Metabolic profiling in human SH-SY5Y neuronal cells exposed to perfluorooctanoic acid (PFOA)

Abstract

Perfluorooctanoic acid (PFOA) is an abundant per- and polyfluoroalkyl substance (PFAS) detected in both indoor and outdoor environments. While studies suggest exposure concerns for humans, studies investigating PFOA-induced neurotoxicity are lacking. To address this gap, we exposed differentiated human SH-SY5Y cells to PFOA (0.1 μM up to 500 μM) at different time points (4, 24, 48, and 72 h) and measured cell viability, Casp3/7 activity, ATP levels, ATP synthase enzyme activity, mitochondrial membrane potential, reactive oxygen species (ROS), oxygen consumption rates for mitochondrial stress test (XFe24 Flux analyzer), glucose utilization, and global metabolome profiles to assess the potential for PFOA-induced neurotoxicity. Treatment with 10 or 100 μM PFOA did not compromise cell viability nor induce cytotoxicity to SH-SY5Y cells over a 48-hour exposure period. However, >250 μM PFOA compromised cell viability, induced cytotoxicity, and induced caspase 3/7 activity at 48 h. ATP levels were reduced in cells treated with 400 μM PFOA for 24 and 48 h, and with 100 μM PFOA and higher at 72 h. ATP synthase activity was inhibited by 250 μM PFOA but was unchanged by PFOA treatment at 200 μM or less. Conversely, mitochondrial membrane potential was reduced by >10 μM PFOA after 24 h. Total ROS was increased with 100 μM PFOA and higher after 4 h of exposure. Several mitochondria-related endpoints (basal respiration, ATP production, maximum respiration) were negatively affected at 250 μM PFOA at both 24- and 48-hour exposure, but were unaltered at concentrations of 100 μM PFOA or less. One exception was mitochondrial spare capacity, which was reduced by 100 μM PFOA after 24-hour exposure. Similarly, glycolysis, glycolytic capacity, and glycolytic reserve of SH-SY5Y cells were not altered by 10 nor 100 μM PFOA. Nontargeted metabolomics was conducted in cells treated with either 10 or 100 μM PFOA for 48 h, as these two concentrations were not cytotoxic and 28 metabolites differed among treatments. Notable was that 10 μM PFOA had little effect on the SH-SY5Y metabolome, and the metabolic profile was not statistically different from media nor solvent controls. On the other hand, 100 μM PFOA shifted the metabolic signature of the neuronal cells, leading to reduced abundance of ATP-related metabolites (adenine, nicotinamide), neurotransmitter precursors (DL-tryptophan, l-tyrosine), and metabolites that protect mitochondria during oxidative stress (betaine, orotic acid, and l-acetyl carnitine). We hypothesize that this metabolic signature may be associated with the reduced mitochondrial membrane potential observed at lower PFOA concentrations. Metabolic shifts appear to precede compromised cell viability, cytotoxicity, and apoptosis. This study generates mechanistic knowledge regarding PFOA-induced neurotoxicity, focusing on mitochondrial oxidative respiration and the neuronal metabolome.