1H NMR-Based Metabolomic Analysis of Sub-Lethal Perfluorooctane Sulfonate Exposure to the Earthworm, Eisenia fetida, in Soil

1H NMR-based metabolomics was used to measure the response of Eisenia fetida earthworms after exposure to sub-lethal concentrations of perfluorooctane sulfonate (PFOS) in soil. Earthworms were exposed to a range of PFOS concentrations (five, 10, 25, 50, 100 or 150 mg/kg) for two, seven and fourteen days. Earthworm tissues were extracted and analyzed by 1H NMR. Multivariate statistical analysis of the metabolic response of E. fetida to PFOS exposure identified time-dependent responses that were comprised of two separate modes of action: a non-polar narcosis type mechanism after two days of exposure and increased fatty acid oxidation after seven and fourteen days of exposure. Univariate statistical analysis revealed that 2-hexyl-5-ethyl-3-furansulfonate (HEFS), betaine, leucine, arginine, glutamate, maltose and ATP are potential indicators of PFOS exposure, as the concentrations of these metabolites fluctuated significantly. Overall, NMR-based metabolomic analysis suggests elevated fatty acid oxidation, disruption in energy metabolism and biological membrane structure and a possible interruption of ATP synthesis. These conclusions obtained from analysis of the metabolic profile in response to sub-lethal PFOS exposure indicates that NMR-based metabolomics is an excellent discovery tool when the mode of action (MOA) of contaminants is not clearly defined.

PFOS present in soil after two, seven and fourteen days of exposure were extracted in triplicate using the procedure based on Higgins et al. [2]. Homogenized and air-dried OECD soil (1 g) was transferred to a 40-mL polypropylene (PP) vial, to which 10 mL of an acetic acid solution (1%) was added. Each vial was then vortexed, sonicated for 15 min in a preheated bath (60 o C) and centrifuged at 4500 rpm (~1500 g) for 2 min using an International Equipment Company 21000 centrifuge (Fisher Scientific). The acetic acid solution was then decanted into another 40-mL PP vial. A 2.5 mL aliquot of a solvent mixture composed of 90:10 (v/v) methanol and 1% acetic acid in Milli-Q (Millipore Synergy UV, Billerica, MA) water was then added to the original PP vial, which was followed by vortexing, sonication for 15 min at 60 o C and centrifugation at 4500 rpm (~1500 g) for 2 min. The supernatant was again transferred to the second PP vial. The acetic acid wash followed by the methanol/acetic acid extraction was repeated and a final 10 mL wash with acetic acid was performed. All washes and extracts were combined for each sample. The total volume of the extracts and washes was approximately 35 mL.

Sample Cleanup
Solid phase extraction (SPE) was performed to concentrate the extracts and to remove the acetic acid, salts, and potential matrix interferences. A 500-mg SUPELCLEAN LC-18 cartridge (SUPELCO, PA, USA) was first conditioned with 10 mL of methanol followed by 10 mL of 1% acetic acid. The extracts were then loaded on to the SPE cartridges that were mounted on a vacuum manifold. The SPE cartridges were then rinsed with 10 mL of Milli-Q water prior to being allowed to dry under vacuum for 2 h prior to elution. PFOS was eluted from the SPE cartridge with 4 mL of methanol and was collected in a 1:1 (v/v) methanol/acetone washed 20 mL glass vials. The eluent was then concentrated to 2 mL under nitrogen and transferred to fresh 20 mL glass vials. The original 20 mL glass vials were then rinsed with 800 L of methanol. The rinse was combined with the eluent and an additional 1200 L of 0.01% aqueous ammonium hydroxide solution was added. The extracts were stored at 4 C until analysis. Prior to analysis the extracts were diluted (10, 000×) into 990 L of 1:1 methanol/water to which 10 L of 13 C 4 -PFOS aqueous internal standard (750 pg/mL) was added in a 2-mL autosampler glass vial.

HPLC-MS/MS Analysis
Analysis of PFOS extracted from the spiked OECD soil was performed using high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). An Agilent 1200 series HPLC was coupled with a 4000 QTRAP triple quadrupole mass spectrometer (Applied Biosystems-MDS Sciex, Concord, ON, Canada). Water and methanol (20 mM Ammonium acetate) were the solvents, which were delivered at a flow rate of 0.5 mL/min. The sample injection volume was 20 µl. The chromatographic separation was obtained using a Kinetex 2.6 µm Phenyl-Hexyl column (4.6 mm i.d. × 100 mm, 2.6 µM; Phenomenex, Torrance, CA). PFOS separation was obtained in 10 mins under gradient conditions, with 65 : 35 methanol:water initial mobile phase, followed by a 0.5 min ramp to 80:20 methanol:water, then a 2 min ramp to 95 : 5 methanol:water, which was held for 4 mins, followed by a 0.5 min ramp back to 65 : 35 methanol:water which was held for 3.50 mins.
The mass spectrometer was operated in negative electrospray ionization multiple reaction monitoring (MRM) mode using previously published methods [3,4]. MRM transition related parameters were optimized for PFOS -(m/z = 499) and SO 3 Figure S1. PCA loadings plots showing the metabolites that were major contributors to the separation observed in the average PCA scores plot of the E. fetida tissue extracts comparing the controls and PFOS exposed earthworms. The abscissa refers to the 1 H NMR chemical shifts (ppm). The loadings plots refer to, two-day exposure (A) PC1 and PC2, (B) PC3 and PC4, seven-day exposure (C) PC1 and PC2, (D) PC3 and PC4, and fourteen-day exposure (E) PC1 and PC2, (F) PC3 and PC4.

Figure S2
PCA loadings plots showing the metabolites that were major contributors to the separation observed in the average PCA scores plot of the E. fetida tissue extracts comparing the controls and earthworms exposed to PFOS for two, seven and fourteen days. The abscissa refers to the 1 H NMR chemical shifts (ppm). The loadings plots refer to, (A) PC1 and PC2, and (B) PC3 and PC4 Figure S3. Histograms of Q 2 Y values for cross-validated PLS models using the binned 1 H NMR spectra as the X-table and random permutations of the PFOS exposure concentrations as the Y variable. Distributions were constructed using 400 permutations of the Y table. The histograms correspond to PLS models constructed for (A) two days of exposure, (B) seven days of exposure and (C) fourteen days of exposure. Figure S4. T-test filtered 1 H NMR difference spectra of the E. fetida tissue extracts obtained by subtracting the mean buckets of the control earthworms from the mean buckets for the PFOS exposed earthworms after two days of exposure and retaining the buckets that were statistically different from the controls at α = 0.05. The t-test filtered 1 H NMR difference spectra are shown after PFOS exposure of (A) 5 mg/kg, (B) 10 mg/kg, (C) 25 mg/kg, (D) 50 mg/kg, (E) 100 mg/kg, and (F) 150 mg/kg. Figure S5. T-test filtered 1 H NMR difference spectra of the E. fetida tissue extracts obtained by subtracting the mean buckets of the control earthworms from the mean buckets for the PFOS exposed earthworms after seven days of exposure and retaining the buckets that were statistically different from the controls at α = 0.05. The t-test filtered 1 H NMR difference spectra are shown after PFOS exposure of (A) 5 mg/kg, (B) 10 mg/kg, (C) 25 mg/kg, (D) 50 mg/kg, (E) 100 mg/kg, and (F) 150 mg/kg. Figure S6. T-test filtered 1 H NMR difference spectra of the E. fetida tissue extracts obtained by subtracting the mean buckets of the control earthworms from the mean buckets for the PFOS exposed earthworms after fourteen days of exposure and retaining the buckets that were statistically different from the controls at α = 0.05. The t-test filtered 1 H NMR difference spectra are shown after PFOS exposure of (A) 5 mg/kg, (B) 10 mg/kg, (C) 25 mg/kg, (D) 50 mg/kg, (E) 100 mg/kg, and (F) 150 mg/kg.