Plant responses to per- and polyfluoroalkyl substances (PFAS): a molecular perspective

Abstract Per- and polyfluoroalkyl substances (PFAS) constitute a large class of toxic manmade compounds that have been used in many industrial and household products. Dispersion of PFAS in the environment has raised concerns because of their persistence and toxicity for living organisms. Both terrestrial and aquatic plants have been shown to take up PFAS from contaminated soil and groundwater, and to accumulate these compounds inside their tissues. Although PFAS generally exert a low toxicity on plants at environmentally relevant concentrations, they frequently impact biomass growth and photosynthetic activity at higher levels. Uptake, translocation, and toxicity of PFAS in plants have been well covered in literature. Although less attention has been given to the molecular mechanisms underlying the plant response to PFAS, recent studies based on -omics approaches indicate that PFAS affects the plant metabolism even a low concentration. The objective of this review is to summarize the current knowledge about the effects of PFAS on plants at the molecular level. Results from recent transcriptomics, proteomics, and metabolomics studies show that low levels of PFAS induce oxidative stress and affect multiple plant functions and processes, including photosynthesis and energy metabolism. These potentially harmful effects trigger activation of defense mechanisms. NOVELTY STATEMENT Although the uptake, translocation, and toxicity of per- and polyfluoroalkyl substances (PFAS) in plants have been well covered in literature, less attention has been given to the molecular mechanisms underlying the plant response to PFAS. Using results from recent transcriptomics, proteomics, and metabolomics studies, this review article aims to summarize the current knowledge about the effects of PFAS on plants at the molecular level. Several reviews have been published on the effects of PFAS on plants, however, none have focused specifically on the molecular mechanisms of PFAS phytotoxicity.


Introduction
Per-and polyfluoroalkyl substances (PFAS) constitute a large class of manmade compounds that have been produced since the mid-twentieth century (Ghisi et al. 2019).The main chemical feature of PFAS is the presence of two distinct moieties: a hydrophobic, highly-fluorinated carbon chain and a polar, sometimes ionizable, functional group (e.g.,carboxylic or sulfonic acid), which confer the molecule surfactant properties (i.e., amphiphilic compounds) (Wang et al. 2017).On the other hand, due to the high strength of the fluorine-carbon bond, PFAS are characterized by high thermal, chemical, and biological stability (Li et al. 2022a).These unique physicochemical properties have led to the use of PFAS in a variety of industrial and household products, including fluoropolymers, coatings, textiles, food contact paper, firefighting foams, and medical devices.To date, more than 12,000 PFAS compounds are recorded on the US EPA Master List of PFAS Substances (US Environmental Protection Agency 2023).
The widespread production and use of PFAS have led to their release into the environment, in which they tend to accumulate due to their high stability.PFAS are today observed in virtually all compartments of the environment, even in remote areas like the Arctic regions (Ghisi et al. 2019).Due to their hydrophobicity and resistance to biodegradation, PFAS tend to bioaccumulate in living organisms, in which they exert variable toxic effects (Ghisi et al. 2019;Dickman and Aga 2022).PFAS have been found to be toxic to all living organisms, including bacteria, algae, plants, fishes, birds, and mammals (Sinclair et al. 2020).Perfluorooctane sulfonic acid (PFOS), one of the major PFAS found in the environment, figures on the list of compounds regulated by the United Nations Stockholm Convention on Persistent Organic Pollutants (UNSCPOP) (Wang et al. 2017).
Plants have been shown to take up PFAS from contaminated soil and water, and accumulate them in their tissues, where they may undergo transformation (Wang et al. 2020).Although the uptake, accumulation, and toxicity of PFAS congeners have been has been the subject of an abundant literature, fewer studies have focused on the effects PFAS on plants at the molecular level.After a brief overview of the uptake, transformation, and toxic effects of PFAS in plants, the present review summarizes the current knowledge about the molecular effects of PFAS on plants.We have collected results obtained from four types of approaches.i.e., transcriptomics (e.g.,RNA sequencing, RT-qPCR), proteomics (e.g.,LC-MS), metabolomics (e.g.,LC-MS, GC-MS), and measurement of enzymatic activities (e.g.,antioxidative enzymes), in an attempt to identify emergent patterns underlying the effects of PFAS on plants.

PFAS uptake and translocation in plants
Plant uptake and translocation of PFAS have been reviewed elsewhere (Ghisi et al. 2019;Wang et al. 2020;Li et al. 2021c;Huang et al. 2021;Mei et al. 2021;Li et al. 2022a), so we will provide hereafter only a brief summary necessary for understanding the molecular mechanisms covered later in the article.
Plants can be in contact with PFAS through biosolidsamended soil, the application of contaminated irrigation water, and/or contaminated groundwater at PFAS production sites (Ghisi et al. 2019).Different PFAS congeners have been shown to be efficiently taken up by a variety of terrestrial and aquatic plants, including wild species and cultivated crops (Ghisi et al. 2019;Zhang et al. 2020;Wang et al. 2020).The extent of PFAS uptake in plants depends on a multitude of factors, including the soil or groundwater PFAS concentration, the PFAS chemical structure, the plant species, and the soil organic matter (Wen et al. 2014;Zhang et al. 2019;Lal et al. 2020).Only PFAS dissolved in the soil solution are available for uptake by plants.Due to their structure, PFAS can adsorb to the soil organic matter via hydrophobic interactions, hydrogen bonding, and/or electrostatic interactions, which therefore limit the available fraction of PFAS in the soil solution (Lesmeister et al. 2021;Qi et al. 2022).The functional group of PFAS also seems to play an important role on the plant uptake.Although perfluorosulfonic acids (PFSAs) are stronger acids than their perfluorocarboxylic acid (PFCA) analogs, both are expected to exist primarily in dissociated, anionic form in groundwater (pH ¼ 5 -7), making them less prone to adsorb to negatively-charged soil particles (Lesmeister et al. 2021).Due to the larger size of the sulfonate group, PFSAs have higher sorption coefficients (K d ) and are more strongly attached to soil particles, than their PFCA analogs (Mei et al. 2021).Bioavailability of organic compounds is highly complex and other factors affect the uptake of PFAS by plants, including pH, salinity, temperature, and even earthworms (Wang et al. 2020;Mei et al. 2021).
Bioavailable PFAS may travel from the soil solution to the root cortex and vascular system through apoplastic, symplastic, or transmembrane routes.Although no specific transporters are known for xenobiotic organics in plants (Mei et al. 2021), several authors have reported involvement of both passive and active carrier-mediated processes involving water channels (aquaporins) and anion channels (Wang et al. 2020).Based on classical models, uptake of organic contaminants in plants is more effective for moderately hydrophobic compounds, with an octanol-water partition coefficient (logK ow ) between 0.5 and 4.5, which is the case for many, but not all PFAS congeners (Mei et al. 2021).The tendency of organic pollutants to enter the roots is commonly expressed by the root concentration factor (RCF), which for PFAS varies widely, depending of the PFAS structure (e.g., length of the carbon chain, polarity), concentration, exposure medium (e.g., soil or hydroponic solution), and plant species.Although the plant characteristics (e.g., plant protein and lipid content) and environmental factors (e.g., soluble organic matter-SOM, pH, temperature) are playing a role, the RCF of PFAS seems to be largely influenced by the hydrophobicity (logK ow ).For hydroponic plants, the RCF of PFAS has been shown to be positively correlated with hydrophobicity (logK ow ), leading some authors to suggest an uptake mechanism involving sorption to the lipid-rich root surface (Mei et al. 2021).
Following the uptake by the roots, PFAS can be transported to the aerial organs via the vascular system (xylem and phloem), which requires PFAS to penetrate the cells (symplastic route) or to cross the Casparian strip (apoplastic route).Some species without a typical Casparian strip (e.g., tomato, carrots) have been shown to accumulate more PFAS in their aboveground tissues, suggesting involvement of the apoplastic route (Mei et al. 2021).Translocation of chemicals in plants are expressed by the translocation factor (TF), which, in hydroponic experiments, has been show to correlate negatively with PFAS hydrophobicity (logK ow ) (Lesmeister et al. 2021).
Although most studies have focused on the uptake and translocation of PFAS by terrestrial plants, a few publications report uptake and translocation in aquatic plants.Using Typha angustifolia grown in soil exposed to a suite of eight PFAS, Zhang et al. (2020) observed accumulation of longer-chain PFAS in the roots and higher translocation of shorter-chain PFAS to the shoots.Studying the uptake of a series of perfluoroalkyl acids (PFAAs) by two submerged macrophytes (Potamogeton wrightii and Ceratophyllum demersum), Li et al. (2021b) similarly observed higher bioaccumulation of long-chain PFAAs and higher translocation of shorter-chain compounds from roots to shoot.

Toxicity of PFAS for plants
Multiple studies have reported the toxicity of PFAS for plants, although the magnitude of toxic effects, often expressed by the inhibition concentration 10% or 50% (IC10 or IC50), varies widely depending on the toxicity endpoint, the PFAS congener, the time of exposure, the growth medium, and the plant species.Toxicity of PFAS for plants has been reviewed elsewhere and will not be extensively covered in this review (Li et al. 2021c;Huang et al. 2021;Li et al. 2022a).The toxicity of chemicals in plants may be quantified based on multiple endpoints, including germination rate and growth/elongation rate, chlorophyll content, and photosynthetic activity (Li et al. 2020c;Li et al. 2021d;Li et al. 2022a).Li et al. (2022a) reported IC50 in a variety of algae and terrestrial plants ranging from < 1.0 to over 4,000 lM.Reported IC50 values of PFASs were shown to be highly variable depending on the toxicity endpoint and exposure time.For instance, IC50s in Arabidopsis thaliana exposed to PFOA for 7 days were lower (higher toxicity) when based on the shoot biomass than on the root elongation.The IC50 values then decreased significantly when calculated over a longer exposure time (21 days vs. 7 days) (Li et al., 2022a).Photosynthesis has been frequently used to determine the phytotoxicity of PFAS.For instance, Qu et al. (2010) reported that a high concentration of PFOS (200 mg L À1 ) decreased the chlorophyll content in wheat, although lower concentrations (0.1 to 10 mg L À1 ) increased this parameter as compared with non-treated plants.Many authors also reported stimulation of the growth when plants were exposed to lower concentrations and inhibition at higher concentrations (hormesis) (Zhou et al. 2016).
In many cases, PFAS toxicity was only observed at concentrations well above environmentally relevant concentrations (ng/L in water and ng/g in soil) (Li et al. 2022a).
The phytotoxicity of PFAS has been shown to be dependent on their chemical structure.Generally speaking, the inhibitory effect of PFAS on plant growth has been shown to increase with the chain length, even though this trend was not always observed.Exposure of wheat seedlings to PFCAs of different chain lengths (C4 to C8, 2,000 lg kg À1 ) showed inhibition of the shoot biomass that increased with the chain length (Li et al. 2022a).On the other hand, the shoot biomass exposed to the same compounds did not show significant changes or even stimulation of the growth.The functional group of PFAS has also been shown to affect PFAS phytotoxicity, with PFSAs being generally more toxic than their PFCA analogs (Huang et al. 2021;Li et al. 2022a).
Often, the toxicity of PFAS in plants has been related to oxidative stress and generation of reactive oxygen species (ROS), such as hydrogen peroxide H 2 O 2 , hydroxyl free radical OH, superoxide ion O 2 -, and singlet oxygen 1 O 2 , which can induce lipid peroxidation and DNA damage.Other toxic effects of PFAS have been mentioned, including alteration of membranes (due to PFAS surfactant properties) and protein secondary structures (Li et al. 2020c;Li et al. 2022a;Ebinezer et al. 2022).Exposure to PFAS has also been shown to create damage to cellular structures, including the rupture of the cell wall (Li et al. 2020c;Li et al. 2022a).

Transformation of PFAS by plants
Being sessile organisms constantly exposed to natural and manmade harmful chemicals, plants have developed detoxification mechanisms allowing them to degrade a range of organic compounds, including PFAS.Based on the observation that plants can metabolize pesticides, Sandermann (1994) introduced the green liver concept, suggesting a detoxification sequence similar to that which occurs in the liver of mammals: phase I involves "activation" of the toxic molecule increasing its reactivity, phase II consists in conjugation of the phase I-activated compound with a molecule of plant origin (e.g., glutathione) therefore reducing its toxicity, and phase III includes sequestration of the conjugates in plant tissues (e.g., vacuole) (Sandermann 1994).These reactions are mediated by specific enzymes, such as cytochrome P-450s (CYP) (phase I), glutathione S-transferases (GST) (phase II), and ABC transporter proteins (phase III).
PFAS have shown resistance to biological degradation, which originates from the strength of the C-F bonds, the three electron pairs surrounding the fluorine atom, and the shielding of carbon atoms by fluorine atoms.Nevertheless biodegradation of PFAS has been reported in multiple organisms, including microorganisms, worms, fish, mammals, and plants (Butt et al. 2014).In fact, the PFAS most frequently detected in the environment (i.e., PFCAs and PFSAs) are believed to have been originally discharged as PFAS precursors, such as fluorotelomer alcohols (FTOHs) and ethyl perfluorooctane sulfonamides (EtFOSs), which were then transformed by biotic or abiotic processes into PFCAs and PFSAs (Zhao et al. 2018).

Molecular mechanisms involved in PFAS uptake, translocation, and transformation
The development of -omics technologies have revolutionized biological sciences, including plant biology.Although the entire genetic material of an organism is referred to as the genome, the expression of genes generates transcripts (messenger RNAs-mRNAs), which together constitute the transcriptome.Transcripts are then translated into proteins, which together constitute the proteome.Proteins, mostly enzymes, mediate biochemical reactions of the metabolism, resulting in the synthesis of a multitude of small molecules forming the metabolome (Abdullah-Zawawi et al. 2022).Studying the transcriptome, proteome, and metabolome constitutes therefore a powerful approach for understanding the molecular of the plant response to stress, including toxic chemicals, such as PFAS.
In the following sections, we attempted to organize the observed molecular effects of PFAS based on the plant molecular functions and biological processes, including response to oxidative stress, photosynthetic processes, protein and amino acid metabolism, lipid metabolism, and response to xenobiotics.Figure 1 summarizes the major molecular effects of PFAS on the plant cell.A summary table (Table S1) is provided as supplemental document.

Reactive oxygen species and response to oxidative stress
One of the most commonly reported effects of toxic contaminants on plants is the induction of oxidative stress and generation of ROS, which can damage multiple plant components, including membranes (e.g., membrane peroxidation), proteins, and DNA (Wielsøe et al. 2015;Huang et al. 2021;Dickman and Aga 2022).ROS in plant tissues are typically removed through enzymatic reactions, involving superoxide dismutase (SOD), catalase (CAT), peroxidase (POD), ascorbate peroxidase (APX), or glutathione reductase (GR).Non-enzymatic removal mechanisms of ROS also occur through reactions with antioxidant molecules, such glutathione (Li et al. 2022a).
Most studies focusing on the effects of PFAS in plants included measurements of the plant biomass and photosynthetic activity (see Section 3 -Toxicity of PFAS for plants).Other parameters frequently measured include the activity of antioxidative enzymes, as well as the concentration of ROS and antioxidant molecules in PFAS-exposed plant tissues.In a study in which the aquatic plants Eichhornia crassipes and Cyperus alternifolius were exposed to PFOS, Li et al. (2020c) reported an increase of the CAT activity and malondialdehyde (MDA).These effects were mostly observed in plants exposed to the highest concentration and were indicative of oxidative stress and lipid peroxidation.Using wheat exposed to PFOA, Zhou et al. (2016) observed a dose-dependent increase of the POD activity (from $18,000 in the controls to $32,000 U g À1 min À1 in plants exposed to 800 mg kg À1 ) and proline content (antioxidant molecule) in exposed seedlings.However, the authors also reported a dose-dependent decrease of CAT activity in exposed seedlings (from 190 in the controls to 6.5 U g À1 min À1 in plants exposed to 800 mg kg À1 ).Often the different makers of oxidative stress used showed variable trends.Another example is provided by Li et al. (2022b) who reported an increased in MDA, H 2 O 2 , and CAT activity in Hydrilla verticillate exposed to a mixture of 12 PFAS congeners (1, 10, and 100 lg L -1 ), although activities of SOD and POD were not affected or even slightly decreased.
One emergent pattern is the stimulation of antioxidative enzymes by low PFAS concentrations and inhibition at higher concentrations, which suggests that elevated PFAS levels can damage the enzymatic antioxidant system.In a study involving exposure of the wetland plant Juncus effuses to seven common PFAS, Zhang et al. (2019) observed that the SOD and CAT activities increased in the shoots but decreased in the roots of exposed plants, which was explained by the inhibition of the enzymes by high amounts of long-chain PFAAs, especially PFOS, accumulating in the roots.On the contrary, low PFAAs concentrations in the shoots resulted in stimulation of the antioxidant system and activation of SODs and CATs.Several other authors have suggested that exposure to higher PFAS concentrations may inactivate antioxidative enzymes, either directly by altering the protein structure or indirectly through the generation of ROS (Qu et al. 2010;Zhou et al. 2016;Li et al. 2020c).
This distinct effect of PFAS at low and high concentration was also reported by Fan et al. (2020) who conducted a whole-genome transcriptomic analysis (RNA sequencing) on the model plant Arabidopsis thaliana: in the shoots accumulating high levels of PFOA, the authors observed downregulation of genes involved in response to oxidative stress, including CAT2, APX1, sAPX, and Fe-SOD, while no changes were observed in roots accumulating low levels of PFOA.These results were consistent with the recorded activities of the antioxidative enzymes APX, CAT, and SOD, which were found to decrease in the shoots following exposure to PFOA, while no significant changes were recorded in the roots.These observations suggest transcriptional regulation of antioxidative enzymes in tissues exposed to PFAS.Liu et al. (2022) analyzed gene expression in the alga C. pyrenoidosa exposed to PFBS and FBSA, which revealed downregulation of SOD (SOD2) and peroxidase genes (PRDX5), and upregulation of glutathione peroxidase genes (GPX).The authors then proposed that these changes would increase H 2 O 2 and decrease free glutathione in exposed tissues, explaining the phytotoxic effect of PFAS.The downregulation of SOD2 was consistent with a decrease of the SOD activity in exposed plants.However, results at the transcriptomic level were not always in agreement with the recorded enzymatic activities.In their experiment with E. crassipes exposed to PFOS, Li et al. (2020c) observed downregulation of a SOD gene (Mn-SOD) in plants exposed to high level of PFOS (10 mg L À1 ), while no significant change in SOD activity was measured.In another transcriptomic analysis, exposure of soil-grown soybean (Glycine max) plants to the short-chain PFBA did not result in significant changes in the level of expression of SOD and CAT, although the activities of these enzymes decreased in plants exposed to even low concentration (100 ng/L), suggesting posttranslational inhibition of the enzymes by PFBA (Omagamre et al. 2022).In this study, enrichment of pathways involved in alkaloid biosynthesis (e.g., isoquinoline, tropane piperidine, and pyridine) led the authors to suggest a non-enzymatic response to ROS.
Several metabolomic studies then reported an increase of plant metabolites involved in response to oxidative stress after exposure to PFAS, also supporting the importance of non-enzymatic responses.For instance, Li et al. (2020a) investigated the metabolome of hydroponic lettuce leaves exposed to PFOA and PFOS (500 and 5,000 ng/L) and observed an increase in antioxidant molecules, including amino acids, such as tryptophan (precursor of glutathione) and phenylalanine (shikimate phenylpropanoid pathway), phenolic compounds, melatonin, N-acetylserotonin glucuronide, thiamin pyrophosphate, and caffeic acid.The same group and others have similarly reported an increase of antioxidant molecules following exposure to PFAS, although in some cases the antioxidant content tended to decrease ).However, the response to oxidative stress in PFASexposed plants revealed a high variability.Although some reports indicated an increase in the antioxidant response in plants exposed to PFAS, in many cases, different oxidative stress markers led to contradictory results.ROS in plants are naturally produced through respiration and photosynthesis, and different types of stress may disrupt cell homeostasis leading to the accumulation of ROS.The complex interactions between ROS and antioxidative mechanisms include feedback loops making difficult to draw a consistent pattern: e.g.,the antioxidant molecule glutathione may increase as a response to oxidative stress or decrease by reaction with ROS (Liu et al. 2022).In addition, besides being toxic species, ROS are also signaling molecules involved in the stress response.Regulation of ROS therefore operates at different levels for either signaling or detoxification purposes (Mittler 2002).

Photosynthesis and energy metabolism
The photosynthetic system is very sensitive to external stimuli and constitutes therefore a popular marker of plant health.Almost every study focusing on the effect of PFAS in plants reported the impact on photosynthetic parameters, including chlorophyll content, chlorophyll autofluorescence, assimilation rate, and/or stomatal conductance.
Often the effects on the photosynthetic system is dependent on the PFAS concentration.For instance, Omagamre et al. (2022) reported an increase of the chlorophyll content in soybean plants exposed to low PFBA concentration ( 100 ng L -1 ), but a decrease at high concentrations (100 lg L -1 and 1 mg L -1 ), which was supported by the downregulation of genes involved in photosynthesis, and the synthesis of chlorophylls and carotenoids.In their study on aquatic plants, Li et al. (2020c) observed that a high level of PFOS (10 mg L À1 ) did not result in significant changes in the chlorophyll content in E. crassipes, while RT-qPCR analysis resulted in downregulation of genes involved in photosystem synthesis (psbA, psbC, and psbD).In C. alternifolius, PFOS resulted in an increase of the chlorophyll content at very low PFOS concentration (0.001 mg L À1 ) and upregulation of psbC at higher concentration (10 mg L À1 ).Tang et al. (2020) investigated the effects of 6:2 chlorinated polyfluoroalkyl ether potassium sulfonate (F53B) on water spinach (Ipomoea aquatica).Using RT-qPCR, the authors reported significant overexpression of genes involved in photosynthesis, including psbA, encoding the PS II-RC (photosystem II reaction center), and rbcL, encoding the RubisCO large subunit, upon exposure to F53B.
Two studies using the alga C. pyrenoidosa reported various effects of PFAS on the chlorophyll content and photosynthetic activities (Li et al. 2021d;Liu et al. 2022).Transcriptomic analysis showed changes in the expression of genes involved in photosynthesis and the energy metabolism.For instance, genes involved in the synthesis of photosynthesis-antenna proteins (light-harvesting complex I and II, LHCA4, LHCB4, and LHCB5) were downregulated by exposure to PFOA and perfluoro-2-methyl-3-oxahexanoic acid (GenX), but upregulated by exposure to the short-chain PFBS.The study also showed differential expression of genes involved in the synthesis of PS II (PsbA, PsbC, and PsbB), protoporphyrin and protochlorophyllide (chlD and por), and cytochrome b6f complex (petN, petC, and petF) in response to PFBS and FBSA.
Beside photosynthetic processes, several authors observed that PFAS also affected the plant energy metabolism, including carbon fixation, ATP synthesis, carbohydrate metabolism, and tricarboxylic acid (TCA) cycle (Ebinezer et al. 2022;Li et al. 2020a;Li et al. 2020b;Li et al. 2021d).For instance, in their metabolomic study on lettuce plants, Li et al. (2020a) and Li et al. (2020b) observed changes in the abundance of key metabolites involved in the TCA cycle, including pyruvate, succinate, and glyoxylate.Similarly, in their study on the alga C. pyrenoidosa, Liu et al. (2022) reported that PFBS and FBSA caused upregulation of genes encoding enzymes of the TCA cycle, such as the ribulosebisphosphate carboxylase large chain (rbcL) and phosphoglycolate phosphatase (PGP).
PFAS have been shown to impact both photosynthesis and the energy metabolism in plants.However, a clear pattern is difficult to identify because the observed effects are frequently inconsistent across studies.This may be due to the variety of methods used to characterize photosynthetic activities (e.g., chlorophyll content, chlorophyll fluorescence, gas exchanges), as well as the nonspecific effects of PFASinduced oxidative stress.For instance, PFAS may induce overexpression of genes involved in photosynthesis, although oxidative damage to proteins or membranes may lead to a reduction of the photosynthetic activity (Li et al. 2022a).

Effect on lipids and membranes
As described earlier, several authors have reported an increase of ROS and MDA in plants exposed to PFAS, which suggests toxic effects through lipid peroxidation (Qu et al. 2010;Chen et al. 2020;Li et al. 2020a;Li et al. 2020c;Li and Li 2021;Li et al. 2022b).On the other hand, because of their surfactant properties, PFAS may increase the permeability of biological membranes, therefore potentially enhancing nutrient uptake by the plants, and hence, increase the plant biomass at low concentrations (hormesis) (Li et al. 2020c;Li et al. 2022a).
At the molecular level, the effects of PFAS on the lipid metabolism and membrane integrity have been supported primarily through proteomic and metabolomic evidence.In their proteomic study of maize plants exposed to a mixture of 11 PFAS, Ebinezer et al. (2022) observed a significant increase of the NSLTP1 protein involved in lipid transport and metabolism.The analysis of fatty acids in exposed tissues confirmed alteration of the abundance of 8 fatty acids, which the authors related to PFAS surfactant properties, as it was observed in non-plant species (Dickman and Aga 2022).Studies on model organisms indicate that PFAS are capable to activate the peroxisome proliferator-activated receptor alpha (PPARa) involved in the regulation of the lipid metabolisma similar mechanism could take place in plants, although currently not demonstrated (Ebinezer et al. 2022).
In their study focusing on the metabolome of lettuce plants exposed to PFOA and PFOS, Li et al. (2020aLi et al. ( , 2020b) ) reported an alteration of the lipid composition and the fatty acid metabolism.They proposed that the observed reduction of linoleic acid was indicative of PFAS-induced membrane modification to improve stress adaptability and ROS removal.Studying the effects of PFOA and PFOS on lettuce roots, Li et al. (2020b) reported a decrease in linolenic acid derivatives (17-hydroxylinolenic acid) which they interpreted as a repair mechanism of damaged membranes.Alteration of the fatty acid metabolism was also observed at the transcriptomic level, as it was reported by Li et al. (2020c) in their study of aquatic plants exposed to PFOS in a constructed wetland.
Although multiple studies have demonstrated that PFAS can affect biological membranes, it is unclear whether these effects originate from lipid peroxidation via PFAS-generated ROS or directly from the PFAS surfactant properties.Molecular dynamics simulations on human cells have shown that PFOS could enter the phospholipid bilayer, and alter the permeability and molecular organization of cellular membranes (Dickman and Aga 2022).Alteration of biological membranes by PFAS can then affect photosynthesis, energy metabolism, and virtually every cell function.

Amino acids and protein metabolism
Although several authors have reported that PFAS affect the amino acid and protein metabolism, their ubiquitous interactions with every other plant processes and functions make difficult to draw clear conclusions.Based on enrichment analysis, transcriptomic studies indicated alteration of the amino acid, peptide, and protein metabolism (Li et al. 2020b;Li et al. 2020c;Ebinezer et al. 2022;Liu et al. 2022).
In their proteomic study of maize exposed to a mixture of PFAS, Ebinezer et al. (2022) identified 75 differentiallyabundant proteins in exposed plants versus control plants (fold change 1.3 or !1.3, p < 0.05).A reduction of proteins involved in translation, including 11 ribosomal proteins, indicated inhibition of protein synthesis.The increased abundance of serine-type endopeptidases suggested a plant response to restore protein homeostasis, which was supported by changes in other proteins related to trafficking, autophagy, and proteasome-mediated ubiquitin-dependent protein catabolic processes.Change in the abundance of proteins involved in protein folding (e.g., molecular chaperones) and posttranslational repression (e.g., ribonucleases, mRNA decapping proteins) also suggested an impact of PFAS on the mRNA turnover under stress conditions.
Interaction with proteins seems to be a major mechanism by which PFAS act on biological systems.As observed with lipids, PFAS may act directly through interaction with protein molecules or indirectly through oxidative damage.Research using human and animal models have pointed toward strong binding of PFAS to proteins.For instance, GenX was shown to bind human serum albumin (HSA) at multiple sites, causing changes in the protein conformation (Dickman and Aga 2022).The capability of PFAS to bind proteins may affect plant functions, such as enzyme affinities and membrane fluidity via alteration of membrane-bound proteins.Although several reports suggest the effects of PFAS on proteins and lipids in plants, little information is available about PFAS interactions with specific enzymes or molecular mechanisms.

Metabolism of xenobiotic compounds
Plants have the capability to metabolize toxic organic compounds through detoxification reactions (green liver model), involving catabolic enzymes (e.g., GST, CYP) (Sandermann 1994).Activation of the xenobiotic metabolism in plants following exposure to PFAS has been mostly identified through transcriptomic analyses, enzymatic activities, and identification of PFAS metabolites.
As described above, the transformation of PFAS precursors, such as FTOHs, includes aand b-oxidation, sometimes followed by conjugation with glutathione, suggesting mediation of both phase I (e.g., cytochrome P-450) and phase II enzymes (e.g., glutathione S-transferase), as it has been observed with other environmental contaminants (Van Aken et al. 2010;Zhang et al. 2016;Zhao et al. 2018b).
In their study of the transformation of FTOHs in soybean plants, Zhang et al. (2016), directly observed an increase of the activities of the phase I enzymes alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH), potentially involved in oxidation of 8:2 FTOH into 8:2 fluorotelomer aldehyde [F(CF 2 ) 8 CH 2 CHO] and 8:2 fluorotelomer carboxylic acid [F(CF 2 ) 8 CH 2 COOH].The increase of the activity of the phase II enzyme GST was consistent with the detection of glutathione conjugates of 8:2 FTOH metabolites (i.e., 8:2 uFTOH-SCysNAcetyl and 8:2 FTUCA-SCys).Another key enzyme involved in phase I detoxification of xenobiotic compounds, CYP, however, did not increase in 8:2 FTOHexposed plants, suggesting that CYP may not be involved in the initial activation of 8:2 FTOH.
Besides evidence from enzymatic activities and PFAS metabolites, transcriptomic studies on plant exposed to PFAS have shown activation of genes involved in the xenobiotic metabolism.Using A. thaliana exposed to PFOA, Fan et al. (2020) reported an increase in the level of expression of several genes potentially involved in the transport of PFAS (41 and 29 genes in exposed shoots and roots, respectively), including ABC transporters, drug transmembrane transporters, and ion and alternative active transmembrane transporters (e.g.,DTX1 (DETOXIFICATION 1), ABCB11, and ABCC3), suggesting involvement of phase III enzymes on the transport of PFAS in plant tissues.Further examination of transcriptomic data made available by Fan et al. (2020) revealed overexpression of a range of phase I, II, and III enzymes potentially involved in the response of A. thaliana to PFAS, including GSTs, CYPs, and ABC transporter proteins.In their study on the effects of the short-chain PFBA on soybean plants, Omagamre et al. (2022) similarly reported upregulation of multiple phase I, II, and III genes, encoding CYPs, ABC transporter proteins, pleiotropic drug resistance proteins, protein DETOXIFICATION, and pathogenesis-related proteins.
Metabolomic analysis of PFAS and their metabolites using LC-MS/MS, as well as comparison with the metabolism of PFAS in other organisms, have allowed advances in understanding the transformation pathway of biodegradable PFAS in plants.However, as with many other contaminants, specific links between enzymes involved in the xenobiotic metabolism and PFAS transformation have not been identified.More research is therefore needed to understand further the catabolic reactions and regulation of the metabolism of PFAS in plants.

Other effects of PFAS on plants
A few reports indicated that PFAS may induce DNA damage in exposed plants, which is consistent with the consequences of oxidative stress and ROS.In their metabolomic study on lettuce exposed to PFOA and PFOS, Li et al. (2020a) observed an increase of 8-hydroxy-deoxyguanosine (8-OHdG), which suggests DNA damage and subsequent activation of DNA repair mechanisms.In their study on water spinach exposed to F53B and chromium, Tang et al. (2020) detected overexpression of the metallothionein gene LaMT2 in exposed plants, which may indicate DNA damage, although chromium may be the major toxicant causing this effect.Finally, exposure of the alga C. pyrenoidosa to PFBS and FBSA was found to result in the downregulation of DNA polymerase a (POLA2 and POLA1), which may indicate a negative effect on DNA replication (Liu et al. 2022).
Besides generation on ROS (which, in addition to be toxic species, are important signaling molecules), PFAS were shown to affect several signaling pathways involving carbohydrates and ethylene.For instance, Omagamre et al. (2022) reported downregulation of transcription factors involved in the ethylene signaling pathway (i.e., NAC, WRKY, and AP2/ERF) in soybean exposed to PFBA, which the authors related to the decreased SOD activity, and subsequently the increase of H 2 O 2 in plant tissues.
Although exposure to xenobiotics is largely regulated at the transcriptomic level, little information is available about the signaling mechanisms involved in the plant response to PFAS.In mammals, the xenobiotic metabolism is activated by specific xenobiotic receptors (e.g., aryl hydrocarbon receptor-AHR, constitutive androstane receptor-CAR).However, to date no xenobiotic receptors have been identified in plants and regulation of the xenobiotic metabolism in plants has been largely understudied.

Conclusion
Although clear conclusions about the molecular responses of plants exposed to PFAS are difficult to draw because of the variety of plant species, PFAS congeners, and experimental approaches involved, a few trends emerged that are worth noting.PFAS, as rather inert chemicals, exert relatively low toxicity on living organisms, including plants.As a consequence, many analyses focusing only on directly-observable plant metrics (e.g., growth rate, photosynthetic activity) might not detect the impact of PFAS at environmentally relevant concentrations.At the molecular scale, on the other hand, low levels of PFAS often induce clear changes in the plant metabolism, which may have long-term implications for plant health and environmental risk assessment.The most frequently-detected mechanisms of PFAS phytotoxicity include the generation of ROS and oxidative stress.Other mechanisms of action of PFAS on plants include the alteration of macromolecules, such as phospholipids, proteins, and DNA, which could be caused directly by interactions with the PFAS molecules or indirectly through PFASmediated oxidative stress.Damage to these macromolecules may then affect virtually all plant functions and systems.
More research based on -omics methods is needed to understand further the underlying molecular mechanisms of the plant response to PFAS.A few key areas for future research are given below.
Transcriptomic analyses of plants exposed to toxic contaminants have allowed identifying genes involved in the toxic response to these contaminants.In the case of PFAS, only a handful of studies have focused on gene expression.More research is needed to characterize the transcriptomic response and understand the gene regulatory network in plants exposed to PFAS.
As in other organisms, the plant response to stress, including toxic chemicals, has been shown to be regulated through epigenetic mechanisms, including DNA methylation, histone modifications, and small interfering RNAs (siRNAs).More research is needed to understand further the epigenome developing in plants exposed to PFAS.This is particularly important to predict the long-term effects of PFAS and the transgenerational adaptation to these compounds.
Most metabolomic studies on plants exposed to PFAS have focused mostly on known metabolites (e.g., amino acids, fatty acids).Advances in mass spectrometry (MS), nuclear magnetic resonance (NMR), and multivariate statistical analysis have recently allowed to perform nontargeted metabolomics, allowing characterization of the whole-plant metabolome, potentially including unknown metabolites and new signaling molecules.
Finally, computational modeling, including molecular docking and molecular dynamics, has allowed characterization of the interactions between PFAS and animal biomolecules.Similar approaches applied in plant science would help understand further the molecular effects of PFAS.

(
Yang et al. 2015;Li et al. 2020b;Li et al. 2021a;Li and Li 2021).Generation of ROS and oxidative stress are frequently observed in studies on plants exposed to toxic chemicals, including dioxins, heavy metals, and PFAS(Wielsøe et al. 2015;Sinclair et al. 2020;Dickman and Aga 2022;Li et al. 2022a

Figure 1 .
Figure 1.Molecular effects of PFAS on plant molecules and cell structures.The blue lightening bolds indicate the direct interactions between PFAS and plant molecules.The red lightening bolds indicate the effects of reactive oxygen species (ROS) on plant molecules.