Optimization of an in vitro method to measure the bioaccessibility of polybrominated diphenyl ethers in dust using response surface methodology

Bioaccessibility of halogenated flame retardants and organophosphate esters in settled dust: Influences of specific dust matrices from informal e-waste and end-of-life vehicle processing areas in Vietnam

Bioaccessibility of halogenated flame retardants (HFRs) and organophosphorus esters (OPEs) is necessarily investigated to provide more accurate risk assessment and information about absorption behavior of these pollutants. In this study, total and bioaccessible concentrations of HFRs (including legacy and alternative substances) and OPEs were determined in settled dust samples collected from Vietnamese e-waste and end-of-life vehicle (ELV) processing areas. Concentrations of both HFRs and OPEs were significantly higher in the e-waste dust than ELV dust. Bioavailability of HFRs and OPEs in dust was determined by using an in vitro assay with human-simulated digestive fluids, dialysis membrane, and Tenax® TA sorptive sink. Bioaccessibility of HFRs was markedly lower than that of OPEs, which could be largely due to higher hydrophobicity of HFRs compared to OPEs. Bioaccessibility of almost hydrophobic compounds were markedly lower in the e-waste dust (containing micronized plastic debris) than in the ELV dust (containing oily materials), suggesting the influence of specific dust matrices on pollutant bioaccessibility. Although the daily uptake doses of selected HFRs and OPEs from dust were markedly higher in the e-waste sites compared to the ELV sites, the direct exposure risk was not significant. Our results suggest that bioaccessibility can partly explain the differences between dust and uptake profiles, which may relate to accumulation profiles of HFRs and OPEs in human samples.

Bioaccessibility of per- and polyfluoroalkyl substances in food and dust: Implication for more accurate risk assessment

Humans are exposed to per- and polyfluoroalkyl substances (PFASs) mainly through oral exposure route, while little is known about their bioaccessibility (BC) in oral matrices. Here, the BC of 13 PFASs in simulated vegetable (VFs) and animal foods (AFs) as well as indoor dust was investigated using a physiology-based extraction test. The BC of PFASs in the AFs (78.5 ± 13.6 %) was distinctly higher than that in the VFs (60.6 ± 13.4 %), because high-saturated and long-chain fatty acids in the animal fat favored formation of more stable micelles. The BC of most long-chain PFASs was positively correlated with the protein content while negatively correlated with the carbohydrate content in the foods. The BC of polyfluoroalkyl phosphate diesters was negatively correlated with the lipid content. The BC of the very long-chain PFASs in the foods was 2.42-6.02 times higher than that in the dust, which might be attributed to their strong sequestration in dust. With the increase in bile salt concentration, the BC of PFASs in food increased and then remained constant, which was related to the changes in fatty acids and stability of the formed micelles. Comparing with the previous results obtained from animal study, the BC obtained in this study has the potential to predict PFAS bioavailability in food.

In vitro assessment of polychlorinated biphenyl bioaccessibility in meat: Influence of fat content, cooking level and consumer age on consumer uptake

In a risk assessment perspective, this work aims to assess the bioaccessibility of PCBs in meat. A standardised in vitro static digestion protocol was set up and coupled with extraction, clean-up and GC × GC-ToF/MS multianalyte method to monitor the fate of PCBs in meat during digestion. Starting with spiked meat, PCB bioaccessibility in 11% fat medium-cooked meat varied in adults from 20.6% to 30.5% according to congeners. PCB bioaccessibility increased to 44.2-50.1% in 5% fat meat and decreased to 6.2-9.1% and to 14.6-19.4% in digestion conditions mimicking infants and elderly, respectively. Intense cooking also decreased PCB bioaccessibility to 18.0-26.7%. Bioaccessibility data obtained with spiked meat were validated with measurements carried out in incurred meat samples. Finally, mean uptake distributions are obtained from a modular Bayesian approach. These distributions feature a lower mode when the fat content is higher, the meat is well-done cooked, and the consumers are older.

An interlaboratory evaluation of the variability in arsenic and lead relative bioavailability when assessed using a mouse bioassay

Animal bioassays have been developed to estimate oral relative bioavailability (RBA) of metals in soil, dust, or food for accurate health risk assessment. However, the comparability in RBA estimates from different labs remains largely unclear. Using 12 soil and soil-like standard reference materials (SRMs), this study investigated variability in lead (Pb) and arsenic (As) RBA estimates employing a mouse bioassay in 3 labs at Nanjing University, University of Jinan, and Shandong Normal University. Two performances of the bioassay at Nanjing University in 2019 and 2020 showed reproducible Pb and As RBA estimates, but increasing the number of mouse replicates in 2020 produced more precise RBA measurements. Although there were inter-lab variations in diet consumption rate and metal accumulation in mouse liver and kidneys following SRM ingestion due to differences in diet composition, bioassays at 3 labs in 2019 yielded overall similar Pb and As RBA estimates for the 12 SRMs with strong linear correlations between each 2 of the 3 labs for Pb (R² = 0.95-0.98 and slope = 0.85-1.02) and As RBA outcomes (R² = 0.46-0.86 and slope = 0.56-0.79). The consistency in RBA estimates was attributed to the relative nature of the final bioavailability outcome, which might overcome the inter-lab variation in diet consumption and metal uptake in mice. These results increased the confidence of use of mouse bioassays in bioavailability studies.

A review on in-vitro oral bioaccessibility of organic pollutants and its application in human exposure assessment

Generally, human oral exposure assessments of contaminants have not considered the absorption factor in the human gastrointestinal tract, thus overestimating human exposure and associated health risk. Currently, more researchers are adding the absorption factor into human exposure assessment, and bioaccessibility measured by in-vitro methods is generally replacing bioavailability for estimation because of the cheap and rapid determination. However, no single unified in-vitro method is used for bioaccessibility measurement of organic pollutants, although several methods have been developed for these pollutants and have shown good in vitro-in vivo correlation between bioaccessibility and bioavailability. The present review has focused on the development of in-vitro methods, validation of these methods through in-vivo assays, determination of factors influencing bioaccessibility, application of bioaccessibility in human exposure assessment, and the challenges faced. Overall, most in-vitro methods were validated using bioavailability, and better in vitro-in vivo correlations were obtained when absorption sinks were added to the digestion solution to mimic dynamic absorption of organic chemicals by small intestine. Incorporating bioaccessibility into the estimation of human exposure by oral ingestion significantly decreases the estimated exposure dose. However, more investigations on bioaccessibility of hydrophobic organic compounds are urgently needed because many challenges for in-vitro methods remain to be overcome.

In vitro oral and inhalation bioaccessibility of hydrophobic organic contaminants (HOCs) in airborne particles and influence of relevant parameters

The bioaccessibility of environmental contaminants has been assessed widely using in vitro simulation; however, the physiological parameters used vary greatly. In this study, we assessed the influence of various physiological parameters and food material on the oral or inhalation bioaccessibility of PM_2.5-bound hydrophobic organic contaminants (HOCs), including halogenated flame retardants (HFRs), organophosphorus flame retardants (OPFRs), and polycyclic aromatic hydrocarbons (PAHs). The results showed that physiologically based pepsin and pancreatin have a small influence on the HOC liberation from particles. The bioaccessibility increased dramatically when the bile salt concentrations exceeding the critical micelle concentration, and application of porcine bile salts probably lead to underestimated bioaccessibility. Protein and carbohydrates significantly increased the bioaccessibility of most HOCs, while a significant bioaccessibility reduction was caused by green tea. The bioaccessibility of most HOCs was not promoted by liquor under normal physiological condition, but was significantly promoted under fast condition. Long residence time of PM_2.5 in the lung (15 days) would result in higher mobilization of PAHs into the lung fluid than short time (one day). However, the inverse time-dependence for OPFRs suggests degradation in the lung fluid. A mechanism of hydrolysis of organophosphorus ester is hypothesized, and the half lives ranged from 17 to 90 days.

Relationships between the bioavailability of polybrominated diphenyl ethers in soils measured with female C57BL/6 mice and the bioaccessibility determined using five in vitro methods

Several in vitro methods for simulating human gastrointestinal digestion have been validated for predicting the bioavailability of heavy metals, but the methods for successfully predicting the bioavailability of organic pollutants are still limited. In this study, we used an adapted fasting in vitro digestion method (Fa-VDM) from the Simulator of the Human Intestinal Microbial Ecosystem and four other in vitro methods comprising In Vitro Gastrointestinal, a physiologically-based extraction test, the unified BARGE method, and Deutsches Institut für Normung e.V. in order to measure the bioaccessibility of polybrominated diphenyl ethers (PBDEs) in soils from an e-waste dismantling town, China, with a Standard Reference Material (SRM2585) as the control. Furthermore, the bioaccessibility data were compared with the bioavailability measured using female C57BL/6 mice. The bioavailability of PBDEs in the soils and SRM2585 were 1.7% to 38.1% and 3.9% to 48.8%, respectively, and the bioaccessibility determined using Fa-VDM were 1.6-55.4% and 6.7-32.1%. There were negative and parabolic correlations between octanol/water partition coefficient for PBDEs and the bioavailability and bioaccessibility, respectively, whereas the H/C ratios and organic matter contents of the soils did not correlate with them. The bioaccessibility data determined by Fa-VDM were generally higher than those obtained using the other four methods, mainly due to the higher bile concentration and larger liquid to solid ratio in the digestion solution in Fa-VDM. There was a significant linear relationship between the results according to the in vivo and in vitro method of Fa-VDM where the slopes varied from 0.83 to 1.16 (R² > 0.73) and intercepts from 0.3%-7.7% for BDE47, 99, 100, and 153 measured using Fa-VDM, thereby indicating that the bioaccessibility assessed by this method can potentially be used to predict the bioavailability of moderately brominated congeners in soils.

Oral bioaccessibility of semi-volatile organic compounds (SVOCs) in settled dust: A review of measurement methods, data and influencing factors

Many semi-volatile organic compounds (SVOCs), suspected of reprotoxic, neurotoxic or carcinogenic effects, were measured in indoor settled dust. Dust ingestion is a non-negligible pathway of exposure to some of these SVOCs, and an accurate knowledge of the real exposure is necessary for a better evaluation of health risks. To this end, the bioaccessibility of SVOCs in dust needs to be considered. In the present work, bioaccessibility measurement methods, SVOCs' oral bioaccessibility data and influencing factors were reviewed. SVOC bioaccessibilities (%) ranged from 11 to 94, 8 to 100, 3 to 92, 1 to 81, 6 to 52, and 2 to 17, for brominated flame retardants, organophosphorus flame retardants, polychlorobiphenyls, phthalates, pesticides and polycyclic aromatic hydrocarbons, respectively. Measurements method produced varying results depending on the inclusion of food and/or sink in the model. Characteristics of dust, e.g., organic matter content and particle size, also influenced bioaccessibility data. Last, results were influenced by SVOC properties, such as octanol/water partition coefficient and migration pathway into dust. Factors related to dust and SVOCs could be used in prediction models. To this end, more bioaccessibility studies covering more substances should be performed, using methods that are harmonized and validated by comparison to in-vivo studies.

Does the source migration pathway of HBCDs to household dust influence their bio-accessibility?

A study was conducted to assess the human bioaccessibility of dust contaminated with hexabromocyclododecane (HBCD) via two migration pathways a) volatilisation with subsequent partitioning to dust particles, and b) abrasion of treated textile fibres directly to the dust. This was achieved using previously developed experimental chamber designs to generate dust samples contaminated with HBCDs emitted from a HBCD treated textile curtain. The generated dust samples were exposed to an in vitro colon extended physiologically based extraction test (CE-PBET). The bioaccessibility of the HBCDs which were incorporated within dust as a result of volatilisation from the curtain material with subsequent partitioning to dust was higher than in dusts contaminated with HBCDs via abrasion of the curtain (35% and 15% respectively). We propose this occurs due to a stronger binding of HBCDs to treated fabric fibres than that experienced following volatilisation and sorption of HBCDs to dust particles.

Distribution pattern of legacy and "novel" brominated flame retardants in different particle size fractions of indoor dust in Birmingham, United Kingdom

This study investigates the particle size distribution of eight polybrominated diphenyl ethers (PBDEs) and five "novel" brominated flame retardants (NBFRs) in settled house dust. Elevated surface dust (ESD) and floor dust (FD) were collected from 5 homes in Birmingham, UK, yielding a total of 10 samples. Each sample was fractionated into three different particle sizes: 125-250 μm (P1), 63-125 μm (P2) and 25-63 μm (P3). Non-fractionated bulk dust samples (BD) were also analysed. BDE-209 predominated, comprising an average 74.3%, 77.3%, 69.2%, and 62.7% ΣBFRs of BD, P1, P2 and P3 respectively. Σ5NBFRs contributed 24.2%, 21.5%, 29.0% and 35.3% ΣBFRs, while Σ7tri-hepta-BDEs represented 1.5%, 1.2%, 1.7%, and 2.0% ΣBFRs. BEH-TEBP was the predominant NBFR contributing 76.9%, 75.1%, 83.1%, and 83.9% ΣNBFRs in BD, P1, P2 and P3 respectively; followed by DBDPE which contributed 20.1%, 21.9%, 14.1% and 13.9% ΣNBFRs. EH-TBB, BTBPE and PBEB were the least abundant NBFRs. Concentrations of Σ7tri-hepta-BDEs and BEH-TEBP in P3 exceeded significantly (P < 0.05) those in P2, with those in P2 exceeding significantly those in P1. In contrast, no significant differences were found between concentrations of BDE-209, EH-TBB, BTBPE, and DBDPE in different particle size fractions. Concentrations of Σ7tri-hepta-BDEs, BDE-209, and BEH-TEBP in ESD exceeded significantly those in FD (P < 0.05). Normalising BFR concentrations to organic carbon content, did not alter these findings. This suggests that differences in BFR concentrations between different particle size fractions are caused by variations in particle surface area to volume ratio, rather than by variations in organic carbon content.

Bioaccessibility of polychlorinated biphenyls in workplace dust and its implication for risk assessment

Human exposure to bioaccessible PCBs via indoor dust is limited around the world. In the present study, the workplace dust sample from commercial office, hospital, secondary school, shopping mall, electronic factory and manufacturing plant were collected from Hong Kong for PCBs analyses. Total PCBs concentrations ranged from 46.8 to 249 ng g(-1), with a median of 107 ng g(-1). Manufacturing plant showed the highest concentration among all of sampling sites. PCB 77 was found as the dominant congeners. The bioaccessibility of PCBs in small intestinal juice ranging from 8.3% to 26.0% was significantly higher than that in gastric condition, ranging from 4.8% to 12.4%. In addition, significant negative correlations (p<0.05) were observed between KOW and bioaccessibility for all workplace dust samples. Risk assessment indicated that the averaged daily dose of dioxin-like PCBs via non-dietary intake of workplace dust, considering the bioaccessibility of PCBs, were much lower than the TDI of dioxins (2.3 pg WHO-TEQ kgbw(-1)d(-1)) established by Joint FAO/WHO Expert Committee on Food Additives.

Concentrations and seasonal variations of polybrominated diphenyl ethers (PBDEs) in in- and out-house dust and human daily intake via dust ingestion corrected with bioaccessibility of PBDEs

The objective of this study was to investigate the concentrations, seasonal variations, bioaccessibility, and associated human daily intake of polybrominated diphenyl ethers (PBDEs) in in- and out-house dust collected in Shanghai, China. The PBDE concentrations varied from 131.6 to 3,886.7 ng g(-1) (with an average of 948.2 ng g(-1)) in in-house dust and from 8.7 to 3,116.3 ng g(-1) (with an average of 290.8 ng g(-1)) in out-house dust during four seasons. The PBDE concentrations in the autumn were the lowest for both in- and out-house dust. Among the detected PBDEs, BDE209 was the predominant congener, accounting for more than 80% of the total PBDE amounts. The bioaccessibility of PBDEs, measured using a simulation system of human gastrointestinal tract, was determined as 14.2-66.4% depending on individual PBDE congeners and showed significant negative correlations with organic matter in dust. After corrected with the bioaccessibility of PBDEs, the human daily intake of PBDEs via dust ingestion was calculated to be 0.4-21.4 and 4.3-40.6 ng day(-1) for an average adult and child in Shanghai, respectively. The values were much lower than most estimates in the literature, in which the bioaccessibility of PBDEs were not taken into account, suggesting that the intake of PBDEs may have been overestimated.