Per-and polyuoroalkyl substances (PFAS) are synthetic chemicals with useful properties such as water and oil repellence and extreme temperature resistance. 1 These properties led to wide industrial and commercial applications, including in semiconductors, reghting foams, non-stick cookware and food packaging, resulting in exposure of humans and wildlife. [2][3][4][5] PFAS are very persistent in the environment, and once in the body, some PFAS accumulate in tissues. 5,6 Studies have detected PFAS in the blood, liver, kidney, heart, muscle and brain of various species. [7][8][9][10] Based on previous studies, PFAS accumulate in the blood due to binding between PFAS and serum albumin. 6,[11][12][13] The brain ensures its normal functions through uptake of oxygen, nutrients and other required substances from the blood. 14 Substance exchange in the cerebral circulation creates the opportunity for PFAS to enter the brain. However, xenobiotics cannot usually move freely into the brain because of cerebral barriers, such as the blood-brain barrier (BBB) and blood-cerebrospinal uid barrier (BCSFB), that protect the central nervous system (CNS, composed of the brain and spinal cord) by allowing needed chemicals in but not toxins and pathogens. 15,16 This barrier function has been shown to also apply to some PFAS. 17 The BBB is the biochemical boundary of endothelial cells that mediates the exchange of substances between the bloodstream and brain. 18 The link between endothelial cells, known as tight junctions, are responsible for limiting paracellular leakage during substance transport. 19 Various transporters located at the surface of endothelial cells exchange chemicals across the cell membrane, such as P-glycoprotein (P-gp), breast cancer resistance protein (BCRP/Bcrp), multidrug resistance proteins (MRPs/Mrps), organic anion transporting polypeptides (OATPs/Oatps), organic anion transporters (OATs/Oats) and organic cation transporters (OCTs/Octs) (Fig. 1). Chemicals bind to transporters and achieve transmembrane transport through the transporters' conformational changes. 20 Based on previous studies, PFAS could enter the brain by disrupting tight junctions to permeate into the brain [21][22][23] or binding to transporters to cross the plasma membrane. [24][25][26] However, studies related to the interaction of PFAS and transporters mainly focus on renal transporters, 27 while the transport of PFAS through similar transporters at the BBB has yet to be veried. In addition, the specic mechanisms by which different PFAS enter the brain is still unclear, but a number or studies have reported the presence of PFAS in the brain. 7,8,10,17,[28][29][30][31][32][33][34][35][36][37] Biomonitoring studies have detected a broad array of PFAS, including peruoroalkyl carboxylic acids (PFCAs), per-uoroalkane sulfonic acids (PFSAs) and PFAS precursors (compounds that have the potential to be degraded to terminal PFAS, including sulfonamides and uorotelomer substances 38 ) in the brain and cerebrospinal uid (CSF) of humans, and in the brain of wildlife species. 8,17,28,37 The CSF is the uid surrounding the brain and spinal cord, and PFAS content in this uid has been used in a small number of studies as a surrogate for PFAS Fig. 1 The structure of the blood-brain barrier (BBB), the biochemical boundary of endothelial cells between the bloodstream and brain. The link between endothelial cells (dark blue rectangles in inset) are tight junctions (TJ), responsible for limiting paracellular leakage during substance transport. Various transporters located at the surface of the BBB exchange chemicals across the cell membrane (abbreviations refer to different transporters as discussed in the text).
Yuexin Cao is a graduate student in the Department of Civil and Environmental Engineering at the University of Pittsburgh. She works with Dr Carla Ng and her current focus is on using computational methods and zebrash models to evaluate the potential hazards of per-and poly-uoroalkyl substances (PFAS) used in the photolithography industry.
Dr Carla Ng is an Assistant Professor in the Department of Civil and Environmental Engineering at the University of Pittsburgh, with a secondary appointment in Environmental and Occupational Health. Her group's research focuses on the development of models for the fate of legacy and emerging chemicals in organisms and ecosystems. Current areas of active research include development of toxicokinetic models for PFAS in organisms and simulating protein-PFAS interactions to understand PFAS fate and toxicity. She is particularly interested in developing mechanistic models to understand and predict the interactions between emerging chemicals, human activity, and ecological systems. content in the brain interstitial uid. 17,37 In addition, various wildlife biomonitoring and animal exposure studies have also detected the accumulation of PFAS, most frequently PFOA and PFOS, in the brain of various species. 7,10,[29][30][31][32][33][34][35][36]39 In terms of the PFAS distribution data related to specic brain areas, the brain stem, hippocampus, hypothalamus, pons/medulla and thalamus are dominant for PFAS accumulation. 29,30 These brainregion-specic studies are critical for connecting the dominant areas in the brain for PFAS accumulation to the toxic effects of PFAS on the brain, but the available studies are limited. The absorption and accumulation of PFAS in the brain highlight the potential for these substances to cause toxic effects. Studies have reported associations between PFAS exposure and behavioral [40][41][42][43][44][45][46] and cognitive [47][48][49][50][51][52] disorders in both animals and humans, but conicting results of the direction of the association are present in these studies, and the mechanisms underlying PFAS-induced neurotoxicity remain poorly understood. Various in vitro studies proposed two main potential mechanisms, including PFAS-induced intracellular calcium alteration in neurons [53][54][55][56][57][58][59][60][61][62][63] and the impacts of PFAS on neurotransmitters. [64][65][66][67][68][69][70][71][72][73] However, most of these in vitro studies focus on PFOA and PFOS. The neurotoxicity of other peruoroalkyl acids (PFAAs) and emerging PFAS still needs to be evaluated.
In this critical review, we surveyed studies related to the absorption, accumulation, distribution and toxicity of a broad array of PFAS in the brain (Table 1 lists those that are the focus of this review; a more exhaustive list for all PFAS analyzed in the reviewed papers can be found in Table S1 in the ESI †). Based on current understanding, we summarized two potential mechanisms for PFAS to enter the brain, including (1) initiating BBB disassembly through the disruption of tight junctions and (2) relying on membrane transporters. To understand the accumulation and distribution of PFAS in the brain of various species, we surveyed studies of PFAS distributions in collected brain samples from biomonitoring studies and controlled exposure experiments. The brain region-specic PFAS distribution may provide links to observed adverse effects. Finally, we reviewed papers discussing the potential neurotoxicity of PFAS, in terms of effects on calcium homeostasis and neurotransmitters, as well as neurobehavioral and cognitive disorders as outcomes of PFAS exposure.
In this critical review, we used the Web of Science to search for studies using the following search terms: PFAS, brain, bloodbrain barrier (or BBB), transporter, accumulation, distribution, exposure and neurotoxicity. This resulted in 65 papers published between 2005 and 2020, which we categorized into three major subcategories corresponding to the review sections to follow: (1) absorption of PFAS in the brain, (2) accumulation and distribution of PFAS in the brain, and (3) potential neurotoxicity of PFAS.
In the absorption section, we identied and reviewed 11 papers. In the accumulation and distribution section, we identied and reviewed 25 papers, classied into PFAS in collected brain samples (15 papers, see Table 2) and controlled PFAS exposure experiments (10 papers, see Table 3). We calculated PFAS brain-to-blood (or brain-to-serum) ratios where paired brain and blood (or serum) data were available and/or PFAS brain-to-liver ratios if paired brain and liver (but not blood) concentrations were available. These ratios are useful to understand PFAS uptake and/or retention rates in the brain relative to overall exposure, since accumulation for many PFAS is greatest in blood and liver. Several of the studies reviewed did not report their raw data or substance-specic PFAS homeostasis, and 10 papers covering effects of PFAS on neurotransmitters. Since this is an emerging area, we did not perform a systematic review that restricted or characterized studies by quality, but rather reported all available evidence.
Several studies have mentioned barrier functions preventing PFAS from entering the brain. ) suggests that PFOA and PFOS may not cross the BBB freely and/or they are efficiently pumped out from the brain by transporters. 37 Similar ndings, that brain-to-blood ratios of PFOA and PFOS are low in humans, based on post-mortem examinations, were reported in the study of Maestri et al. (2006). 28 In these two early studies, the sample sizes were limited, and the PFAS analysis mainly focused on PFOA and PFOS. Therefore, the results may not necessarily represent the general population, and may not be generalizable to other PFAS given known differences in their toxicokinetics based on animal studies. 74,75 In 2018, Wang et al. pointed out the barrier effect is one potential factor inuencing the penetration of PFAS from the serum into CSF, since PFAS concentrations in CSF are 2 to 3 orders of magnitude lower than in the serum. In addition, they mentioned inammation could increase the permeability of the brain barriers. Albumin CSF-to-serum ratios are strongly correlated with PFAS CSF-to-serum ratios, which may provide an alternative explanation to the barrier theory, since PFAS are known to bind to albumin in the blood. While their study had a relatively large sample size (223 serum-CSF pairs), and analyzed a broad array of PFAS, including PFCAs, PFSAs and emerging alternatives, such as 8:2 Cl-PFESA and 6:2 Cl-PFESA (trade name: F-53B), Wang et al. (2018) noted that the results might not represent the general population, since the paired serum and CSF samples were collected from hospital patients in the Neurological Department. They also mentioned the bias that may come from using the PFAS content in the CSF to represent the PFAS level in the brain interstitial uid. 17 Obviously, measuring chemicals within the brain interstitial uid is challenging. Using the drug content of CSF as a surrogate for the drug content of brain interstitial uid has been demonstrated as feasible, by showing the generated error is less than 3fold. 76 With regard to specic barrier functions, studies indicate different PFAS cross the BBB with varying efficiencies. 7,17 For instance, a pilot whale study by Dassuncao et al. (2019) suggested that certain long-chain PFAS, specically PFDoA, PFTrA, PFTeA and PFDS, may cross the BBB through a process related to the signicantly higher phospholipid levels measured in the brain (though the specic mechanism was not evaluated), while short-chain PFAS may not penetrate the BBB effectively. 7 The current understanding of the potential pathways for chemicals to enter the brain includes (1) initiating BBB disassembly mainly through disrupting tight junctions, and (2) binding to transporters to complete transmembrane transport. However, although PFAS have been detected in the brain and CSF, the mechanisms by which PFAS enter and remain in the brain are unclear. Understanding the possible mechanisms is critical, both for investigating strategies to block PFAS entering the brain so as to limit their adverse effects, and for understanding how to select and design safer replacements for these chemicals. In this section, we review what is known about the uptake of PFAS in the brain and CSF.
Several studies have reported PFOS-induced endothelial discontinuity in the brain. 22,23,77 More specically, PFOS may disrupt tight junctions in brain endothelial cells by triggering the PI3K/Akt signaling pathway. PI3K is a critical regulator of the permeability of endothelial cells. This signaling pathway has been demonstrated via in vitro experiments with the PI3K inhibitor, which blocks PFOS-induced endothelial disassembly. 77 In another in vitro human microvascular endothelial cell model, PFOS provokes the production of reactive oxygen species (ROS). The existence of ROS induces actin lament remodeling, which is directly associated with increased endothelial permeability. 22 Most recently, Yu et al. (2020) reported that PFOS can penetrate the BBB by disrupting the structure of tight junctions and/or decreasing the expression of tight junction proteins (e.g., claudin-5 and occludin). The disrupted tight junctions could then initiate BBB disassembly. Astrocyte hypertrophy and damage have also been found to exacerbate the disassembly of the BBB, as the interaction of endothelial cells and astrocytes is critical for regulating the BBB. PFOS disrupts these interactions and promotes the disruption of the BBB. 23 However, these studies only focused on PFOS. It is still unknown whether other PFAS enter the brain through disrupting the integrity of brain barriers.
The other potential pathway for PFAS entering the brain is by interacting with transport proteins. The traffic of many toxic substances across brain barriers relies on active transport mediated by transporters. 78 Various efflux and inux transporters are expressed at brain barriers, including P-gp, BCRP/ Bcrp, MRPs/Mrps, OATPs/Oatps, OATs/Oats and OCTs/Octs, as illustrated in Fig. 1. 20 While the transport of PFAS through transporters at the BBB has yet to be veried, previous studies related to the interaction of PFAS and similar transporters expressed in other tissues may provide useful insight. For example, several studies have indicated that PFAS renal clearance is mediated by OATs/Oats, 24,25 and PFAS renal reabsorption is moderated by Oatps. 26 In addition, previous in vitro research has investigated the impacts of PFAS on the P-gp transporter, which is one of the most studied efflux transporters at the BBB. 79 Specically, PFOA and PFOS could signicantly inhibit human P-gp, and this inhibition increased with PFAS dose in an in vitro experiment, while the interaction of P-gp with other compounds of low molecular weight (less than 300 Da) was not observed. 80 Another in vitro study on the marine mussel (Mytilus californianus) found that PFOA, PFNA, PFDA and PFHxS have inhibitory effects on P-gp in a chain-length-dependent manner.
That is, longer-chain PFAS caused more severe inhibition of Pgp than shorter-chain PFAS under the same PFAS exposure dose. The mechanism by which PFNA inhibits P-gp is indirect, which means PFNA disrupts the transporter function rather than competing for binding sites with P-gp substrates. But the inhibitory effect of PFNA and PFDA on P-gp is reversible, and exposure of P-gp to PFNA induces the synthesis of new P-gp transporters. 81 Furthermore, an in vitro experiment investigating the interaction of PFOA and PFOS with four types of transporters located at the blood-testis barrier 82 showed both PFOA and PFOS inhibited the activity of the BCRP, P-gp, MRP1 and MRP4, among which the BCRP transporter could transport PFOA as its substrate, while P-gp did not transport any of the PFAS analyzed. 83 This nding of P-gp is in line with the in vitro P-gp study by Stevenson et al. (2006) mentioned previously. 81 When the PFAS acts as an inhibitor of an efflux transporter, such as with P-gp, it reduces the ability of the transporter to effectively remove xenobiotics (including PFAS) from the tissue where it is expressed. Alternatively, when the PFAS acts as a substrate of a transporter, it could compete for binding sites with the normal substrates of the transporter, and thereby limit the transport of the normal substrates, as with the BCRP and PFOA mentioned here. If the transporter is an efflux transporter, then any PFAS that acts as a substrate will be eliminated as would an endogenous substrate. 84 Fatty acid transporters are another potential PFAS transporter group. Greaves et al. (2013) rst found a correlation between long-chain PFCAs (C10-C15) and nonpolar free fatty acids in the brain of polar bears. However, the method they used could not isolate the specic fatty acid types. 29 A recent study in pilot whales demonstrated that phospholipid (one type of fatty acid) concentrations were predictive of the distribution of long-chain PFAS (C12-C14 PFCAs and PFDS) in the brain. 7 The brain takes up the majority of its needed fatty acids from the blood. In order to enter the brain, long-chain fatty acids rely on transporters to cross the BBB. 85 Long-chain PFCAs (C10-C15) may have similar mechanisms to long-chain fatty acids to penetrate the BBB due to their similar structures. 29 Based on the papers discussed in this section, the existence of cerebral barriers prevents xenobiotic chemicals from entering and accumulating in the CNS, but PFAS may enter the brain by initiating BBB disassembly mainly through disrupting tight junctions and/or by relying on transporters to complete transmembrane transport. PFAS are amphiphilic substances composed of a hydrophilic "head' and a hydrophobic carbon-uorine "tail", potentially leading to their ability to cross the BBB. Based on quantitative structure-activity relationship (QSAR) modelling of the relationship between chemical structures and their ability to cross the BBB, molecular weight less than 400 to 600 Da, lipophilicity and protein binding affinity are major factors in CNS penetration. 86 Different PFAS may match this description: the "smaller molecular weight" that makes it easier to penetrate tight junctions falls within the molecular weight range of PFAAs with chain length between C4 and C11. Moreover, previous studies have emphasized that the hydrophobic interactions between uorinated carbon "tails" and the binding pocket of proteins allow PFAS to be substrates of membrane transporters. 87 Compared to long-chain PFAS, shortchain PFAS (which are less hydrophobic) may enter the brain less effectively and/or may be efficiently pumped out from the brain by transporters. For long-chain PFCAs, especially C10-C15 PFCAs are likely to enter the brain through interacting with transporters. 7,29 However, the specic mechanisms of PFAStransporter interactions in the brain are not well understood, and it is also possible that some PFAS may enter the brain through other as yet unidentied mechanism(s). Given the phaseout of many long-chain PFAS and the advent of emerging PFAS alternatives, it is necessary to extend the PFAS types being investigated with respect to uptake in the brain to understand the factors that regulate the absorption of diverse PFAS in the brain. In addition, while in vivo and biomonitoring studies are limited due to the invasive nature of sampling the brain, other methods may be complementary to the current focus on in vitro experiments. For example, computational simulations are an increasingly powerful tool to provide us a better understanding of the uptake of xenobiotic chemicals 88 at the BBB that could be applied to PFAS.
To understand the accumulation and distribution of PFAS in the CNS, we reviewed studies reporting PFAS concentration in human brains and CSF, and in the brains of wildlife species. Basic information related to the samples in these studies is listed in Table 2. The mean PFAS concentrations in the brains, blood and livers in these studies are in (Table S2 in the ESI †).
PFAS have been detected in the brain and CSF of humans. Based on these studies, PFAS content in human CSF is relatively lower than in human brain. 8,17,28,37 The mean concentrations of PFAS in these human samples show a consistent trend, namely that PFCA concentrations decrease with their chain length. 8,17 Another study by Pérez et al. (2013) detected higher mean concentration of PFHxA in the brain of cadavers compared to other PFAS analyzed in their study, 8 but this observation is not supported by the patterns we found in other studies for humans and wildlife. Furthermore, a recently published study suggests these observations may need to be taken with some caution due to potential for the analytical method employed and contamination to generate erroneous results for short-chain PFAS like PFBA. 89 In general, the number and sample sizes of studies related to the distribution of PFAS in the human brain are limited, and some of these studies only focused on PFOA and PFOS. 28,37 Further, the experimental data in these studies are either from autopsy or hospital patients. It is therefore not clear whether PFAS distribution in these samples can represent the distribution in the general healthy population.
In addition to humans, PFAS have also been detected in the brain of various wildlife species. 7,10,[29][30][31][32][33][34][35][36]39 The dominant detected PFAS are C6-C14 PFCAs, and C6, C8 and C10 PFSAs. The concentration of PFCAs with 6 to 11 carbons increases with chain length in the brain. 7,[29][30][31]34,36 The concentration of PFCAs with 11 to 15 carbons shows a uctuating trend wherein the concentration of PFCAs with an odd number of carbons are higher than those with an even number of carbons. 7,29,31,33 According to Greaves et al. (2013), the difference between odd and even chain length PFAS may indicate the presence of precursors of PFAS in biota, such as uorotelomer alcohols (FTOHs), which degrade to both odd and even carbon chain length PFAS. 29 However, this could also be related to the PFAS source: electrochemical uorination (ECF) and telomerization are the two primary methods of PFAS manufacturing, the former yielding both odd and even chain length PFAS, and the latter producing PFAS with an even number of carbons. 90 Thus, sources containing more ECF-derived PFAS would also have a higher proportion of odd chain-length PFAS. Among the PFSAs, PFOS is always dominant in the brain of wildlife, 10,[29][30][31]33 likely due to the wide application of PFOS historically. Additionally, several studies also detected peruorooctane sulfonamide (PFOSA), a PFOS precursor, in brain samples. 7,29,31,35 The existence of PFOSA could increase the content of PFOS in brains.
Among all the papers reviewed that were associated with PFAS distribution in the CNS, several of them reported paired blood (or serum) and brain PFAS concentrations and paired liver and brain PFAS concentrations (see ESI, Table S2 †). The calculated PFAS brain-to-blood (or serum) ratios in wildlife species increased with chain length, suggesting PFAS with longer chain length can enter or remain in the brain more easily. 29,31,33,34 This trend is consistent with the study of Wen et al. (2017) on zebrash (Danio rerio) that long-chain PFAS can outcompete short-chain PFAS for transporters and binding positions, suggesting long-chain PFAS bind to transporters and are transported more effectively than short-chain PFAS. 91 This is in contrast with the more variable data reviewed for humans. This may be because we don't have enough human brain samples to see the trends, or because the mechanisms of accumulation and distribution of PFAS in humans and wildlife species are different, but the former is more likely. focused on the brain region-specic PFAS distribution in polar bears (Ursus maritimus). 29,30 Polar bears are the top predators in their food web, and therefore have higher exposure to bioaccumulative chemicals such as long-chain PFAS. 92 Polar bear samples in these two studies were collected in similar geographical locations in East Greenland, although at different times (see Table 2). Compared to the polar bears hunted in 2006, the mean concentration of PFCAs in the brains of polar bears collected from 2011 to 2012 increased, while the mean concentration of PFSAs decreased. This trend was also reected in the dominant PFAS in each brain region. PFOS was the dominant PFAS in four of eight brain regions in 2006 harvested polar bears, but in only one of the brain regions of polar bears collected from 2011-2012 (Fig. 2C andD). 29,30 This decline likely results from the phase-out of PFOS production in the early 2000s, as was posited in the study by Rigét et al. (2013), who detected the annual average PFAS concentration in the liver of East Greenland polar bears from 1984 to 2011, and found liver PFOS content decreased since 2006. 93 Long-chain C11-C15 PFCAs and PFOS are the major PFAS detected in polar bear brains (Fig. 2A). 29,30 39,94 Compared to other tissues, the dominant PFAS in polar bear brains have longer chain lengths. Greaves et al. (2013) mentioned that the high concentration of longer-chain PFCAs may result from unique transport mechanisms into the brain. Their study was the rst to explore the relationship between PFAS concentration and nonpolar free fatty acids content. They found a positive correlation between long-chain PFCAs (mainly C11-C15 PFCAs) and lipid content. The brain is a lipid-rich tissue, providing a more nonpolar environment for the accumulation of longchain PFCAs, which are more hydrophobic. 29 In terms of the total PFAS content in each brain region, the brain stem, hippocampus, hypothalamus, pons/medulla and thalamus have higher PFAS content than other brain areas. 29,30 These regions are closer to the incoming bloodstream and receive the freshest blood, providing the PFAS in the blood opportunity to accumulate in these brain regions rst. 29 The accumulation of PFAS in these brain regions may have implications for neurotoxicity, as we will discuss in Section 5. Further studies are needed to explore the distribution and accumulation of a broad array of PFAS in the brain and connect them to the neurotoxicity of PFAS.
Various short-term and long-term exposure experiments at a wide range of PFAS concentrations have been conducted on gilthead bream, 95 rainbow trout, 96 zebrash, 97-99 carp 100,101 and rats [102][103][104] to investigate the accumulation and distribution of PFAS in the brain and other tissues. PFAS exposure time, dosage, chain length, functional groups and the age of the test organism have all been shown to affect the accumulation of PFAS in the brain (Table 3).
In terms of PFAS dosage, some studies reported a positive relationship with PFAS brain accumulation. 100,101 Giari et al. (2016) exposed common carp (Cyprinus carpio) to 200 ng L À1 and 2 mg L À1 PFOA, respectively, for 56 days. They found PFOA concentration in carp brain is lower than the limit of detection (0.4 ng g À1 wet weight) at 200 ng L À1 exposure, while they detected PFOA accumulation in brain samples at 2 mg L À1 exposure with the mean concentration of 0.45 ng g À1 wet weight. 101 Dong et al. (2019) also did not detect PFOA in the crucian carp (Carassius auratus) brain at 0.2 mg L À1 exposure at the seventh exposure day. 100 However, another study on rats (Rattus norvegicus) found no obvious difference in PFOA concentration in the rat brain aer 28 days of exposure to either 5 mg kg À1 day À1 or 20 mg kg À1 day À1 PFOA, indicating the saturation of PFOA-protein binding sites at low exposure concentration. PFOA could bind to various proteins in the brain, but increased PFOA elimination through urine or feces will occur when binding sites are saturated. This study also tested the accumulation of PFOS in the brain and found high level of PFOS in the rat brain (146 mg g À1 ) at 20 mg kg À1 day À1 PFOS exposure, while the increase in PFOS bioconcentration was not proportional to the increase in PFOS exposure concentration. Cui et al. (2009) suggested that higher PFOS concentration may cause more serious impacts on the integrity of the BBB, leading to more PFOS penetration into the brain. In addition, the concentration of PFOS in the brain is higher than that of PFOA under the same exposure dose and time, indicating the elimination rate of PFOS might be lower than that of PFOA. 102 The different ndings for PFOA and PFOS might result from their different acid functional groups (carboxylate vs. sulfonate) or the presence of an additional uorinated carbon in PFOS. The study by Wen et al. (2019) pointed out that PFAA accumulation in the zebrash brain is associated with PFAA chain length and functional group. Specically, the accumulation of PFAAs in the brain increases with the peruorinated carbon chain length. This trend may be due to the greater hydrophobic forces that enhance longer chain PFAA binding to proteins. 99 In addition, according to Wen et al. (2017), longer-chain PFAAs might compete for protein binding sites and transporters with shorter-chain PFAAs so as to lead to the observed differences in their bioconcentration potentials. 91 With regard to functional group, compared to PFCAs, PFSAs with the same peruorinated carbon chain length are more accumulative in zebrash brain, since more hydrogen bonds can be formed between amino acid residues and the sulfonate functional group than with the carboxylate functional group. 99 In addition to PFAS dosage and functional groups, differences among individuals in PFAS exposure experiments could also affect the results of PFAS accumulation in the brain. For example, mice at different postnatal ages were exposed at the same dosage of PFOS (50 mg kg À1 body weight). Liu et al. (2009) found higher level of PFOS in younger mice aer the same PFOS exposure, suggesting the development of the BBB function with age provides added protection from xenobiotic accumulation. 104 However, no obvious sex differences in PFAS brain accumulation has been reported in these PFAS exposure studies.
Exposure experiments with PFAS precursors have also been conducted. For example, the study by Zabaleta et al. (2017) explored the exposure of gilthead bream (Sparus aurata) to 29 mg g À1 8:2 polyuoroalkyl phosphate diester (8:2 diPAP), which is a precursor of PFOA, and detected a high level of PFOA (mean concentration 3.7 ng g À1 ) in their brain aer 7 days exposure. However, further studies on the accumulation of PFAS precursors in the brain are needed to investigate whether some PFAS precursors may be more toxic than their degradation products, 105 and to improve the understanding of emerging PFAS.
In nature, organisms are exposed to various chemical contaminants through multiple pathways. The presence of other substances may affect the bioaccumulation and distribution of PFAS in organisms. An in vivo study on the impacts of single-walled carbon nanotubes on the bioaccumulation of PFOS in zebrash tissues found the bioaccumulation of PFAS declines with the increase of nanotube dose, because the adsorption of PFOS to the carbon nanotubes reduces the bioavailability of PFOS to zebrash. 97 This suggests various other environmental contaminants could impact the bioaccumulation of PFAS in organisms, but our understanding of this eld is still not well-established.
Experiments associated with PFAS accumulation and distribution in the brain should be done with particular care. Various studies used aquatic organisms to explore the accumulation and distribution of PFAS in the brain. As mentioned by Vidal et al. (2019) water temperature is critical in the design of studies on aquatic organisms since the distribution and accumulation of PFAS may be affected by metabolic rates, which in aquatic organisms is oen closely tied to water temperature. Specically, they found the brain-to-blood ratios of PFOS increases with water temperature in rainbow trout (Oncorhynchus mykiss). 96 Indeed, ectotherms are very sensitive to temperature, which affects their rates of respiration, consumption, and growth and thereby affect most key toxicokinetic parameters. 106 Ulhaq et al. (2015) mentioned that the brain is a tissue with complex blood vessels, leading to the mixing of PFAS in the blood with PFAS in the brain during experiments, which could also affect the interpretation of brain data. 98 In order to reduce invasive experiments, studies have proposed alternative noninvasive biomonitoring methods to measure internal PFAS exposure. 103,[107][108][109][110] For example, Gao et al. (2015) used hair as an indicator of PFAA exposure, indicating the correlation between average concentrations of PFAAs in hair and brain can reach up to 0.86 or more for PFNA and PFOS. 103 However, studies of using hair as a biomarker of PFAS exposure are still quite limited, and results vary by PFAS types, 103,110 subject population 110 and gender. 103 More studies are needed to explore whether hair can be a reliable biomarker for PFAS exposure by testing different PFAS in more species and optimizing the analytical methods for PFAS detection and quantication in the hair. Taken together, differences in analytical techniques in different studies and challenges associated with experiments may impact our ability to compare results across studies.
The studies reviewed in the previous sections demonstrate that PFAS accumulate in and distribute through the brain, which highlights the importance to better understand the toxicity of PFAS in the CNS. 29,30 In this review, we surveyed 13 studies related to the associations of PFAS exposure with behavioral and cognitive disorders, mainly including attention decit hyperactivity disorder (ADHD), fetal congenital cerebral palsy, learning disorders, memory dysfunction, and intellectual disability. [40][41][42][43][44][45][46][47][48][49][50][51][52] In addition, various in vitro PFAS exposure experiments have been conducted, mainly on hippocampal neurons, to further explore the mechanisms of PFAS toxicity in the brain. Hippocampal neurons are promising subjects, since the hippocampus is one of the dominant brain areas for PFAS accumulation as discussed above. Also, the hippocampus is related to learning and memory. 111 Here we reviewed 21 papers related to the two most studied potential mechanisms of PFAS neurotoxicity: (1) PFAS-induced intracellular calcium alteration in neurons, [53][54][55][56][57][58][59][60][61][62][63] and ( 2) the impacts of PFAS on neurotransmitters. [64][65][66][67][68][69][70][71][72][73] 5.1 Associations of PFAS exposure with behavioral and cognitive disorders
PFAS have been identied as potential neurobehavioral toxicants, e.g. as inducers of behavioral disorders. Multiple studies have explored the prevalence of PFAS exposure and ADHD, but conicting results exist in these studies, including positive, negative and no associations. [40][41][42][43][44][45] Specically, a Norwegian birth cohort study with 1199 mother-child pairs found that higher PFOS concentration in breast milk (collected before infants reached 2 months) increased the odds of ADHD in children (around 13 years old; odd ratio ¼ 1.77, 95% condence interval: 1.16, 2.72). The positive association between early-life PFOS exposure and ADHD was sex-specic, showing stronger association in girls than boys. 40 Negative associations of PFAS prenatal exposure with ADHD were also reported in multiple studies. For example, a questionnaire-based study with 282 subjects found prenatal PFNA exposure was negatively related to ADHD in 7 year-old children. 41 The study by Stein and Savitz (2011) reported the negative prevalence of PFOA exposure and ADHD in 5-to-18 year-old children living in areas where drinking water was contaminated by PFOA. 42 Stein et al. (2014) indicated that the negative association might be because PFOA could slightly activate peroxisome proliferator-activated receptor (PPAR) gamma, acting like PPAR-gamma agonists, which harbors neuroprotective and anti-inammatory functions. Similar functions of activating PPAR-gamma between PFOA and PPARgamma agonists suggests PFOA might also have neuroprotective function. 43 This explanation of negative association between PFOA exposure and ADHD might be extended to other PFCAs due to their similar structures. In addition, no signicant association was found between prenatal PFAS exposure and parent-reported ADHD in 18 month-old children, but the sample size (n ¼ 59) was small in that study. 44 Another study with 4826 mother-child pairs also did not nd the prevalence of PFOS and PFOA prenatal exposure and ADHD (odds ratios ranging from 0.96 to 1.02), but in their stratied analyses, increased association of PFAS exposure and ADHD were found in female infants, and in infants from nulliparous or loweducated mothers. The sex-dependent results might result from different endocrine-disrupting effects of PFAS on estrogen, that thereby cause different impacts on males and females. 45 Besides these prenatal PFAS exposure studies, researchers also explored the relationship between PFAS levels in children's blood and their ADHD symptoms. For example, the study by Stein et al. (2014) found sex-specic prevalence of serum PFOA content and ADHD in 6-to-12 year-old children. That is, serum PFOA level was positively associated with ADHD in boys, but negatively in girls. 43 In addition to ADHD, Liew et al.
(2014) conducted a case-cohort study and found higher concentrations of PFOA, PFOS and PFHpS in maternal plasma could increase the risk of cerebral palsy only in male infants, which might result from the limited sample size of female infants and/or the existence of the sex-related mechanisms, which need to be further explored. 46 Evidence is mixed in these human data, indicating further replications is needed to better understand the associations of PFAS exposure with human behavioral disorders.
According to animal exposure experiments, both short-chain and long-chain PFAS could induce cognitive disorders. [47][48][49] The neonatal exposure of mice to PFHxS affected cognitive function in a long-lasting or even persistent manner. 47 PFDoA decreased the ability of adult rats to recognize novel objects in a dosedependent manner; the cognitive decit became more severe as PFDoA concentration increased in the brain. 48 Another PFOS exposure study indicated that both prenatal and postnatal PFOS exposure decreased the spatial learning and memory abilities in rat offspring, and the reduction induced by prenatal PFOS exposure was more severe. 49 In addition, PFAS-induced cognitive decits have also been reported in humans. [50][51][52] In the study by Skogheim et al. (2020) with 944 mother-child samples, the PFAS concentration in maternal plasma was used to represent child prenatal PFAS exposure. They observed weak negative associations between non-verbal working memory in preschool children and their prenatal exposure to PFAS, including PFOS, PFOA and PFHpS; and weak positive prevalence of verbal working memory in preschool children and their prenatal exposure to PFAS, including PFNA, PFDA and PFUnDA. 50 Similarly, positive association between higher PFAS serum concentrations and cognition limitations (self-reported difficulty remembering) in 1766 adults between 60-85 years old was reported by Power et al. (2013). 51 However, Vuong et al. (2019) did not observe signicant associations between either prenatal or childhood PFOS and PFHxS exposure and the alteration of cognitive functions based on the Full Scale Intelligence Quotient (FSIQ) measurement of 8 year-old children. Aer stratied analyses, they found positive associations between prenatal PFOA exposure and higher IQ in females, and between childhood PFOS exposure and higher IQ in males. 52 Their ndings may reect some other indicators, rather than a causal relationship.
To sum up, we found conicting results in these studies focusing on the association of PFAS exposure with behavior and cognitive disorders, especially in humans. This is probably because these studies have diverse sample sizes, and/or diverse experimental subjects with different ages, living areas and health conditions. It is also worth noting that most of the studies related to ADHD use parentally-reported symptoms, which might compromise the accuracy of the results. Several studies indicated sex-specic associations in stratied analyses, but it is difficult to determine the mechanism underlying these associations since the number of existing studies and sample sizes are limited and their results are inconsistent. Further studies are needed in this eld to validate these ndings. It is worth further exploring the mechanisms underlying such PFASinduced neurotoxicity. In the following sections, we reviewed two potential mechanisms of PFAS neurotoxicity.
Calcium (Ca 2+ ) is responsible for mediating multiple neuronal processes, such as proliferation, synaptogenesis, apoptosis, and neurotransmitter secretion. 112,113 Various PFAS exposure studies have reported effects of PFAS on calcium homeostasis in neurons, which is considered to be one potential mechanism of PFAS neurotoxicity. [53][54][55][56][57] The PFAS-induced calcium increase in neurons is either from extracellular calcium inux or calcium store release (Fig. 3). 113 Liao et al. (2008) found PFOS could induce the inux of extracellular calcium through L-type voltage-gated calcium channels (L-VGCCs) in rat hippocampal neurons. 53 Another study by Liu et al. (2011) found both PFOA and PFOS could signicantly increase the calcium concentration in cultured rat hippocampal neurons. The increased calcium was mainly released from intracellular calcium storage organs, such as mitochondria and the endoplasmic reticulum (ER), 112 and mediated by inositol 1,4,5-trisphosphate receptors (IP 3 Rs) and ryanodine receptors (RyRs) at the surface of calcium stores. 54 Studies have linked calcium overload to neuron dysfunction and even to cell apoptosis. 53,54 Specically, aer acute exposure of hippocampal neurons and brain slices to PFOS, the increased intracellular calcium potentiated synaptic
transmission, which represents the communications between neurons. In addition, PFOS-induced intracellular calcium overload also provoked neuronal excitement, which could lead to neuronal injury. In terms of long-term implications, the exposure to PFOS affected the normal structure and functions of neurons. 53 A further study by Liao et al. in 2009 pointed out that the effects of PFAS on rat hippocampal neurons depend on the chain-length, the degree of uorination and functional groups of PFAS. Specically, the disturbance of neuronal activities by PFAS increased with the uorinated carbon chain length and the uorination level. Compared with peruorinated carboxylates, peruorinated sulfonates had stronger effects on neurons. 58 Additionally, Liu et al. (2011) also observed the increase of ROS in calcium-overloaded neurons. ROS could induce oxidative stress events, which may eventually lead to cell death. 54 Furthermore, Dusza et al. (2018) pointed out the rise of calcium release depends on age, since they found exposure to PFOS increased the calcium release in brain microsomes in adult rats, but not in neonatal rats. 56 Studies have also reported PFAS-induced alteration of calcium-dependent signaling molecules, a potential molecular mechanism of PFAS-induced neurotoxicity since these molecules are critical for determining the structure and functions of neurons. 54,57,[59][60][61] Ca 2+ /calmodulin-dependent protein kinase II (CaMKII), cAMP-response element binding protein (CREB) and calcineurin (CaM) are critical calcium signaling down-stream molecules. 54,59 CaMKII participates in synaptogenesis and plays important roles in learning and memory. 59,61 CREB is critical in neuronal growth and the formation of long-term memory. 114 CaM is important for neuron survival and cognition. 54 PFOS was shown to increase the expression of CaMKIIa and phosphorylated CREB in adult male rat cortex and hippocampus. 59 The expression of CaM signicantly increased in both PFOA and PFOS treated rat hippocampal neurons. 54 Furthermore, to probe the developmental neurotoxicity of PFAS, studies investigated various calcium-dependent signaling molecules in different developmental stages of mice aer PFAS exposure. 60,61 Liu et al. (2010a) detected that the expression of Nmethyl-D-aspartate receptor subtype-2B (NR2B), CaM, CaMKIIa and CERB changed both under prenatal and postnatal PFOS exposure in mice. NR2B expression is related to learning ability and memory; CaM can respond to calcium concentration changes, which has been detected in a PFAS exposure study we mentioned previously; 54 the change of CaM will further impact its downstream molecule CaMKIIa; and CERB is related to neuronal growth. They suggested the relationship between the alterations of the expression of these calcium-related signaling molecules and cognitive decits. Based on their results, PFOS could reach the brain at the embryo stage, and further induce adverse effects to the CNS postnatally. 60 In addition to these molecules, PFOS and PFOA were also shown to increase the level of growth-associated protein-43 (GAP-43), synaptophysin and tau in mouse hippocampus and cerebral cortex aer neonatal exposure. These proteins play important roles in synaptogenesis, neuronal development, and growth. The neonatal stage is a critical brain development period, and PFAS-induced overexpression of these proteins at the neonatal stage affects the healthy development of the mouse brain. 61 Finally, PFNA could also induce increased intracellular calcium concentration and CaMKII expression in rat pheochromocytoma-12 (PC12) cells. These alterations could result in oxidative stress in cells and ultimately lead to cell apoptosis. 57 This observation is in line with Wang et al. (2015), who found prenatal and postnatal PFOS exposure could increase hippocampal neuron apoptosis in rat offspring. The increase of apoptosis is in a similar manner to the calcium increase in neurons, suggesting the rise of intracellular calcium is one of the potential mechanisms of neuron apoptosis. Specically, Wang et al. (2015) indicated that PFOS-induced calcium disturbance in neurons injured calcium signaling pathways, then induced neuronal apoptosis, and eventually could cause behavioral decits, such as ADHD and response inhibition. 55 However, studies also found approaches to reduce PFAS-induced neuronal dysfunctions. For example, Oh et al.
(2018) pointed out phycoerythrin-derived peptide of Pyropia yezoensis (PYP) could alleviate PFOS-induced calcium disorder. 62 A recent study by Zhang et al. (2020) found blueberry anthocyanins (ANT) could reduce PFOA-induced neurotoxicity in Dugesia japonica in terms of locomotion reduction, oxidative stress and neurotransmitter dysregulation. 63 These studies provide insights for alleviating PFAS-induced neuronal toxic effects, but the mechanisms underlying these protective strategies remain to be explored.
The second most studied potential mechanism of PFAS neurotoxicity is neurotransmitter dysfunction. Neurotransmitters are chemicals generated by neurons that are responsible for signal transmission. 115 Neurotransmitter levels in the brain are related to the activation of neurons and signal transmission among neurons. 64 Studies have reported the implications of PFAS on neurotransmitters in the brain, mainly dopamine, [64][65][66][67][68][69][70] glutamate, 64,66,68,[70][71][72] acetylcholine and the cholinergic system. 66,68,69,73 According to various exposure experiments, PFOS and PFOA could alter dopamine concentration in the brains of rat, mouse and frog, but the direction of the alteration was not consistent across studies. [64][65][66] Yu et al. (2016) applied a highthroughput targeted metabolomics approach to analyze the PFOA-induced neurotoxicity in male mice, and found the increase of dopamine concentrations in 0.5 mg PFOA kg À1 body weight day À1 exposure group. 64 In terms of different brain regions, PFOS increased the dopamine concentration in the prefrontal cortex and hippocampus in adult mice aer 28 days PFOS exposure, but the alteration of dopamine content in amygdala was not signicant. 65 However, another PFAS exposure study on Northern leopard frog (Lithobates pipiens) found PFOS and PFOA decreased dopamine in the brain. In addition, this study suggested long-term developmental PFAS exposure could reduce the amount of dopaminergic neurons. Leopard frogs can be more relevant for the study of these neurons compared to rodents, since leopard frogs have neuromelanincontaining dopaminergic neurons, similar to those affected by Parkinson's disease in humans. Therefore, Foguth et al. (2019)
Environmental Science: Processes & Impacts contended that further studies on frogs are needed to explore the relationship between PFAS-induced dopamine alteration and Parkinson's disease. 66 In addition to monitoring dopamine content in the brain, detecting alterations of the gene expression of dopamine receptors further helps to explore the potential molecular mechanism of PFAS neurotoxicity. 65,67 To understand the effects of PFOS on the development of CNS, neonatal mice were exposed to PFOS during development.
Aer 24 hours of PFOS exposure, the transcription of dopamine receptor-D5 decreased in mouse cerebral cortex. At 2 months post exposure, the transcription of dopamine receptor-D2 was reduced in mouse hippocampus. 67 Similar ndings have been reported in the study by Salgado et al.
(2016), namely the gene and protein expression of D1 and D2 receptors in rat prefrontal cortex and hippocampus changed aer exposure to PFOS. 65 Both D1 and D2 receptors play important roles in cognition and memory. Another study by Hallgren and Viberg (2016) considered the decreased transcription of dopamine receptor-D2 in hippocampus may be related to cognition disorders in adult mice. However, they did not explain the reduced D5 receptor in cerebral cortex due to the lack of developmental roles of the D5 receptor in the cerebral cortex among literature studies. 67 In addition to decreasing dopamine, Long et al. (2013) found the exposure of adult mice to PFOS could increase hippocampal glutamate, which is another critical neurotransmitter related to learning and memory. 70 Another study found the glutamate concentration in the brain decreased aer exposing mice to 2.5 mg PFOA kg À1 body weight day À1 for 28 days. 64 Similar results have been reported by Foguth et al. (2020), who included both PFOS alone (10 ppb) and PFAS mixture (4 ppb PFOS, 3 ppb PFHxS, 1.25 ppb PFOA, 1.25 ppb PFHxA and 0.5 ppb PFPeA) exposure groups in their study and found both of these exposures resulted in signicantly decreased glutamate concentrations in the brains of Northern leopard frogs in a similar degree. 68 The low glutamate level in the brain could cause adverse effects on synaptic plasticity and memory. 116 Furthermore, an in vitro study on rat cerebellar granule neurons mentioned that PFOS and PFOA increased glutamate concentration, and in turn induced glutamate excitotoxicity, which means excessive glutamate leads to excessive stimulation of its receptors, and even to cell injury and eventually death. 71,117 However, the degree of excitotoxicity induced by PFOS and PFOA were different, which the authors suggest may be due to the different mechanisms of neurotoxicity caused by PFOS and PFOA. In addition, the glutamate excitotoxicity also varied from the developmental stages of cultured neurons. 71 Liao et al. (2009) found PFOS ranging from 0.1 to 100 mM altered the glutamate-activated current in rat hippocampal neurons. 72 However, the exposure of leopard frog to PFOA and PFOS did not signicantly change the glutamate level in the brain. 66 The study by Foguth et al. (2020) detected the alteration of diverse neurotransmitters in Northern leopard frogs exposed to PFAS, among which, they found PFAS could alter these tested neurotransmitters, especially acetylcholine. Specically, PFOS and the PFAS mixture described above signicantly increased acetylcholine level in the later developmental stage of frogs, but the mechanism behind this acetylcholine rise was not clear. 68 In addition, the exposure of neonatal mice to PFOS and PFOA damaged the adult cholinergic system, even at low PFAS exposure dose (1.4 mmol kg À1 body weight). 73 However, Foguth et al. (2019) did not observe signicant change of acetylcholine levels in leopard frog brain aer PFOA and PFOS exposure. 66 Based on these observations, the effects of PFAS on neurotransmitters are complex. As Slotkin et al. (2008) mentioned, the mechanism of PFAS impacts on neurotransmitters vary by PFAS types. 69 In addition to PFAS type, the alterations of PFAS to neurotransmitters may also depend on PFAS exposure time and dosage, animal species, and brain/neuro-developmental stages, and these factors are important to consider when comparing across studies.
To sum up, PFAS-induced intracellular calcium alteration in neurons and the impacts of PFAS on neurotransmitters are two major potential mechanisms of PFAS neurotoxicity. There is also a potential link between the effects of PFAS on calcium homeostasis and its effects on neurotransmitters because it is known that the increase of intracellular calcium can trigger neurotransmitter secretion. 54 With respect to all of these PFAS neurotoxicity studies, the majority of them focus on PFOS and PFOA, but researchers found long-chain PFAS can enter the brain more effectively than short-chain PFAS, as was also highlighted in the preceding sections on PFAS absorption, accumulation, and distribution in the brain. Although longchain PFAS have been phased out and replaced by diverse emerging PFAS, they are still present in organisms and in the environment. Information on the neurotoxicity of long-chain and emerging PFAS is still lacking. In addition, many studies have mentioned the different adverse effects resulting from diverse PFAS types, but it is necessary to further explore the specic mechanisms behind observed differences. 54,69 This review has focused primarily on observations of direct effects of PFAS on the brain and associated outcomes, but there are additional, potentially important, indirect impacts of PFAS, for example through disruption of thyroid hormone function in pregnant women which could affect neurodevelopment of the fetus. [118][119][120] Such indirect effects also merit further consideration. Furthermore, more studies are needed to explore PFAS neurotoxicity at the molecular level. Currently, it is still difficult to connect the potential neurotoxic mechanisms to specic brain diseases. Critical data gaps remain not only for neurotoxicity, but also in the whole eld of PFAS toxicology. Exploring these toxicological issues faces similar dilemmas: the existence of thousands of untested PFAS, and the lack of the quantication of the potential health effects associated with PFAS exposure. Data and tools are needed to establish the link between PFAS exposure and toxicity. The framework of adverse outcome pathways (AOPs), for example, which identify the specic molecular events required to cause a toxic effect, could help to analyze the risk of more PFAS more accurately with fewer resources, since in vitro experiments and in silico approaches can be alternatives to in vivo studies to test and screen molecular events linked to specic toxic effects. View Article Online
Research on PFAS absorption, accumulation, distribution and toxicity in the brain is increasing, but many knowledge gaps remain. PFAS may enter the brain through initiating BBB disassembly and/or relying on transporters located at the BBB, but diverse PFAS with different chain-length and functional groups have different abilities to enter the brain. Future studies are needed to specify the mechanism of each PFAS entering the brain, and how the uptake efficiencies are affected by differences in PFAS structure and properties. Aer entering the brain, PFAS have the potential to distribute to and accumulate in different areas of the brain. The available studies related to PFAS distribution in various brain regions are quite limited, as are PFAS accumulation data in human brains. Indeed, experimenting on the brain is invasive and should be done with particular care since the brain is a vulnerable tissue with complex blood vessels. As a result, to reduce the invasive experiments and to make the PFAS-related brain studies more accessible, it will be helpful to nd surrogates (such as the CSF and hair) that can represent PFAS concentration in the brain. In addition to in vivo methods, 3-D tri-culture models have been used for drug screening and disease modeling in the brain. 122,123 Although this technology has not yet been used in the study of PFAS, it is a potentially powerful path forward to explore the absorption, accumulation and effects of PFAS in the brain with an in vitro system that more closely mimics in vivo activity.
Computational methods may likewise be useful alternatives or complements to experiments to explore PFAS toxicokinetics in the brain. More studies are needed to explore the characteristics of the accumulation of PFAS in the brain and the brain region-specic PFAS distribution. These studies help to understand the specic toxic effects to the CNS induced by PFAS since the brain is composed of various regions which are responsible for mediating different functions, such as learning, memory, emotions and movement. In this review, we summarized PFASinduced toxic effects including behavioral and cognitive decits. Although learning and memory disorders have been observed, the link between PFAS exposure to specic diseases, such as Alzheimer disease and Parkinson's disease, remains to be explored. Two primary mechanisms of PFAS-induced neurotoxicity have been proposed: disrupting calcium homeostasis and the alteration of neurotransmitters. However, the existence of disconnects across studies on the toxicity of PFAS in the brain and on potential neurotoxicity mechanisms makes interpretation difficult. For example, studies related to the prevalence of prenatal/postnatal PFAS exposure and behavioral and cognitive disorders and in vitro studies exploring mechanisms of PFAS neurotoxicity do not evaluate consistent PFAS types, exposure concentrations, or model organisms. Finally, to understand PFAS in the brain more comprehensively, we expect future studies to be better aligned between the accumulation of PFAS and PFAS toxicity in the brain. Currently, PFOA and PFOS are overrepresented in the literature. Given that other PFAS have been shown to accumulate in the brain, and the general lack of studies on emerging PFAS, it is important to identify the neurotoxicity of these other ubiquitous environmental contaminants.
In this review, we show the existing evidence from multiple perspectives (epidemiological, in vivo, and in vitro) that PFAS do enter and accumulate in the brain, and there are indications they may have an effect. The importance of addressing gaps in our understanding is that there are potentially thousands of PFAS (with at least hundreds in active use) that haven't been tested: the lack of toxicity of some of them does not mean that the others will be safe. It is important to determine whether "emerging" or replacement PFAS may have more profound neurological effects than others, and to connect the understanding of the absorption, distribution and toxicity of PFAS in the brain.
This journal is © The Royal Society of Chemistry 2021Environ. Sci.: Processes Impacts
Published on 10 September 2021. Downloaded by University of Pittsburgh on 10/8/2021 6:16:08 PM.View Article Online