Adverse PFAS effects on mouse oocyte in vitro maturation are associated with carbon‐chain length and inclusion of a sulfonate group

Abstract Objectives Per‐ and polyfluoroalkyl substances (PFAS) are man‐made chemicals that are widely used in various products. PFAS are characterized by their fluorinated carbon chains that make them hard to degrade and bioaccumulate in human and animals. Toxicological studies have shown PFAS toxic effects: cytotoxicity, immunotoxicity, neurotoxicity, and reproductive toxicity. However, it is still unclear how the structures of PFAS, such as carbon‐chain length and functional groups, determine their reproductive toxicity. Methods and Results By using a mouse‐oocyte‐in‐vitro‐maturation (IVM) system, we found the toxicity of two major categories of PFAS, perfluoroalkyl carboxylic acid (PFCA) and perfluoroalkyl sulfonic acid (PFSA), is elevated with increasing carbon‐chain length and the inclusion of the sulfonate group. Specifically, at 600 μM, perfluorohexanesulfonic acid (PFHxS) and perfluorooctanesulfonic acid (PFOS) reduced the rates of both germinal‐vesicle breakdown (GVBD) and polar‐body extrusion (PBE) as well as enlarged polar bodies. However, the shorter PFSA, perfluorobutanesulfonic acid (PFBS), and all PFCA did not show similar adverse cytotoxicity. Further, we found that 600 μM PFHxS and PFOS exposure induced excess reactive oxygen species (ROS) and decreased mitochondrial membrane potential (MMP). Cytoskeleton analysis revealed that PFHxS and PFOS exposure induced chromosome misalignment, abnormal F‐actin organization, elongated spindle formation, and symmetric division in the treated oocytes. These meiotic defects compromised oocyte developmental competence after parthenogenetic activation. Conclusions Our study provides new information on the structure‐toxicity relationship of PFAS.

persistent organic pollutants. 4 Human are regularly exposed to PFAS through inhalation, dermal exposure, food, and drinking water. 5 Two subcategories of PFAS (Table 1), perfluoroalkyl carboxylic acids (PFCA) and perfluoroalkyl sulfonic acids (PFSA) have drawn great attention in recent years due to their demonstrated neurotoxicity, 6,7 developmental toxicity, 8 immunotoxicity, 9 hepatotoxicity, 10 and especially reproductive toxicity. 11,12 Unfortunately, both PFCA and PFSA have been detected in humans. For instance, one of the PFSA, perfluorooctanesulfonic acid (PFOS), has a median serum concentration as high as 12.70 ng/ml. 13 In terms of female reproductive toxicity, PFCA and PFSA have been shown to be able to pass through the blood-follicle barrier and can be detected in follicular fluid. 14 Clinical evidence shows that PFCA and PFSA are associated with a late age at menarche, irregular menstrual cyclicity, and early menopause. 14 In vitro studies have demonstrated the direct cytotoxicity of PFCA and PFSA on mouse oocyte maturation. Oocyte maturation releases oocytes from dictyate arrest and prepares them for fertilization. Dramatic morphological changes occur during this process, including germinal vesicle breakdown (GVBD) and polar-body extrusion (PBE).
Various epigenetic regulations are known to be involved in oocyte maturation, including histone acetylation, phosphorylation, and SUMOylation. 15 The breakdown of nuclear envelop, GVBD, exposes the chromosomes to many environmental toxicants, such as iodoacetic acid, 16 PM 10 17 and several PFAS chemicals. [18][19][20][21] Iodoacetic acid, for example, has been shown to induce DNA damage and cause chromosome misalignment at the metaphase I stage. 16 PFOS exposure has also been shown to alter histone methylation levels with increased H3K4me3 and decreased H3K9me3 19 and, furthermore, modulate maternal-to-zygotic transition. 22,23 Mitochondria also play important roles during oocyte maturation because large amounts of ATP are required for continuous transcription and translation. 24 Mitochondrial DNA (mtDNA) in oocytes can encode various functional proteins including ATP synthase, cytochrome oxidase, NADH, and pan-reductase. Therefore, mitochondrial dysfunction caused by PFAS can result in excess ROS generation, dissipation of the mitochondrial membrane potential, and early apoptosis. These mitochondria-related problems have been found in oocytes 18,19 and other cell types. [25][26][27] Recently, the cytotoxicity of individual long-chain PFAS on mouse oocytes has been widely studied. However, the data on the cytotoxicity of short-chain PFAS, such as PFBA and PFBS, are still missing.

| Animals
In this study, we used two mouse strains, CD-

| Mouse oocyte in vitro maturation
Female mice (4-6 weeks old) were euthanized and dissected for ovary collection. The ovaries were washed in pre-warmed M2 media, which contained 100 μM IBMX, before the isolation of cumulus-oocytecomplexes (COCs the intensity of signals from the ROI was analysed using Fiji software.
Fixed oocytes were permeabilized in PBS with 0.1% Triton X-100 for 20 min at room temperature, followed by a 20-min incubation in blocking solution (0.3% BSA and 0.01% Tween-20 dissolved in PBS). Primary antibody incubation was performed at room temperature for 1 h.
Oocytes were then washed three times (7 min each) in blocking solution.

| Time-lapse confocal microscopy
Germinal vesicle oocytes were cultured in pre-warmed and equilibrated maturation medium and imaged over time under a 40Â immersion oil objective using a Leica TCS SP8 confocal microscope equipped with a microenvironmental chamber to regulate the temperature and CO 2 at 37 C and 5% in humidified air. SiR-tubulin (Cytoskeleton NC0958386) was added to the maturation medium to label microtubules. 30,31 Bright-field and 647 nm wavelength images acquisition were started at 1.5 h after collection (30-45 min collection time), in which the oocytes were at the GVBD stage. Time-lapse images were taken every 30 min. Z-plane images were captured to span the entire oocyte at 7 μm Z-intervals.

| Parthenogenesis
Metaphase II oocytes (in vitro matured with DMSO, PFHxS or PFOS) were activated to produce parthenogenetic embryos, which were cultured in Ca 2+ /Mg 2+ -free CZB maturation medium supplemented with 10 mM of Strontium Chloride (SrCl 2 ; Sigma 255521) and 5 μg/ml of cytochalasin D (Sigma C2743) for 3 h. The oocytes were then washed, transferred, and incubated in kalium simplex optimized medium (KSOM) supplemented with 5 μg/ml of cytochalasin D for an additional 3 h. The oocytes were then washed and cultured in KSOM for 48 h before assessing parthenote cleavage using a Leica DMi8 microscope.

| Statistical analysis
All experiments were repeated at least three times. The data were presented as mean ± SEM. One way analysis of variance (ANOVA) was used to compare means between multiple groups, followed by Tukey post hoc procedure. Unpaired two-tailed t-tests were used to compare means between two groups. All analysis was done using R (version 4.0.3, Vienna, Austria). Graphs were made using OriginPro 2020 (Northampton, MA, USA). Comparisons were considered significant at *p < 0.05, **p < 0.01, and ***p < 0.001.

| PFAS with a long carbon chain impede mouse oocyte maturation
The reproductive toxicity of PFAS chemicals correlate with their concentrations 19-21 ( Figure 1). However, how the structure of PFAS chemicals influences their toxicity is unknown. To determine the effect of carbon-chain length on the toxicity of PFAS, comparisons of GVBD rate, PBE rate, and relative PB size (size ratio) were conducted among PFCA and PFSA groups. We found the toxicity of PFSA is positively correlated with the carbon-chain length. Specifically, GVBD and PBE rates were normal in the PFBS-and PFHxS-treated oocytes.
However, as the carbon-chain length increased to eight, PFOS treatment resulted in a lower GVBD rate (32.84 ± 7.21% vs. 76.03 ± 1.11% in the control, p < 0.001, Figure 1B,C) and a lower PBE rate (24.93 ± 8.14%, compared to 73.84 ± 1.78% in the control, p < 0.001, Figure 1B,D). Interestingly, we noticed that although PFHxS-treated oocytes had normal GVBD and PBE rates, like PFOS, some oocytes extruded a large PB ( Figure 1B), a phenotype that has been qualitatively described in previous PFAS studies. 17,19,32 Therefore, we calculated the average size ratio between the first PB and its oocyte for each treatment, as described in Section 2.5. We found the average size ratio was 0.14 in the control. In contrast, the ratios were 0.31 in 600 μM PFHxS group and 0.52 in the PFOS group, respectively (p < 0.001, Figure 1B,E). Together, these results suggest a positive relationship between the toxicity of PFAS and the carbon-chain length.
3.2 | PFSA has higher and unique toxicity effects on mouse oocyte maturation compared to PFCA  in the PFOA group, both with p < 0.001, Figure 2B). Based on our finding, we concluded that PFSA has a higher reproductive toxicity than PFCA with the same carbon-chain length.
Next, due to the abovementioned difference, we asked whether PFSA and PFCA can disrupt oocyte maturation in different ways. We increased the concentration of PFOA to 1000 μM to induce abnormality. We found at this concentration, the GVBD and PBE rates decreased to 55 ± 4% and 31 ± 4%, but unlike PFHxS-and PFOStreated groups, the size ratio remained within a normal range ( Figure 2D). This finding indicates that the large PB is a unique phenotype that is associated with PFSA, but not PFCA. To further study the specific toxic effects of PFSA, we selected 600 μM PFHxS-and PFOS-treatment groups to check the mitochondrial function and cytoskeleton structures in mouse oocytes.

| PFHxS and PFOS elevate the level of reactive oxygen species (ROS)
For all cells that undergo aerobic respiration, redox homeostasis is maintained between the ROS generation and scavenging. 34 Various PFAS chemicals have been reported to break the redox balance in mouse oocytes, including PFNA, 20 PFOS, 19 and PFOA. 21 Excess ROS induces oxidative stress, leading to DNA damage, protein malfunction, and lipid chain breakage. 35 To determine if PFHxS can induce oxidative stress in oocytes and to confirm previous results for PFOS, 19 we used DCFH-DA as a probe to detect intracellular H 2 O 2 and oxidative stress. 36 After passively diffusing into the oocytes, DCFH-DA is cleaved by esterase to form DCFH, which can be oxidized to DCF and emit green fluorescence signal. 37 We found PFHxS and PFOS ele-

| PFHxS and PFOS induce mitochondrial depolarization
A positive loop exists between mitochondrion-derived ROS accumulation and mitochondrial depolarization, 38 an event that is universally associated with apoptosis and cell death. 39 To quantify the mitochondrial membrane potential change after PFHxS and PFOS exposure, TMRM, a small cationic lipophilic fluorescent indicator, 40 was used to bind negatively charged mitochondrial membrane. 41 Our results show that the red signal of TMRM was dimmer in PFHxS and PFOS treatment groups, again, depending on the carbon-chain lengths ( Figure 3C). Statistically, the red signal intensity is 3.14 ± 0.09 in control, while it is 2.81 ± 0.10 in the PFHxS-treatment group (p < 0.05 vs. control) and 2.12 ± 0.06 in the PFOS-treatment group (p < 0.001 vs. control and PFHxS group). Therefore, we concluded that 600 μM

| PFHxS and PFOS cause chromosome misalignment and abnormal assembly of spindle and F-actin
Abnormal first PB morphology, including fragmented and enlarged PB, is associated with poor outcomes of in vitro fertilization (IVF). [42][43][44] Therefore, the morphology of the first PB is used as a prognostic factor regarding egg quality in IVF clinics. 42 oocytes compared to control oocytes (4.76%). Importantly, we observed a significant increase in spindle migration failure in PFHxS (20%) and PFOS-treated oocytes (77.78%) when compared to untreated control oocytes (0%). Interestingly, we also found that spindle length-to-width ratio ( Figure 4C) and spindle-length-to-oocyte-diameter ratio ( Figure 4D) significantly increase in PFHxS-and PFOStreated oocytes when compared to untreated controls. This spindle elongation potentially overcomes spindle migration failure. In addition, the chromosome misalignment rate was also increased dramatically as indicated by increased metaphase-plate width after PFHxS and PFOS exposure ( Figure 4E).

| PFHxS and PFOS compromise the developmental competence of oocytes
Parthenogenic activation was used to evaluate the developmental potential of matured oocytes. 49 Considering the mitochondria-and cytoskeleton-related problems we discussed above, we examined the developmental competence of PFHxS-and PFOS-treated oocytes following parthenogenetic activation. We found that cleavage rates of parthenogenetically activated PFHxS-and PFOS-treated oocytes were significantly lower (70.4% and 62.2%, respectively, p < 0.05) than that of control oocytes (92.2%, Figure 6). These results indicate that PFHxS and PFOS not only hinder the nuclear maturation of the oocyte, but also compromise its potential to develop into embryos.

| DISCUSSION
As persistent organic chemicals, PFAS have been detected ubiquitously, even in the liver samples from polar bears in Alaska. 50 In humans, epidemiological studies show that human exposure to PFAS either prenatally or postnatally is related to reproductive defects in both males and females. [51][52][53] Previous studies mostly focused on the toxicity of individual long-chain PFAS, such as PFOA and PFOS. Short chain PFAS such as PFBA were often neglected due to their relatively low bioaccumulation effect. 54 In this study, using a mouse oocyte in vitro maturation system, we demonstrated that the toxicity of PFAS increases with the elongated carbon chain and the inclusion of a sulfonate group. Taking advantage of time-lapse confocal microscopy and immunocytochemistry, we also found aberrant cytoskeleton structure organization in PFAS-treated oocytes.
In terms of the carbon-chain-length effect, consistent with our finding, in other cell types including human colon carcinoma (HCT116) 55 and human hepatocarcinoma (HepG2), 56  We also observed that PFSA toxicity is higher than that of PFCA, which is the case for other cell types including Sertoli cells. 60 The higher cytotoxicity of PFSA is probably due to several reasons. First, the lipophilicity (octanol-water partition coefficient, KOW) of PFCA and PFSA are both very high, implying their ability to cross the cell membranes is limited by their transfer at a membrane/water interface (sorption affinity to artificial phospholipid membranes, KMW), 58 which is higher in PFSA. 61 To prove this, a novel method needs to be developed to precisely detect the amount of PFAS inside a single oocyte. Spindle assembly and migration are also important for successful oocyte maturation and fertilization. In normally developed metaphase I oocytes, microtubules build up barrel-shaped spindles that facilitate chromosome alignment at the metaphase plate. Under the forces of F-actin, the spindles migrate from the cell center to a sub-cortical location to allow an asymmetric division. 75 However, we found many oocytes in PFOS-treated oocytes underwent symmetric division, instead. Our data showed significant increases of F-actin fluorescent intensity around the spindle region after PFSA treatment, together with the loss of key F-actin structures like spindle-associated actin cage. In mouse oocytes, perturbing F-actin dynamic prevents spindle migration towards the cortex. Indeed, treating mouse oocytes with jasplakinolide, a cyclo-depsipeptide actin stabilizer 76 or F-actin inhibitors prevented spindle migration during meiosis I. 45,48 Therefore, these aberrant actin filaments could block the migration of the spindle leading to large PB extrusion in PFSA-treated oocytes. 19 Another unique phenotype we observed in the PFSA treatment groups was elongated spindles. This could be explained by the fact that sulfonate buffers can induce the polymerization of tubulin to enlarge spindles. 61 However, Verlhac et al. proposed that the elongated spindle could be a compensation for oocytes with unmigrated spindles to extrude a normal-sized PB. 77 For example, they found in mos À/À oocytes which lack mitogen-activated protein (MAP) kinase activity, 78,79 non-migrating spindles elongate so that one pole can be closer to the cortex while the other pole remained near the oocyte center. 77 Therefore, some of the mos À/À oocytes still can extrude their first PBs of normal sizes, which is also the case for PFHxS-and PFOS-treated oocytes. However, even though some of the PFHxS-and PFOStreated oocytes can still reach metaphase II stage, their potential to be fertilized and to develop to embryos is compromised ( Figure 6).
In summary, from a female reproduction perspective, we demonstrated that PFAS with a longer chain and a sulfonate group are more toxic and revealed their toxic mechanisms. From an environmental health perspective, short chain PFAS may be less toxic than long chain PFAS according to our results. However, the health concerns regarding their endocrine-disrupting effects still need more research. 80