No abstract

With the phasing out of long-chain per- and polyfluoroalkyl substances (PFASs), production of a wide variety of alternative PFASs has increased to meet market demand. However, little is known about the bioaccumulation potential of these replacement compounds. Here, we developed a modeling workflow that combines molecular docking and molecular dynamics simulation techniques to estimate the relative binding affinity of a total of 15 legacy and replacement PFASs for human and rat liver-type fatty acid binding protein (hLFABP and rLFABP). The predicted results were compared with experimental data extracted from three different studies. There was good correlation between predicted free energies of binding and measured binding affinities, with correlation coefficients of 0.97, 0.79, and 0.96, respectively. With respect to replacement PFASs, our results suggest that EEA and ADONA are at least as strongly bound to rLFABP as perfluoroheptanoic acid (PFHpA), and as strongly bound to hLFABP as perfluorooctanoic acid (PFOA). For F-53 and F-53B, both have similar or stronger binding affinities than perfluorooctanesulfonate (PFOS). Given that interactions of PFASs with proteins (e.g., LFABPs) are important determinants of bioaccumulation potential in organisms, these alternatives could be as bioaccumulative as legacy PFASs, and are therefore not necessarily safer alternatives to long-chain PFASs.

Physiologically based pharmacokinetic (PBPK) modeling is a powerful in silico tool that can be used to simulate the toxicokinetics and tissue distribution of xenobiotic substances, such as perfluorooctanoic acid (PFOA), in organisms. However, most existing PBPK models have been based on the flow-limited assumption and largely rely on in vivo data for parametrization. In this study, we propose a permeability-limited PBPK model to estimate the toxicokinetics and tissue distribution of PFOA in male rats. Our model considers the cellular uptake and efflux of PFOA via both passive diffusion and transport facilitated by various membrane transporters, association with serum albumin in circulatory and extracellular spaces, and association with intracellular proteins in liver and kidney. Model performance is assessed using seven experimental data sets extracted from three different studies. Comparing model predictions with these experimental data, our model successfully predicts the toxicokinetics and tissue distribution of PFOA in rats following exposure via both IV and oral routes. More importantly, rather than requiring in vivo data fitting, all PFOA-related parameters were obtained from in vitro assays. Our model thus provides an effective framework to test in vitro-in vivo extrapolation and holds great promise for predicting toxicokinetics of per- and polyfluorinated alkyl substances in humans.

In this study, we determined the abundance of 8 antibiotics (3 tetracyclines, 4 sulfonamides, and 1 trimethoprim), 12 antibiotic-resistant genes (10 tet, 2 sul), 4 antibiotic-resistant bacteria (tetracycline, sulfamethoxazole, and combined resistance), and class 1 integron integrase gene (intI1) in the effluent of residential areas, hospitals, and municipal wastewater treatment plant (WWTP) systems. The concentrations of total/individual targets (antibiotics, genes, and bacteria) varied remarkably among different samples, but the hospital samples generally had a lower abundance than the residential area samples. The WWTP demonstrated removal efficiencies of 50.8% tetracyclines, 66.8% sulfonamides, 0.5 logs to 2.5 logs tet genes, and less than 1 log of sul and intI1 genes, as well as 0.5 log to 1 log removal for target bacteria. Except for the total tetracycline concentration and the proportion of tetracycline-resistant bacteria (R (2) = 0.330, P < 0.05), there was no significant correlation between antibiotics and the corresponding resistant bacteria (P > 0.05). In contrast, various relationships were identified between antibiotics and antibiotic resistance genes (P < 0.05). Tet (A) and tet (B) displayed noticeable relationships with both tetracycline and combined antibiotic-resistant bacteria (P < 0.01).

A recent OECD report estimated that more than 4000 per- and polyfluorinated alkyl substances (PFASs) have been produced and used in a broad range of industrial and consumer applications. However, little is known about the potential hazards (e.g., bioactivity, bioaccumulation, and toxicity) of most PFASs. Here, we built machine-learning-based quantitative structure–activity relationship (QSAR) models to predict the bioactivity of those PFASs. By examining a number of available molecular data sets, we constructed the first PFAS-specific database that contains the bioactivity information on 1012 PFASs for 26 bioassays. On the basis of the collected PFAS data set, we trained 5 different machine learning models that cover a variety of conventional models (e.g., random forest and multitask neural network (MNN)) and advanced graph-based models (e.g., graph convolutional network). Those models were evaluated based on the validation data set. Both MNN and graph-based models demonstrated the best performance. The average of the best area-under-the-curve score for each bioassay is 0.916. For predictions on the OECD list, most of the biologically active PFASs have perfluoroalkyl chain lengths less than 12 and are categorized into fluorotelomer-related compounds and perfluoroalkyl acids and their precursors.