentially decrease nuclear receptor transcriptional activity due to physicochemical properties that enables them to bind to nuclear receptors (Yang et al., 2011). Leaching of additives and monomers from plastics into its contents such as food is often accelerated if the product is exposed to common-use stresses such as ultraviolet radiation in sunlight and/or moist heat via boiling or dishwashing. However, many biodegradable polymers are sensitive to water and moisture so will not be suitable for use in dishwashers. Even when using materials that initially are not known to leach hormonally active compounds, the stresses of manufacturing can change chemical structures or create chemical reactions to convert a non- hormonally active compound into a chemical with endocrine disrupting activity (Yang et al., 2011).

Polarity and solubility properties are important factors for migration due to the interactions between the polymer, migrants and the food simulant. If the migrant has poor solubility in the food simulant it will be retained in the polymer matrix, while a simultaneous effect based on a high affinity between food simulant and the polymer may lead to absorption by the polymer matrix. In addition, absorption of solvents may cause swelling of the polymer matrix, therefore, enlarging the gaps between the molecules, leading to the migration of small particles and degradation products (Zygoura et al., 2011; Fortunati et al., 2012). Results also differed for the same bio-based polymer under the same standardised testing conditions between leaching into more polar dH₂O and less polar MeOH. Therefore, the use of more polar and less polar solvents should be assessed to ensure the ability to detect more polar and less polar leachate activity as plastic containers may hold either type of liquid or a liquid that is a mixture of more polar and less polar solvents (e.g., milk) (Yang et al., 2011).

Solubility, temperature and exposure time are significant factors controlling the migration of leachates. In agreement with our study, Muncke (2014) also found that the extent of chemical migration from food packaging into food and food simulant is highly correlated with temperature and duration of exposure. The migration of hormonally active leachates out of the poultry feather based polymer, eggshell/polypropylene (40:60), PLA and poultry litter ash/polypropylene (10:90) occurred after migration testing was implemented under conditions representing long term storage at room temperature or below or heating up to 100°C. In contrast, under conditions representing short term storage at room temperature or below, none of the bio-based polymers appeared to leach substances with hormonal activity.

Endocrine disrupting and toxic packaging leachates are a major health concern for the consumer and environment. Whilst all the bio-based polymers investigated in this study presented either negligible or less hormonal leachate activity in comparison to BPA, ideally it is desirable to develop packaging with zero toxicological hazard. It should also be noted that whilst the current study carried out initial investigations at the level of estrogen, androgen and progestogen nuclear receptor transcriptional activity, further investigations should also investigate other endocrine-related pathways, acting through different mechanisms of action.

The traditional petroleum based polymer PP did not present hormonal activity but did leach a potentially toxic compound which reduced metabolic activity within our estrogen responsive MMV-Luc cell line. The bio-based polymers eggshell/polypropylene (10:90), PHB, PBAT, TPS, PBS, PCL and PLA/PBAT blends did not present any hormonal or cellular cytotoxicity activity at any of the standardised testing conditions and thus may be potentially safer candidates for further development for the food contact packaging industry. The poultry feather based polymer and PLA and the formulations eggshell/polypropylene (40:60) and poultry litter ash/polypropylene (10:90) presented estrogenic activity, with the latter also presenting anti-androgenic activity. These developmental polymers and their formulations may be more suitably streamed towards further development in non-food packaging areas.

The major finding in this study is that altering formulations and their constituent levels may eradicate toxic and or hormonal activity. For example, the estrogenic and toxic activity observed in the eggshell and PP formulations was eliminated in the eggshell/polypropylene (10:90) formulation which consists of relatively lower eggshell and PP concentration as compared to the other formulations or PP alone. The mechanism by which this happens remains to be elucidated but effects may be diluted with compound reduction or alternatively eggshell may bind and eliminate the toxic elements of PP. Similarly the estrogenic activity observed in PLA was eliminated in PLA/PBAT blends.

In conclusion, this study is the first to demonstrate that combining toxicological analysis with the development of new bio-based and or manipulating mixtures of inorganic additives with a traditional synthetic polymer may eradicate hormonal and toxic leachates from packaging material, thus providing a useful approach for safer packaging development.