Perfluorooctane sulfonic acid (PFOS) is an anthropogenic chemical found in aqueous film-forming foams (AFFFs) and many consumer products. Despite its environmental ubiquity and persistence, little is known about the effects of PFOS on stress levels in wild animals. Here, we examined PFOS bioaccumulation and correlations between PFOS exposure and oxidative stress in snapping turtles (Chelydra serpentina) downstream of Griffiss Air Force Base in Rome, New York, a known source of AFFF contamination. Maximum concentrations of PFOS were extrapolated as 41.7 ppb and 498 ppb in water and turtle samples, respectively; in comparison, the proposed United States national advisory concentration of PFOS in drinking water is currently 0.00002 ppb. PFOS concentrations declined with distance downstream from the base and were lower in other local water bodies. Indices of oxidative stress were positively correlated with plasma PFOS concentrations in snapping turtles. These data illuminate the potential for bioaccumulation and increasing oxidative stress levels associated with PFOS contamination in a wild population of aquatic turtles.

On 6 October 2020, water samples were collected for PFAS analysis from six locations: one 5 km upstream of the contamination point in Sixmile Creek, four 0–1500 m downstream of the contamination point in Sixmile Creek, and one in the Erie Canal, 2,250 m downstream of the contamination point. Stream water samples (30 mL, collected in triple-rinsed Falcon centrifuge tubes) were stored at −20o °C until analysis.

Twenty-nine blood samples were collected from snapping turtles between June and September of 2021. Turtle sampling locations included three sites downstream of the contamination point in Sixmile Creek, one site at Stony Creek Pond adjacent to a Department of Defense Fuel Depot fuel storage facility, which was also known to have PFAS contamination, and three control sites (Delta Lake, South Otselic Fish Hatchery, and Camroden Pond) with no known PFAS contamination. Snapping turtles were captured in hoop nets, and Tomahawk traps baited with tuna. Traps were checked within 24 hours after baiting. Captured turtles were weighed, and sex was determined by the shape and position of the cloaca. Blood samples (~ 0.1 mL) were taken from the dorsal cervical vessels into heparinized tubes and stored on ice until centrifugation. Plasma was stored at −80 °C until analysis. Turtles were sampled within 10 min after they were removed from the traps and were released at the point of capture immediately after sampling.

PFAS quantification
Methods for the extraction of PFAS from water and plasma samples were modified from Taniyasu et al. (2005) and the EPA’s Method 537.1 (Shoemaker and Tettenhorst 2020). All samples were prepared and analyzed in duplicate. To prepare water samples for analysis, 10 mL samples were first spiked with the internal standard for PFOS [sodium perfluoro-1-[^₁₃^C₄]-octane sulfonate (MPFOS); purity > 98%; Sigma-Aldrich] at a concentration of 16 ng/mL and then passed through conditioned Oasis WAX (6 cc, 150 mg, 30 µm) solid-phase extraction cartridges (Wellington Laboratories) at a rate of 1 drop/s. WAX cartridges were conditioned by washing columns with 2 mL of 100% HPLC-grade acetonitrile (Sigma-Aldrich) followed by 2 mL of water. After samples were added, the cartridges were washed with 3 mL of 20 mM ammonium acetate followed by 3 mL of 100% methanol. Target analytes were eluted with 3 mL of 2% ammonium hydroxide in methanol and collected into conical polypropylene test tubes. The eluate was evaporated using a TurboVap concentrator with compressed air. One hundred µL of 60:40 water: acetonitrile solution was added to the test tube residues for reconstitution before HPLC/MS/MS analysis. To prepare plasma samples for analysis, turtle plasma (25µL) was transferred to a centrifuge tube with 100% acetonitrile (75µL) and 0.22 µg/mL MPFOS internal standard solution (10µL) for a final concentration of 20 ng/mL. After centrifugation, the supernatant (approximately 50 µL) was extracted and filtered through a conditioned 0.45-micron sample filter-tipped syringe (using the conditioning methods described above) with distilled water (50µL) into a conical vial within a glass chromatography vial.

LC/MS/MS analysis and quantification
Analysis of PFOS was performed using an ultra-high-performance liquid chromatograph and mass spectrometer (LC/MS/MS), Thermo Vanquish HPLC LTQ-XL ion trap mass spectrometer, operated in the electrospray negative ionization mode (electrospray needle: 4.5 kV; capillary: 275 °C). Five µL of sample extracts were injected into an Accucore Vanquish C18 30 × 2.1 mm column (ThermoFisher; upper-pressure limit 1000.0 bar; 1.50 µm particle size; 2.1 mm diameter). At a flow of 0.250 mL/min and pressure of approximately 4600 psi (divert valve set to source), the flow gradient was started at 60% 95:5 water: acetonitrile and 40% 5:95 water: acetonitrile, increased to 70% 5:95 water: acetonitrile at 3 min, then returned to the starting condition at 5.3 min. Due to the limitations of the ion trap in detecting the expected m/z 80 product ion, PFOS, and MPFOS were detected using the ion current obtained from the deprotonated molecule, m/z 499 and m/z 503, respectively. Calibration curves were created separately for water and plasma analysis. Analytical standards for PFOS (purity > 85%) were purchased from Sigma-Aldrich. For water samples, a curve was created from six standard samples prepared at 0.8, 4, 8, 40, 80, and 160 ng/mL of the PFOS standard, diluted in a 60:40 solution of water: acetonitrile with MPFOS internal standard (R2 = 0.988). For plasma samples, a calibration curve was created from six standard samples prepared at 0.4, 1.6, 8, 40, and 80 ng/mL of the PFOS standard, diluted in a 60:40 solution of water: acetonitrile with MPFOS internal standard (R2 = 0.974). Internal standards were maintained at concentrations of 16 ng/mL and 20 ng/mL MPFOS for water and plasma, respectively. Concentrations of PFOS in each sample were quantified based on the ratio of the response of the analytical standard concentrations to internal standard concentrations. Quality control samples were not run with these analyses.

Oxidative stress measurements
Oxidative stress levels in turtle plasma from Sixmile Creek were measured using the Arbor Assays DetectX® DNA Damage ELISA Kit (K059), following the manufacturer’s guidelines for plasma. The DNA Damage assay measures DNA and RNA deoxidized guanosine species, specifically 8-hydroxy-2′-deoxyguanosine (8-OHdG), which is a marker of oxidative stress. Plasma samples were diluted in a 1:8 ratio and run in duplicate. The endpoint absorbance values at 450 nm were read on a Bio‐Rad Benchmark Plus Microplate Spectrophotometer. Raw data were converted to DNA damage levels using the MyAssays (www.myassays.com) calculator tools designed specifically for these assays.