Evaluation and validation of methodologies for the extraction of per- and polyfluoroalkyl substances (PFASs) in serum of birds and mammals

Association and binding interaction between per- and polyfluoroalkyl substances and maternal thyroid hormones: A case study based on a prospective birth cohort in Wuxi, China

BACKGROUND AND AIM

The relationship between prenatal exposure to per- and polyfluoroalkyl substances (PFASs), a well-known endocrine disruptor, and thyroid hormones (THs) levels remains unclear. Therefore, this study aimed to investigate this relationship in a birth cohort during the second trimester.

METHOD

This prospective study included 562 pregnant women in the Wuxi Birth Cohort from 2019 to 2021 and quantified the serum concentrations of 23 PFASs and 5 THs. Multiple statistical models were used to assess the associations between individual or combined PFASs concentrations and THs, while molecular docking simulated the interactions between PFASs and four thyroid-related proteins.

RESULTS

The median concentration of ∑23PFASs was 71.91 ng/mL, with perfluorovaleric acid (PFPeA) (18.13 ng/mL) emerging as the predominant PFAS. Most PFASs were negatively associated with maternal free thyroxine (FT4) and thyroid-stimulating hormone (TSH) levels, whereas perfluorobutane sulfonate (PFBS) was positively correlated with TSH levels. A similar trend was observed in the weighted quantile sum (WQS) model, in which combined PFASs exposure was inversely associated with the FT4 and TSH levels. Molecular docking results showed that compared with TH natural ligand thyroxine (T4), perfluorooctanoic acid (PFOA), perfluorooctane sulfonate (PFOS), and 6:2 chlorinated polyfluorinated ether sulfonate (6:2 Cl- PFESA) exhibited relatively high binding affinity with thyroid-related proteins (-6.6 to -9.8 kcal/mol vs. T4: -5.6 to -8.6 kcal/mol). Furthermore, PFASs with medium chain lengths and sulfonic acid groups exhibited enhanced protein-binding properties.

CONCLUSION

PFASs exposure may affect THs homeostasis during pregnancy. Moreover, different types and concentrations of PFASs have different effects on THs in the maternal serum.

A High Efficiency Method for the Extraction and Quantitative Analysis of 45 PFAS in Whole Fish

This study describes and validates a new method for extracting perfluoroalkyl and polyfluoroalkyl substances (PFAS) from whole-body fish tissue, demonstrates that freeze-dry preservation of tissue conserves bioaccumulative PFAS, and details a method demonstration on Lake Michigan fish. While fish filets are more commonly analyzed for their significance to human health, whole fish are useful to determine ecological impacts, but published methods such as EPA 1633 do not produce reliable results for this more challenging matrix. Here we show that lipid removal technology produces clean extracts without the need for solid-phase extraction or evaporative concentration, which often lead to loss of some PFAS. This method achieves an accuracy of 96 ± 9% for the detection of 45 PFAS while also offering benefits of a simple procedure, reduced processing time, and decreased waste generation compared to multistep cleanup and concentration methods. A test of freeze-drying demonstrated that compounds detected in Great Lakes fish were retained, but volatile compounds including sulfonamide precursors and ethanols were lost. To demonstrate field performance, the entire method was applied to whole-fish composites from Lake Michigan. Results from these samples reveal that the PFAS concentration was driven by collection location, while the distribution of PFAS was dictated by fish species.

A novel high-throughput quantitative method for the determination of per- and poly-fluoroalkyl substances in human plasma based on UHPLC-Q/Orbitrap HRMS coupled with isotope internal standard

A novel method for the quantitative analysis of 56 per- and polyfluoroalkyl substances (PFASs) in human plasma was established on the basis of ultrahigh performance liquid chromatography tandem quadrupole Orbitrap high-resolution mass spectrometry (UHPLC-Q/Orbitrap HRMS) in combination with accurate customized mass databases and isotopic internal standards. A streamlined, high-throughput, and high-recovery (RE) sample pretreatment method was developed. The method's performance was evaluated in terms of linearity, limit of quantification, RE, repeatability, reproducibility, and matrix effect. The proposed method was applied in the simultaneous analysis of 56 PFASs in human plasma, and its results demonstrated high sensitivity, accuracy, and precision. The optimized method was implemented to analyze PFASs in 135 plasma samples, and 12 components were detected. The comparative analysis of the results from 135 plasma samples with domestic and international studies revealed elevated contents of PFOA, PFOS, PFBA, and PFTrDA, the moderate amounts of PFHxS, PFUdA, PFBS, and PFHpS, and the low concentrations of PFNA and PFDA. Notably, GenX was detected in human plasma for the first time. This finding suggests that the study region is contaminated with this substance. Correlation analysis revealed a strong relationship among PFNA, PFDA, and PFUdA, implying that these substances may have similar exposure sources.

Transplacental transfer of perfluorinated and poly-fluorinated substances in maternal-cord serum and association with birth weight: A birth cohort study, China

Although the effects of traditional perfluorinated and polyfluorinated substances (PFASs) exposure have been extensively explored, research on novel PFASs remains limited, and there is a lack of data regarding their placental transfer and fetal impact. Herein, we aimed to examine maternal and fetal PFASs exposure levels, placental transfer efficiency (TTE), and the consequences of prenatal exposure on birth weight. The study included 214 mother-child pairs recruited in Wuxi birth cohort from 2019 to 2021. Twenty-three PFASs were quantified in maternal serum during the second trimester and umbilical serum during delivery. Median concentrations of ∑23PFASs in maternal and cord sera were 9.34 and 6.88 ng/mL, respectively. The novel alternatives exhibited elevated levels of maternal and fetal exposure, such as perfluorovaleric acid (PFPeA, 2.00 ng/mL and 1.66 ng/mL, respectively) and perfluorohexane sulfonate (PFHxS, 1.77 and 1.14 ng/mL, respectively). With increasing carbon chain length, the TTE of perfluorocarbonic acid (PFCAs) displayed a pattern of initially decreasing before subsequently increasing, with novel alternatives exhibiting a relatively high TTE. Multiple linear regression showed that exposure to perfluorobutane sulfonate (PFBS) and PFPeA in cord serum positively correlated with the birth weight of female infants (β = 231.04 g, 95% confidence interval [CI]: 21.73-440.36; β = 121.26 g, 95% CI: 29.51-213.00). No nonlinear relationship was observed between cord serum PFASs and birth weight. The weighted quantile sum (WQS) regression analysis has reaffirmed that PFPeA and PFBS were predominant contributors to the positive correlation observed between the mixture of PFASs and birth weight. Our findings suggest that novel PFASs may exhibit a heightened susceptibility for transplacental transfer and that exposure to PFBS and PFPeA during pregnancy could be linked to increased birth weight.

Characterization of Chemical Exposome in A Paired Human Preconception Pilot Study

Parental preconception exposure to synthetic chemicals may have critical influences on fertility and reproduction. Here, we present a robust LC-MS/MS method covering up to 95 diverse xenobiotics in human urine, serum, seminal and follicular fluids to support exposome-wide assessment in reproductive health outcomes. Extraction recoveries of validated analytes ranged from 62% to 137% and limits of quantification from 0.01 to 6.0 ng/mL in all biofluids. We applied the validated method to a preconception cohort of Australian couples (n = 30) receiving fertility treatment. In total, 36 and 38 xenobiotics were detected across the paired biofluids of males and females, respectively, including PFAS, parabens, organic UV-filters, plastic additives, antimicrobials, and other industrial chemicals. Results showed 39% of analytes in males and 37% in females were equally detected in paired serum, urine, and reproductive fluids. The first detection of the sunscreen ingredient avobenzone and the industrial chemical 4-nitrophenol in follicular and seminal fluids suggests they can cross both blood-follicle/testis barriers, indicating potential risks for fertility. Further, the blood-follicle transfer of perfluorobutanoic acid, PFOA, PFHxS, PFOS, and oxybenzone corroborate that serum concentrations can be reliable proxies for assessing exposure within the ovarian microenvironment. In conclusion, we observed significant preconception exposure to multiple endocrine disruptors in couples and identified potential xenobiotics relevant to male and female fertility impairments.

Development and validation of the UHPLC-MS/MS method for the quantitative determination of 25 PFAS in dried blood spots

Per- and polyfluoroalkyl substances (PFAS) are anthropogenic fluorine-containing compounds largely used in industrial and consumer applications. They tend to bioaccumulate in the human body after intake from various sources in daily life. Following repeated exposure to PFAS, a broad range of adverse health outcomes has been reported. Consequently, monitoring PFAS levels in human blood is of paramount importance for public health policies. In contrast with traditional venipuncture, dried blood spots (DBS) constitute a reliable, cheap, and less invasive technique to allow microsampling by capillary blood collected on a specific device. This work aimed to develop and validate an innovative analytical method, combining quantitative DBS with UHPLC-MS/MS instrumentation to identify and quantify 25 PFAS. The extraction procedure was developed and optimized within the range 2-100 ng/mL. Specifically, fortified blood was applied on Capitainer®B devices providing 10 μL of blood volume through a microfluidic channel. After 3 h of drying, the extraction was performed by methanol under sonication, followed by centrifugation. Then, the extraction solvent was evaporated; the residue was reconstituted with the mobile phase solution. The validated method evidenced good sensitivity, with limits of detection ranging from 0.4 ng/mL (PFODA, PFOS) to 1.0 ng/mL (PFOA, 3,6-OPFHpA). The ± 20% acceptability criteria established for intra- and inter-day precision and accuracy were fulfilled for all analytes. High recovery-above 80%-was recorded, whereas significant matrix effect resulted in ion enhancement (> 50%) for 13 analytes. In conclusion, the proposed workflow proved to be reliable, fit for purpose, and easily adaptable in the laboratory routine.

Comparison of high-resolution mass spectrometry acquisition methods for the simultaneous quantification and identification of per- and polyfluoroalkyl substances (PFAS)

Simultaneous identification and quantification of per- and polyfluoroalkyl substances (PFAS) were evaluated for three quadrupole time-of-flight mass spectrometry (QTOF) acquisition methods. The acquisition methods investigated were MS-Only, all ion fragmentation (All-Ions), and automated tandem mass spectrometry (Auto-MS/MS). Target analytes were the 25 PFAS of US EPA Method 533 and the acquisition methods were evaluated by analyte response, limit of quantification (LOQ), accuracy, precision, and target-suspect screening identification limit (IL). PFAS LOQs were consistent across acquisition methods, with individual PFAS LOQs within an order of magnitude. The mean and range for MS-Only, All-Ions, and Auto-MS/MS are 1.3 (0.34-5.1), 2.1 (0.49-5.1), and 1.5 (0.20-5.1) pg on column. For fast data processing and tentative identification with lower confidence, MS-Only is recommended; however, this can lead to false-positives. Where high-confidence identification, structural characterisation, and quantification are desired, Auto-MS/MS is recommended; however, cycle time should be considered where many compounds are anticipated to be present. For comprehensive screening workflows and sample archiving, All-Ions is recommended, facilitating both quantification and retrospective analysis. This study validated HRMS acquisition approaches for quantification (based upon precursor data) and exploration of identification workflows for a range of PFAS compounds.

Validating blood microsampling for per- and polyfluoroalkyl substances quantification in whole blood

Microsampling allows the collection of blood samples using a method which is inexpensive, simple and minimally-invasive, without the need for specially-trained medical staff. Analysis of whole blood provides a more holistic understanding of per- and polyfluoroalkyl substances (PFAS) body burden. Capillary action microsamplers (Trajan hemaPEN®) allow the controlled collection of whole blood as dried blood spots (DBS) (four 2.74 µL ± 5 %). The quantification of 75 PFAS from DBS was evaluated by comparing five common extraction techniques. Spiked blood (5 ng/mL PFAS) was extracted by protein precipitation (centrifuged; filtered), acid-base liquid-liquid extraction, trypsin protease digestion, and weak anion exchange (WAX) solid-phase extraction with analysis by high-performance liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS). Filtered protein precipitation was the most effective extraction method, recovering 72 of the 75 PFAS within 70 to 130 % with method reporting limit (MRL) for PFOS of 0.17 ng/L and ranging between 0.05 ng/mL and 0.34 ng/mL for all other PFAS. The optimised method was applied to human blood samples to examine Inter- (n = 7) and intra-day (n = 5) PFAS blood levels in one individual. Sixteen PFAS were detected with an overall Σ16PFAS mean = 6.3 (range = 5.7-7.0) ng/mL and perfluorooctane sulfonate (branched and linear isomers, ΣPFOS) = 3.3 (2.8-3.7) ng/mL being the dominant PFAS present. To the authors knowledge, this minimally invasive self-sampling protocol is the most extensive method for PFAS in blood reported and could be a useful tool for large scale human biomonitoring studies.

Distribution of per- and poly-fluoroalkyl substances and their precursors in human blood

Many studies have examined per- and poly-fluoroalkyl substances (PFASs) in human blood. However, the distribution of PFASs in human blood remains not well known, especially for perfluorooctane sulfonate (PFOS) precursors. In this study, human blood samples (n = 162) were collected from general Chinese population, and then the isomer-specific partitioning of PFASs between human plasma and red blood cells (RBCs) were investigated. Perfluorooctanoate (PFOA) and PFOS were consistently the predominant PFASs in both human plasma and RBCs. In human blood, among C₄-C₇ perfluoroalkyl carboxylates (PFCAs), the calculated mean mass fraction in plasma (F_p) values increased from 0.76 to 0.82 with the increasing chain length. C₇-C₁₃ PFCAs exhibited a trend of gradually decreasing mean F_p with chain length. Among PFAS precursors, 6:2 fluorotelomer phosphate diester had the highest mean F_p value (0.87 ± 0.11). Calculated F_p values of N-methyl perfluorooctanesulfonamide (N-MeFOSA) and N-ethyl perfluorooctanesulfonamide (N-EtFOSA) were 0.66 ± 0.13 and 0.70 ± 0.12, respectively. Individual branched isomers consistently had greater F_p values than their corresponding linear isomers for PFOA, PFHxS, and perfluoroctane sulfonamide. To our knowledge, this study first reports the distribution of N-MeFOSA and N-EtFOSA in human blood, contributing to the better understanding of the occurrence and fate of PFASs in humans.

Field-Based Distribution and Bioaccumulation Factors for Cyclic and Aliphatic Per- and Polyfluoroalkyl Substances (PFASs) in an Urban Sedentary Waterbird Population

The field-based distribution and bioaccumulation factor (BAF) for per- and polyfluoroalkyl substances (PFASs) were determined in residential Black Swans (Cygnus atratus) from an urban lake (Melbourne, Australia). The concentrations of 46 aliphatic and cyclic PFASs were determined by HPLC-MS/MS in serum and excrement from swans, and water, sediment, aquatic macrophytes, soil, and grass samples in and around the lake. Elevated concentrations of ∑₄₆PFASs were detected in serum (120 ng mL^-1) and excrement (110 ng g^-1 dw) were strongly related indicating a potential noninvasive sampling methodology. Environmental concentrations of PFASs were consistent with a highly impacted ecosystem and notably high concentrations of perfluoro-4-ethylcyclohexanesulfonate (PFECHS, 67584-42-3; C₈HF₁₅SO₃) were detected in water (27 ng L^-1) and swan serum (16 ng mL^-1). In the absence of credible putative alternative sources of PFECHS input to the lake, we propose that the use of high-performance motorsport vehicles is a likely source of contamination to this ecosystem. The BAF of perfluorocarboxylic acids increased with each additional CF₂ moiety from PFOA (15.7 L kg^-1 ww) to PFDoDA (3615 L kg^-1 ww). The BAF of PFECHS was estimated as 593 L kg^-1 ww, which is lower compared with that of PFOS (1097 L kg^-1 ww).