*Article* **Multi- and Transgenerational Effects of Developmental Exposure to Environmental Levels of PFAS and PFAS Mixture in Zebrafish (***Danio rerio***)**

**Alex Haimbaugh 1 , Chia-Chen Wu 1 , Camille Akemann 1 , Danielle N. Meyer 1 , Mackenzie Connell 2 , Mohammad Abdi 2 , Aicha Khalaf 2 , Destiny Johnson <sup>2</sup> and Tracie R. Baker 1,2,3, \***


**Abstract:** Per- and polyfluoroalkyl substances (PFASs) are ubiquitous in the environment and are tied to myriad health effects. Despite the phasing out of the manufacturing of two types of PFASs (perfluorosulfonic acid (PFOS) and perfluorooctanoic acid (PFOA)), chemical composition renders them effectively indestructible by ambient environmental processes, where they thus remain in water. Exposure via water can affect both human and aquatic wildlife. PFASs easily cross the placenta, exposing the fetus at critical windows of development. Little is known about the effects of low-level exposure during this period; even less is known about the potential for multi- and transgenerational effects. We examined the effects of ultra-low, very low, and low-level PFAS exposure (7, 70, and 700 ng/L PFOA; 24, 240, 2400 ng/L PFOS; and stepwise mixtures) from 0–5 days post-fertilization (dpf) on larval zebrafish (*Danio rerio*) mortality, morphology, behavior and gene expression and fecundity in adult F0 and F1 fish. As expected, environmentally relevant PFAS levels did not affect survival. Morphological abnormalities were not observed until the F1 and F2 generations. Behavior was affected differentially by each chemical and generation. Gene expression was increasingly perturbed in each generation but consistently showed lipid pathway disruption across all generations. Dysregulation of behavior and gene expression is heritable, even in larvae with no direct or indirect exposure. This is the first report of the transgenerational effects of PFOA, PFOS, and their mixture in terms of zebrafish behavior and untargeted gene expression.

**Keywords:** PFAS; PFAS mixtures; epigenetics; zebrafish; transgenerational

#### **1. Introduction**

Per- and polyfluoroalkyl substances (PFASs) are a class of chemicals constituted by a polar head group attached to a chain of C-F bonds. The unique chemistry of these compounds renders them effectively indestructible and, thus, a prime candidate for highheat industrial processes and long-lasting consumer goods such as non-stick cookware and waterproofed outerwear. The utility of PFASs is offset by their bioaccumulation and toxic health effects. PFAS are detected virtually everywhere—in diverse wildlife, multiple environmental matrices, and in >99% of the general public [1–3]. Drinking water is a significant source of exposure in humans [4], and drinking water treatment plants are not designed to remove these contaminants from source water. Likewise, wastewater treatment plants do not intentionally filter out PFASs. PFASs are commonly found to be in the parts per trillion (ppt; ng/L) range in both untreated and treated drinking water [5] and wastewater [6]. The widespread low-level exposure warrants investigation into the health effects on wildlife and humans.

**Citation:** Haimbaugh, A.; Wu, C.-C.; Akemann, C.; Meyer, D.N.; Connell, M.; Abdi, M.; Khalaf, A.; Johnson, D.; Baker, T.R. Multi- and Transgenerational Effects of Developmental Exposure to Environmental Levels of PFAS and PFAS Mixture in Zebrafish (*Danio rerio*). *Toxics* **2022**, *10*, 334. https:// doi.org/10.3390/toxics10060334

Academic Editors: Lisa Truong and Robyn L. Tanguay

Received: 24 April 2022 Accepted: 17 June 2022 Published: 18 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

PFASs readily cross the placental barrier, potentially exposing a fetus during sensitive time periods during development [7]. Chemical assault during critical windows in development can have effects later in life; this thinking stems from the developmental origins of health and disease (DOHaD) hypothesis [8]. DOHaD posits that the timing of the exposure is crucial in determining the result. Placental transfer of PFASs necessitates the study of early-life exposure and the heritable effects of exposure. As most people have small amounts of many types of PFASs in their bodies, it is of great general interest to study the effects of low-level exposure, including to mixtures, on developing organisms.

The two most common PFASs carried by the general population are perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS), and they are usually detected at higher levels than other types of PFASs. PFOS is present at approximately three times the levels of PFOA in humans [9]. Choosing a relevant exposure concentration is of importance when planning translational experiments to realistically inform public health. The general population carries serum PFAS levels in the µg/L range (>999 ng/L), with an average of 1.42 µg/L PFOA and 4.25 µg/L PFOS [9]. The Environmental Protection Agency (EPA) health advisory limit for drinking water of PFOA, PFOS, or their combined concentration in mixture is 70 ppt, or 70 ng/L. Much early work characterizing PFAS toxicity used, understandably, high dose experiments to define outcomes, such as the concentration at which 50% of exposed organisms die (LC50). We now know that PFAS levels, while ubiquitous in all environmental compartments, are typically at ng/L or µg/L levels in water. A study of treated water from 25 drinking water treatment plants across the United States found a median concentration of 19.5 ng/L for 17 PFASs combined, with a maximum sum of 1.1 µg/L (1100 ng/L) [5]. Our group has previously shown that mean concentrations of PFOA and PFOS in a waterway that provides drinking water in a major metropolitan area in Michigan were 2.2 ng/L and 2.9 ng/L, respectively [10]. In order to advance public health knowledge of exposures at both environmentally relevant levels and levels encompassing the EPA health advisory, we chose exposure concentrations for PFOA of 7, 70, and 700 ng/L and for PFOS at approximately 3× higher concentrations of 24, 240, and 2.4 µg/L (2400 ng/L), a ratio similar to reported human levels. The mixture concentrations contain half of each exposure level per chemical (e.g., the ultra-low mixture concentration contains 3.5 ng/L PFOA and 12 ng/L PFOS). Throughout this report, we will refer to these nominal concentrations of 7 ng/L PFOA exposure and 24 ng/L PFOS exposure as the "ultra-low" exposure level, the 70 ng/L PFOA and 240 ng/LPFOS exposure as "very low", and the 700 ng/L PFOA and 2.4 µg/L PFOS exposure as "low". Exposures in other studies within the ng/L range are referred to as "low", µg/L range as "moderate", mg/L range as "high", and g/L range as "very high".

As individual PFASs are seldom discovered in the environment or treated drinking water alone, it is critical to study mixtures at environmental levels. PFAS mixtures are increasingly studied, but their effects are still unclear and often unpredictable, especially at different concentrations. Ding et al. [11] characterized the 1:1 mixture of PFOA and PFOS at high concentrations to be synergistic towards early-life lethality in zebrafish, while increasing the PFOA:PFOS ratio resulted in antagonism, then additivity. In another study, individual PFASs alone significantly changed swim behavior in exposed fish at moderate levels, but a mixture of nine PFAS had no effect at environmental levels [12]. We sought to address this gap by characterizing a low-level mixture of the two most commonly detected PFAS.

There is emerging evidence that PFAS exposure confers heritable effects on later generations via epigenetic mechanisms [13] rather than direct genotoxicity. Epigenetic modifications to DNA or chromatin serve as a "biological memory" of environmental history that modulate gene regulatory networks in current and future generations [14]. Toxicoepigenetic initialization in the directly exposed organism can be perpetuated across multiple generations. When an individual is directly exposed, the exposure indirectly affects germ cells residing in the individual. "Multigenerational" (F1) effects are seen in the generation following the directly exposed (F0) generation. Even if the exposure ceases,

indirect germ cell exposure has occurred and can present phenotypically in this next generation's life. Effects are considered "transgenerational" when observed in the subsequent (F2) generation, which has never been directly or indirectly exposed. Zebrafish exposures with explicit epigenetic outcomes through multiple generations have not been conducted for PFAS. However, in the F0 generation, Bouwmeester et al. [15] found that moderaterange PFOA exposure increased methylation associated with *vtg1*, a gene involved in fertility. Limited epigenetic studies have been done in rodents. Tian et al. (2019) found that non-specific methylation therapy administered with PFOS to F0 females resulted in better birth outcomes in F1 pups than F0 PFOS exposure without methylation therapy [16]. The potential heritability of PFAS exposure effects is pertinent, as measures taken now to prevent or reduce exposure could magnify public health benefits to the next generation(s) at scale. The results of the current study suggest that epigenetic mechanisms mediate each generation's response to exposure in terms of behavior and gene expression.

The zebrafish is an ideal model system for conducting early-life research on waterborne contaminants over multiple lifetimes. Zebrafish have been a useful, popular model in developmental toxicology due to their easy visibility, high *n*-values, quick generation time, and high homology with the human genome [17]. Additionally, the EPA plans to eliminate funding for mammalian vertebrate research completely by 2035 [18], positioning the zebrafish as a pertinent alternative model organism. From the outset, zebrafish eggs have a transparent chorion through which development can be observed from the single-cell stage to free-swimming larvae at five days post-fertilization (5 dpf). Zebrafish are prolific breeders, producing > 300 eggs per week, and are sexually mature in ~3 months [19], meaning transgenerational effects can be observed in about one year. They have been utilized as an ideal transgenerational model due to all of the above-mentioned benefits and the external fertilization of eggs, which reduces the number of generations compared to mammalian models [20].

This study aims to advance understanding of the short- and long-term health effects of developmental exposure to environmentally relevant levels of PFASs using the zebrafish model organism. After exposing embryonic zebrafish to environmental levels of two prevalent PFASs, PFOS and PFOA, and a mixture of the two chemicals (referred to throughout simply as "mixture") for the first 5 days of life, we found that swimming behavior and gene expression at 5 days post-fertilization (dpf) was affected by at least one concentration of all chemicals in all three generations (F0–F2). Pathway analysis of gene expression revealed upregulated pathways of immunotoxicity, movement disorders, and endocrine disruption. Adult fecundity (eggs produced per female) was statistically increased in the PFOA-exposed F0 generation and decreased in the F1 generation. Morphological abnormalities at 5 dpf were not observed until the F1 and F2 generations. As expected at these low doses, survival was uniformly unaffected by exposure.

It is the authors' aim that these results inform decision-making regarding safe contaminant limits in drinking water and in aquatic habitats. The federal health advisory limit set by the EPA for PFOS, PFOA, and their mixture is currently 70 ng/L [21], while some states legislate much lower levels. This study provides the first report of multigenerational effects of PFOA exposure on behavior and of mixture exposure on behavior and gene expression, supporting findings in other PFOS studies showing these endpoints are affected multigenerationally. Further, we show novel transgenerational effects on behavior and gene expression following low-level exposure to any PFAS during early life. Future efforts should include complex mixtures, and PFAS replacements, including "short-chain" alternatives, will be critical to study as well.

#### **2. Materials and Methods**

#### *2.1. Animal Husbandry of Adult Fish*

Adult AB strain zebrafish were maintained on a 14:10 h light:dark cycle, as previously described [22], on a recirculating system of RO water buffered to a neutral pH with Instant Ocean© salts (Spectrum Brands, Blacksburg, VA, USA) at 27–30 ◦C. Ammonia and nitrite levels remained at 0 ppm. Fish were fed twice daily (Aquatox Fish Diet, Zeigler Bros Inc., Gardners, PA, USA) and supplemented with brine shrimp (Artemia International, Fairview, TX, USA). All zebrafish use protocols were approved by the Institutional Animal Care and Use Committee at Wayne State University, according to the National Institutes Health Guide to the Care and Use of Laboratory Animals (Protocol 16-03-054; approved 4 August 2016).

#### *2.2. PFAS Exposures*

#### 2.2.1. Spawning Procedure

To obtain F0 embryos, adult stock zebrafish were spawned in a 2:1 female:male ratio (at least 4 trios per concentration) (Figure S1) in the environmental conditions described above. Sexes were separated overnight by a plastic divider in a spawning tank and were allowed to spawn at 08:00 the next morning. Spawning tanks contained a slotted insert through which eggs fell to the bottom, away from the adults. Embryos were harvested after 2 h of spawning activity.

#### 2.2.2. Egg Cleaning

Eggs were incubated at 27 ◦C in 58 ppm bleach for 10 min, rinsed with RO water, and then placed back in their normal environment of a weak salt solution (600 mg/L salt in RO water) containing Instant Ocean© salts (Spectrum Brands, Blacksburg, VA, USA).

#### 2.2.3. Exposure Protocol

Perfluorooctanoic acid (PFOA) (CAS# 335-67-1, Sigma, St. Louis, MO, USA, 95% purity) and perfluorooctane sulfonic acid (PFOS) (CAS# 1763-23-1, Sigma, 99.4% purity) were used for stock solutions. From these stock solutions, serial dilutions in RO water buffered with Instant Ocean© salts were carried out each day of the exposure to reach the nominal concentrations of 7, 70, and 700 ng/L for PFOA; 24, 240, and 2400 ng/L PFOS; and a mixture with half of the individual concentrations and 1:1 volume ratios (e.g., the ultra-low mixture concentration would contain 3.5 ng/L PFOA and 12 ng/L PFOS). The control was exposed to RO water buffered with Instant Ocean salts; 30 embryos (≤4 hpf) were placed into a well of a 6-well Falcon plate with 8.5 mL of their respective chemical concentration or buffered water (controls). Solutions were replenished daily with approximately 90% fresh solution. Larvae were maintained in an incubator at 27 ◦C. On day 5, all larvae were rinsed three times in buffered water solution to end the exposure before proceeding with further assays.

#### *2.3. Survival and Abnormality Screening*

Survival was recorded on day 5 post-fertilization. Embryos or larvae were considered dead if the heart was stopped. On day 5, all hatched survivors were screened via light microscope for cardiac edema, yolk sac edema, presence of swim bladder, and bent spine. Student's *t*-test was used to determine the statistical significance of each concentration compared to control in terms of the percent total abnormalities. Assays were repeated a minimum of 5 times, with at least 150 larvae per concentration (Table S1). Each repetition was performed on a different day with different larvae.

#### *2.4. Behavioral Analysis*

The behavioral assay measuring swim distance in light and dark cycles was performed and analyzed as previously reported [20]. Briefly, healthy (no morphological abnormalities) 5 dpf larvae from control and exposed groups were acclimated to a well plate for ≥1 h, then loaded into a DanioVision Chamber (Noldus Information Technology, Wageningen, The Netherlands), which alternated four light and dark cycles for three min each following a chamber acclimation period. Raw data were exported to Noldus EthoVisionXT14, and average distance moved (cm) was analyzed using ANOVA and Tukey's HSD tests in custom R scripts (File S1). The assay was replicated at least three times for each chemical or

mixture, with at least 68 fish per concentration in each replicate (Table S1). Each repetition was performed on a different day with different larvae. Larvae were euthanized after the behavioral assay and not used for any further endpoints.

#### *2.5. RNA-Seq and Pathway Analysis*

At 5 dpf, five larvae were euthanized and pooled to create one sample, and at least 3 samples per concentration were analyzed for gene expression (Table S1). Each repetition was performed on a different day with a different cohort of larvae. Larvae were pooled to represent the ratio of healthy:abnormal larvae observed during the morphological abnormality assay. For example, if 20% of all low-level PFOA larvae presented abnormalities, 1 of each 5 pooled larvae would present an abnormality, while the other 4 were healthy. Larvae, once euthanized in 16.7 mg/mL tricaine methanesulfonate, were placed in RNALater. This was drained according to the manufacturer's instructions (i.e., between 1–7 days later) and then stored at −80 ◦C. Storage of larvae, RNA isolation, cDNA library preparation, sequencing, differential expression analysis, and pathway analysis were performed as previously reported [22]. Briefly, RNA isolation was performed with the Qiagen RNeasy Lipid Mini Kit (Qiagen, Hilden, Germany). cDNA libraries were prepared using the Quantseq™ 3 ′ mRNA-seq kit (Lexogen, Vienna, Austria). RNA and cDNA concentrations were measured with a Qubit™ 2.0 fluorometer (Invitrogen, Carlsbad, CA, USA), and cDNA quality was also assessed with an Agilent TapeStation 2200 (Agilent Technologies, Santa Clara, CA, USA). F0 samples were sequenced on Illumina® MiSeq™ (Illumina, CA, USA) and F1–F2 were sequenced on Illumina® HiSeq 2500™ (Illumina, CA, USA) using the Lexogen Bluebee® Genomics Platform (Bluebee, Rijswijk, The Netherlands). F0 reads were aligned to *Danio rerio* genome Build GRCz10, and F1–F2 reads were aligned to *Danio rerio* genome Build GRCz11; differential expression analysis was determined via DESeq2. Differentially expressed genes (DEGs) with log2 fold changes ≥0.75 or ≤−0.75, *p*-values <0.01, and ≥50 analysis-ready molecules were analyzed with Ingenuity Pathway Analysis (IPA®) software (Qiagen Bioinformatics, Redwood City, CA, USA).

#### *2.6. Fecundity Assay*

At sexual maturity and dimorphism (4–6 weeks), fish were spawned in a 1:1 male:female ratio in order to attribute the number of eggs produced to each individual female (Figure S1). Fish were not spawned more than once per week. Four randomly-chosen pairs per concentration and control were used per experiment (16 total spawning tanks) (Figure S1, Table S1). After two acclimation sessions of spawning, experiments were replicated a minimum of three times, and a minimum of 6 clutches per concentration were analyzed. Males and females were separated overnight by a plastic divider. At 08:00, dividers were removed and spawning allowed for 2 h. Then, each clutch was cleaned (as described in Section 2.2.2), and eggs were imaged for later quantification. Student's two-tailed *t*-test was used to determine the average number of eggs per female for each concentration and chemical.

#### *2.7. Sex Ratio*

At maturity, fish were visually assessed for female or male secondary sex characteristics. Chi-squared tests were used to determine the statistical significance of any concentration compared to control (Table S1). In F0 fish, dissection was performed for validation. Fish were euthanized in 1.67 mg/mL tricaine methanesulfonate (Syndel, Ferndale, WA, USA) for 10 min.

#### **3. Results**

#### *3.1. F0 Generation*

Table 1 shows significant endpoints in all chemicals and concentrations.

**Table 1.** Endpoints of PFAS exposure in zebrafish (*Danio rerio*) across all chemicals, concentrations, and generations. Survival, morphological abnormalities, swim distance, fecundity, sex ratio: percent change. DEGs: number. Blue: decreased endpoint. Orange: increased endpoint. Grey: both increased and decreased endpoints.



#### **Table 1.** *Cont.*

#### 3.1.1. F0 Survival and Abnormalities

No statistically significant larval abnormalities or mortality were observed in any concentration of any chemical or mixture. Ultra-low PFOS exposure approached significance with a slightly higher rate of abnormalities (*p* = 6.3 × 10−<sup>2</sup> ) (Table S1).

#### 3.1.2. F0 Behavior

#### PFOA

Direct PFOA exposure significantly decreased larval swimming distance in both dark and light cycles at every concentration, with the exception of the low concentration in the light (Figures 1 and 2) (*p* < 1 × 10−<sup>8</sup> ; *p* <1 × 10−<sup>8</sup> ; *p =* 2.1 × 10−<sup>2</sup> ; ultra-low, very low, low exposure in the dark, respectively) (*p =* 3.4 × 10−<sup>4</sup> ; *p* < 1 × 10−<sup>8</sup> ; ultra-low and very low exposure in the light, respectively).

#### PFOS

Direct PFOS exposure had no effect on larvae from any concentration in the dark. In the light, very low and low exposure groups were significantly hypoactive (Figures 1 and 2) (*p =* 4 × 10−<sup>7</sup> , 1.2 × 10−<sup>3</sup> , respectively).

#### Mixture

Direct exposure to the mixture of PFOA and PFOS resulted in increased swimming distance in larval zebrafish (*Danio rerio*), regardless of light/dark setting (Figures 1 and 2) (dark: *p =* 1.6 × 10−<sup>3</sup> ; *p =* 8.1 × 10−<sup>4</sup> ; *p* < 1 × 10−<sup>8</sup> ; ultra-low, very low, low exposure, respectively) (light: *p =* 6.7 × 10−<sup>4</sup> ; *p =* 6.4 × 10−<sup>5</sup> ; *p* < 1 × 10−<sup>8</sup> ; ultra-low, very low, low exposure, respectively).

#### 3.1.3. F0 Transcriptomics

The full lists of DEGs for all chemicals and concentrations in the F0 generation can be found in Table S3; the top five up- and downregulated DEGs are shown in Table 2. Venn diagrams illustrating the overlap of generation-specific DEGs (all concentrations combined) are in Figure 2. Venn diagrams illustrating the overlap of F0 DEGs for each chemical (all concentrations combined) are in Figure 3. DEGs are considered significant at *p* < 0.01 and log2FC of ≥0.75 or ≤−0.75. Pathway analysis could not be performed due to an insufficient number of DEGs.

**Figure 1.** Locomotion following PFAS exposure in dark and light. Yellow: PFOA. Blue: PFOS. Green: mixture. (**a**) Top panel: F0 generation. Middle panel: F1 generation. Lower panel: F2 generation. \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001; ANOVA with Tukey pairwise test. 0: no exposure. UL: ultra-low exposure. VL: very low exposure. L: low exposure. (**b**) Simplified representation of significant behavioral direction. Upwards arrow: hyperactivity. Downwards arrow: hypoactivity. Two arrows: discordance between one or more concentrations on hyper- vs. hypoactivity. n.s.: not significant.

**Figure 2.** Number of DEGs in each generation for each chemical.

**Table 2.** Top 5 up- and downregulated DEGs in each chemical, concentration, and generation of zebrafish (*Danio rerio*) and the pathways affected.



#### **Table 2.** *Cont.*

**Figure 3.** Number of DEGs from each chemical in every generation. Size of the circle indicates the proportion of genes expressed; color of the circle indicates the chemical.

#### PFOA

− − Exposure to the ultra-low level of PFOA had no effect on differential gene expression. At very low exposure, only *rpe65a* was significantly changed (LFC −0.89, *p =* 9.7 × 10−10). Basic cellular functions were impacted by low exposure. Of the 14 genes that were differentially expressed (DEGs) (11 up, 3 down), *tmem14c* was the most upregulated (LFC 0.96) and *atp6v0e1* the most downregulated (LFC −0.85).

#### PFOS

Exposure to the ultra-low level of PFOS significantly increased the expression of the insulin receptor substrate *irs2a* (LFC 0.77). Of 54 DEGs, the inflammatory response gene *irg1l* was the most upregulated following very low exposure (LFC 1.72); the most downregulated was the kinesin *kif3c* (LFC −1.24). Low exposure to PFOS did not elicit gene expression changes.

#### Mixture

Exposure to ultra-low levels of PFAS mixture significantly downregulated six genes. The most downregulated was the isomerase *fkbp9* (LFC −0.92). Very low exposure had no effect on gene expression. Low exposure induced the upregulation of two genes, the amino acid transporter *slc6a19a.1* (LFC 0.82) and the nucleoside biosynthesis gene *entpd8* (LFC 0.78).

#### 3.1.4. F0 Fecundity

PFOA

Early-life PFOA exposure did not significantly affect adult female egg production at any concentration (Figure S2). Full fecundity data are shown in Table S2.

#### PFOS

Fecundity trended downwards with increasing concentrations of PFOS but did not reach statistical significance (low exposure: *p =* 7 × 10−<sup>2</sup> ) (Figure S2). Full fecundity data are shown in Table S2.

#### Mixture

Early-life mixture exposure did not significantly affect fecundity at any concentration (Figure S2). Full fecundity data are shown in Table S2.

#### 3.1.5. F0 Adult Body Weight/Length

Adult body weight or length was unaffected by any concentration of any chemical significantly. Females exposed to PFOA at very low and low levels trended towards being significantly heavier (*p =* 5.1 × 10−<sup>2</sup> , 6.5 × 10−<sup>2</sup> , respectively); 15–17 fish were evaluated per concentration, with one replicate per exposure group.

#### 3.1.6. F0 Sex Ratio

No chemical or concentration affected the sex ratio in the F0. Low mixture exposure approached a significant decrease in the male:female ratio (0.73, *p* = 5.3 × 10−<sup>2</sup> , *n* = 22–23). The control male:female ratio for PFOA was 1.11 (*n* = 19–25), for PFOS was 2.09 (*n* = 29–34), and for the mixture was 2.67; 19–34 fish were evaluated per concentration, with one replicate per exposure group.

#### *3.2. F1 Generation*

Table 1 shows significant endpoints for all chemicals and concentrations.

#### 3.2.1. F1 Abnormalities and Survival

No statistically significant abnormalities or mortality were observed in any concentration of PFOS or the mixture. At low PFOA exposure, a significant decrease was observed in abnormalities (*p* = 2.9 × 10−<sup>2</sup> ) (Table S1), and at ultra-low exposure, a significant increase was observed for survival (*p* = 3.6 × 10−<sup>4</sup> ).
