Adverse Maternal, Fetal, and Postnatal Effects of Hexafluoropropylene Oxide Dimer Acid (GenX) from Oral Gestational Exposure in Sprague-Dawley Rats

Abstract

Background: Hexafluoropropylene oxide dimer acid [(HFPO-DA), GenX] is a member of the per- and polyfluoroalkyl substances (PFAS) chemical class, and elevated levels of HFPO-DA have been detected in surface water, air, and treated drinking water in the United States and Europe.

Objectives: We aimed to characterize the potential maternal and postnatal toxicities of oral HFPO-DA in rats during sexual differentiation. Given that some PFAS activate peroxisome proliferator-activated receptors (PPARs), we sought to assess whether HFPO-DA affects androgen-dependent development or interferes with estrogen, androgen, or glucocorticoid receptor activity.

Methods: Steroid receptor activity was assessed with a suite of in vitro transactivation assays, and Sprague-Dawley rats were used to assess maternal, fetal, and postnatal effects of HFPO-DA exposure. Dams were dosed daily via oral gavage during male reproductive development (gestation days 14-18). We evaluated fetal testes, maternal and fetal livers, maternal serum clinical chemistry, and reproductive development of F1 animals.

Results: HFPO-DA exposure resulted in negligible in vitro receptor activity and did not impact testosterone production or expression of genes key to male reproductive development in the fetal testis; however, in vivo exposure during gestation resulted in higher maternal liver weights ([Formula: see text]), lower maternal serum thyroid hormone and lipid profiles ([Formula: see text]), and up-regulated gene expression related to PPAR signaling pathways in maternal and fetal livers ([Formula: see text]). Further, the pilot postnatal study indicated lower female body weight and lower weights of male reproductive tissues in F1 animals.

Conclusions: HFPO-DA exposure produced multiple effects that were similar to prior toxicity evaluations on PFAS, such as perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA), but seen as the result of higher oral doses. The mean dam serum concentration from the lowest dose group was 4-fold greater than the maximum serum concentration detected in a worker in an HFPO-DA manufacturing facility. Research is needed to examine the mechanisms and downstream events linked to the adverse effects of PFAS as are mixture-based studies evaluating multiple PFAS. https://doi.org/10.1289/EHP4372.

Figure 1.

Schematic diagram of study designs for evaluating maternal, fetal, and postnatal effects of oral gestational hexafluoropropylene oxide dimer acid (HFPO-DA) exposure. Both (A) fetal and (B) postnatal study designs used oral gavage dosing from gestation day (GD) 14–18 at the indicated exposure levels. Fetal plasma HFPO-DA concentration (*) was only evaluated at doses of $0–30\mathrm{mg}/\mathrm{kg}\text{per}\text{day}$. AGD, anogenital distance; NR, nipple retention; PND, postnatal day; PPAR, peroxisome proliferator-activated receptor; PPS, preputial separation; VO, vaginal opening.

Figure 2.

Expression of significantly up-regulated genes (ANOVA, $p<0.0001$) from peroxisome proliferator-activated receptor (PPAR) signaling pathway gene arrays in (A) fetal ($n=6$ for control, $n=3$ for treated) and (B) maternal ($n=5$ for control, $n=3$ for treated) livers following gestation day (GD) 14–18 oral maternal exposure to hexafluoropropylene oxide dimer acid (HFPO-DA). Upper portions (above break) display significantly altered genes common to both fetal and maternal livers, lower portions display genes differentially altered between fetal and maternal livers. Cell values represent significant ($p<0.01$) dose-level fold-induction values relative to control livers [cells with no value were not significantly different from control (see Table S2 for gene descriptions, and Tables S3 and S6 for complete gene expression data)]. Legend indicates fold-induction compared with control with darker shaded genes more highly expressed. Genes with fold-induction $>25\text{-fold}$ of control were beyond the scale of the legend. Ctl, control.

Figure 3.

(A) Maternal body weight gain during gestation day (GD)14–18 dosing period and (B) maternal liver weight on GD18. Data points represent individual replicates (control, $n=9$; $1–30\mathrm{mg}/\mathrm{kg}$, $n=6$; $62.5–500\mathrm{mg}/\mathrm{kg}$, $n=3$), bars and whiskers represent $\text{mean}\pm$ standard error, and asterisks represent significant differences compared with control values (*, $p<0.05$; **, $p<0.01$; ***, $p<0.001$; ****, $p<0.0001$). Statistical significance was determined using analysis of variance; for liver weight analysis, body weight was included as a covariate.

Figure 4.

Concentrations of (A) total triiodothyronine (${\mathrm{T}}_3$), (B) total thyroxine (${\mathrm{T}}_4$), and lipids [(C) cholesterol, (D) triglycerides, (E) high-density lipoproteins (HDL), and (F) low-density lipoproteins (LDL)] in maternal serum following oral hexafluoropropylene oxide dimer acid (HFPO-DA) dosing from gestation days (GD) 14–18. Dam serum was collected on GD18 approximately 2 h after final oral dose. Data points represent individual replicates (control, $n=6$; treated, $n=3$), bars and whiskers represent $\text{mean}\pm$ standard error, and asterisks represent significant differences compared with control values using analysis of variance (*, $p<0.05$; **, $p<0.01$; ***, $p<0.001$; ****, $p<0.0001$). $<\text{DL}$, values below radioimmunoassay detection limit.

Figure 5.

Maternal serum and fetal plasma hexafluoropropylene oxide dimer acid (HFPO-DA) concentrations ($\text{mean}\pm$ standard error, $n=3–9$; see Table S10) as a function of oral dose following maternal exposure from gestation day (GD) 14–18. Samples were collected on GD18 approximately 2 h after final oral dose. (A) Full maternal dose range modeled using exponential one-phase association and (B) low dose range modeled using linear regression (95% confidence intervals shaded). Fetal plasma was collected only from the low dose range ($1–30\mathrm{mg}/\mathrm{kg}\text{per}\text{day}$).

Figure 6.

Dose–response curves (four-parameter logistic regression) and 5% effect estimates [${\text{EC}}_5$ with 95% confidence intervals (CIs)] for the most sensitive end points [(A) maternal liver weight, (B) maternal liver Ehhadh gene expression, (C) maternal serum total triiodothyronine $\left({\text{tT}}_3\right)$, and (D) total thyroxine $\left({\text{tT}}_4\right)$] as a function of maternal serum hexafluoropropylene oxide dimer acid (HFPO-DA) concentration. Dam serum HFPO-DA concentrations represent those measured on gestation day (GD)18 following GD14–18 dosing. Data points represent $\text{mean}\pm$ standard error, (A) control $n=9$, $1–30\mathrm{mg}/\mathrm{kg}\text{per}\text{day}$ $n=6$, $62.5–500\mathrm{mg}/\mathrm{kg}\text{per}\text{day}$ $n=3$; (B–D) control, $n=6$; treated, $n=3$.

Figure 7.

Comparison of mean maternal Sprague-Dawley rat serum hexafluoropropylene oxide dimer acid (HFPO-DA) concentration from (A) 1- and (B) 125-mg/kg per day exposure groups and individual human plasma HFPO-DA concentrations from workers in an HFPO-DA manufacturing facility in the Netherlands (DuPont 2017). Horizontal lines indicate various margins of internal exposure (MOIE) levels as compared with individual worker plasma concentrations.