Functional and Transcriptional Disruption of Hepatic ABC Transporters by Environmentally Relevant Short‐Chain Per‐ and Polyfluoroalkyl Substances (PFAS)

ABSTRACT

Per‐ and polyfluoroalkyl substances (PFAS) represent a large class of > 11,000 synthetic organofluorine compounds that are globally distributed and consistently detected in human serum. Recent studies suggest that environmentally relevant exposures to short‐chain PFAS can disrupt ATP‐binding cassette (ABC) transporters, which regulate xenobiotic disposition and are critical for human health. This study examined the effects of short‐term (45 min) and long‐term (48 h) exposures to 1 nM or 1 µM of four short‐chain PFAS, perfluorobutanesulfonic acid (PFBS), perfluorohexanoic acid (PFHxA), perfluorohexanesulfonic acid (PFHxS), and 6:2 fluorotelomer alcohol (6:2 FTOH), using differentiated HepaRG hepatocytes. Functional assays evaluated changes in efflux activity, while gene expression analysis quantified transcriptional responses across 38 ABC transporters. Following 48‐h exposures, both 1 nM and 1 µM treatments significantly decreased the retention of fluorescent substrates CMFDA, BODIPY‐cholesterol, Hoechst 33342, and Rhodamine 123 relative to controls, indicating enhanced efflux transporter activity. Consistent with these results, transcriptional analysis revealed significant upregulation of multiple ABC genes, including ABCG2 (BCRP), ABCB1 (P‐gp/MDR1), and ABCC1 (MRP1). In contrast, short‐term exposures produced no measurable effects on efflux activity. Together, these findings demonstrate that environmentally relevant concentrations of short‐chain PFAS can alter the expression and activity of key hepatic ABC transporters. Such changes may disrupt the handling of endogenous compounds and pharmaceuticals, raising concerns that low‐level PFAS exposure could influence drug disposition and human health outcomes.

Low‐dose short‐chain PFAS exposures enhance hepatic ABC transporter activity and expression, revealing that environmentally relevant concentrations can reprogram xenobiotic transport pathways and potentially influence drug metabolism, bile acid handling, and lipid homeostasis, highlighting a molecular link between chronic PFAS exposure and altered hepatocellular function with implications for human health.

1. Introduction

Per‐ and polyfluoroalkyl substances (PFAS) are fluorinated hydrocarbons ranging in chain length and structure which have been used in manufacturing since at least the 1950s [1]. Due to the strength of their fluorine bonds, these chemicals have half lives of thousands of years in the environment, leading to their environmental accumulation [2]. Long‐chain PFAS, defined generally as carboxylic acids with carbons numbering C ≥ 7 or sulfonic acids with C ≥ 6 [3, 4], have been phased out due to their ability to bioaccumulate in humans and cause cancer in rodents, in favor of short‐chain and partially fluorinated alternative PFAS [5]. While short‐chain PFAS have been shown to be less cytotoxic than legacy PFAS [6], and do not have as clear of evidence for their ability to cause overt health effects, recent work has focused more on the human health implications of short‐chain PFAS exposure due to their omnipresence in both the environment and the human body. Short‐chain PFAS have been detected in the serum of over 99% of the general population, with concentrations typically in the nanomolar range [7, 8, 9]. While previous work has indicated in vitro that some of the most prevalent short‐chain PFAS: PFBS, PFHxA, PFHxS, and 6:2 FTOH, are not causing overt cytotoxicity at relevant exposure concentrations [6], novel work suggests these low dose short‐chain PFAS are in fact having a significant impact to cellular health and phase III biotransformation [10]. Phase III biotransformation, responsible for the transport of endogenous substrates, medications, hormones, and vitamins across cell membranes throughout the body, playing a major role in bodily distribution, biotransformation, and excretion, is a central focus within the field of pharmacology [11]. ATP binding cassette (ABC) transporters and solute‐carrier proteins (SLC) are large transport protein superfamilies which primarily efflux and import substrates respectively, with varied substrate specificity and overlap between transporters [12, 13]. While ABC and SLC transporters play major roles in phase III biotransformation and are identified by the FDA as critical to drug development [14], environmental contaminants such as PFAS, which we are continuously unintentionally exposed to, have not been considered in this manner. Novel work from Collier et al. (2025) has indicated that ABCG2 protein expression is significantly altered in vitro following short‐chain PFAS exposure [15], however the potential functional impact of such changes or broader impact to transporters more generally has not been evaluated. Other work has indicated the role of OATPs, from the SLC superfamily, in physically transporting PFAS through cell membranes, and the ability of long‐chain PFAS to alter ABCG2 transport from vesicles following 10 µM and 100 µM exposures [16, 17]. As both PFAS, for which over 11,000 chemicals have been identified [1, 4], and ABC transporters, of which there are 48 sub members [12], are broad topics intricately involved in human health, we aim here to better characterize the extent of ABC transporters impacted on a gene expression level by short‐chain PFAS, and to determine the functional impact of such low dose PFAS exposures on membrane transport activity.

Short‐chain PFAS PFBS, PFHxA, PFHxS, and 6:2 FTOH were chosen for this study based on environmental relevance and evidence of impact on ABCG2 from previous works [15]. Further, these chemicals represent both carboxylic and sulfonic acid PFAS, with chain lengths ranging from four to eight, as well as alternative branched PFAS such as 6:2 FTOH which is partially fluorinated. Dosages of 1 nM and 1 µM were chosen in order to accurately represent human exposures which have been measured in the 1 nM range, as well as a high dose exposure more representative of those highly exposed to these chemicals in an occupational setting [7, 8, 9]. Further, such exposures have previously been demonstrated to be below EC₅₀ values for HepaRG cells [6]. HepaRG were chosen for this model due to the known impacts of PFAS exposure on the human liver, as well as the ability to differentiate HepaRG into hepatic and biliary like cells with established functional expression of many membrane transport proteins including OATP2B1, OATP1B3, OATP1B1, ABCC2, ABCC3, ABCC4, ABCC5, ABCC6, ABCB1, and ABCG2 [18, 19]. While primary human cells are the zenith of in vitro work, HepaRG have been established as a robust model with more similar gene expression to primary cells than other immortalized liver cell lines [20].

ABC and OATP transporters play an intricate role in human health. Impairment of ABC and OATP transport in the liver are associated with drug induced liver injury, which can occur through disruption of bile acid transport or intracellular retention of drugs [21, 22], while overexpression is associated with treatment resistance in cancer and other health conditions [13, 23, 24, 25]. Additionally, differences in ABC transport are associated with gender, age, and ethnicity differences in the efficacy of immunosuppressant medications [26]. In order to better understand how relevant environmental exposures may impact the ability of human cells to transport both endogenous and pharmaceutical substrates normally in vivo, a relevant cell line with endogenous expression of these transporters and retained pathways for the regulation of these proteins [18, 19], rather than vesicles or overexpressing cells, were chosen. While animal models have been used to evaluate the functional transport of ABC and OATP in vivo, evidence of species specific and sex specific differences in ABC and OATP transport and regulation indicate a need for human specific considerations of membrane transport [12, 27, 28]. ABC primarily efflux substrates across cell membranes, OATP primarily import substrates, with some transporters expressed in the liver localized to the sinusoidal or canalicular membranes [12, 29]. Further, as alteration of a single transporter has been implicated in adverse human health outcomes [21], and ABCG2 protein expression has been shown to be increased following low dose short‐chain PFAS exposure in HepaRG [15], we aim to understand these changes and their implications for human health more broadly. We aim here to evaluate the ability of low dose short‐chain PFAS exposure to alter gene expression of 38 ABC transporters, and to determine how such changes impact the functional membrane transport of prototypical ABC and OATP substrates in a differentiated HepaRG liver model. The decision to profile 38 ABC transporters rather than a small subset was based on the hypothesis that short‐chain PFAS, through their surfactant‐like and lipid‐active nature, could modulate not only classical xenobiotic transporters but also lipid‐handling (ABCA), peroxisomal (ABCD), and mitochondrial‐associated (ABCB6, ABCB8) transporters. This approach enabled detection of broader adaptive or compensatory transcriptional responses that would not be captured by a narrow toxicant‐focused panel. This study was also designed to bridge molecular and functional endpoints in a human‐relevant liver model. Rather than assessing transporter‐specific binding, we aimed to determine whether environmentally relevant PFAS concentrations can globally modulate transporter networks. This assay utilizing both PFAS exposure and fluorescent substratesallows the direct observation of net uptake‐efflux dynamics in the same exposure context that drives gene expression changes.

2. Materials and Methods

2.1. Chemicals and Reagents

Perfluorobutanesulfonic acid (PFBS), perfluorohexanoic acid (PFHxA), 6:2 fluorotelomer alcohol (6:2 FTOH), perfluorohexanesulfonic acid (PFHxS), and dimethyl sulfoxide (DMSO) were obtained from Sigma‐Aldrich (St. Louis, MO). All PFAS chemical stock solutions were prepared in analytical grade DMSO. More details on the chemicals listed here are provided in Table 1. Williams E medium, trypsin‐EDTA, Penicillin/streptomycin, and phosphate‐buffered saline (PBS) were obtained from Life Technologies (ThermoFisher Scientific, Waltham, MA). Fetal bovine serum was obtained from Corning (Woodland, CA). Hydrocortisone hemisuccinate and human recombinant insulin were obtained from Sigma‐Aldrich.

Table 1.

Short‐chain PFAS used in this study and corresponding chemical and sourcing information.

Name	Abbreviation	CAS	Formula	MW
Perfluorobutanesulfonic acid	PFBS	375‐73‐5	C4HF9O3S	300.09
Perfluorohexanoic acid	PFHxA	307‐24‐4	C6HF11O2	314.05
Perfluorohexanesulfonic acid	PFHxS	3871‐996	C6HF13KO3S	438.20
6:2 Fluorotelomer alcohol	6:2 FTOH	647‐42‐7	C8H5F13O	364.10

Abbreviation: MW, Molecular weight in g/mol.

2.2. Cell Culture

For both expression and functional transport assays, HepaRG were cultured using 5% CO₂, 37°C, and Williams E medium supplemented with insulin, glutamax, hydrocortisone hemisuccinate, antibiotic, and FBS. Experiments were performed with 3–4 biological replicates per cell line and treatment. For all assays, cells were plated at the same density and differentiated to hepatocytes for 2 weeks in media containing 1.7% DMSO prior to experimentation [30].

2.3. Gene Expression Assays

Cells were plated and differentiated in Corning 6‐well plates as described above. Following differentiation, cells were treated with 1 µM PFBS, PFHxA, PFHxS, or 6:2 FTOH, or with 0.1% v/v DMSO in low serum media. All PFAS were dissolved in DMSO. Cells were then washed, pelleted, and RNA was extracted using Invitrogen PureLink RNA mini kit. RNA quality was assessed using a nanodrop instrument. cDNA was synthesized using Thermo RevertAid First Strand cDNA Synthesis kit, and TaqMan master mix and TaqMan ABC human array plates used for RT qPCR. ABC transporters that were included in the analysis are listed in Supplemental Table 1 along with selected substrates and alternate names. Plates were run on a Quant Studio 6 instrument, and CT values calculated in Quant Studio software. Beta‐glucuronidase (GUSB) housekeeping gene and DMSO‐treated controls were used to normalize all data.

2.4. Functional Transport of Prototypical Fluorescent Substrates

Established fluorescent substrates of ABC and OAT and known inhibitors of ABC and OATP were chosen in order to well represent substrates of as many ABCs as possible. Building upon protocols developed by Birsak et al. (2013) [31] and Grewal et al. (2017) [25], we aim to use differentiated HepaRG with endogenously expressed transporters in order to evaluate how relevant exposure to low dose short‐chain PFAS may impact the ability of the human liver to transport substrates. Such assays allow for evaluation of altered transport, as well as the implication of specific ABC or OATP involved in such functional changes.

Cells were grown in Corning 96‐well clear‐bottom plates and differentiated as described previously [30]. Cells were then treated with 1 nM or 1 µM of short‐chain PFAS in low serum media, or 0.1% v/v DMSO in low serum media. Fluorescent substrate and inhibitor concentrations as well as general methods were chosen based on those used by Birsak et al. (2013) [31]. Following 48 h, all cells were changed to media containing 1 nM or 1 µM PFAS, 2 µM inhibitor, or DMSO control plus 10 µM fluorescent substrate. Cells were allowed to uptake substrate for 45 min at 5% CO₂ and 37°C, followed by a wash with cold serum‐free media on ice to inhibit transport activity and remove residual substrate.

For ABC substrate assays, cells were then changed to media containing 1 nM or 1 µM PFAS, 0.1% v/v DMSO control, or 2 µM inhibitor in low‐serum media. Cells were allowed to efflux retained substrate for 45 min at 5% CO₂ 37°C. For fluorescein uptake by OATP, no efflux phase was performed. Plates were then washed with ice‐cold PBS and lysed using 50/50 MeOH/DI water (v/v). Plates were read using a Thermo Varioskan instrument at the appropriate excitation and emission for each substrate. Fluorescent substrates, known transporters, and sources are provided in Table 2, while inhibitors, their sources, and specific transporters known to be inhibited by them are provided in Table 3.

Table 2.

Fluorescent substrates used in the study.

Substrate	Transporters	References
BODIPY‐cholesterol (sterol C‐24)	ABCA1, ABCG1, ABCG5, ABCG8	[32, 33]
CMFDA	ABCB11, ABCC2, ABCC5	[22, 34]
Fluorescein	OATP1A2, OATP1B1, OATP1B3, OATP1C1, OATP2A1, OATP2B1, OATP3A1, OATP4A1, OATP4C1, OATP5A1, OATP6A1, ABCC1	[35, 36, 37]
Rhodamine 123	ABCB1	[25]
Hoechst 33342	ABCG2, ABCB1, ABCC1	[31]

Note: These fluorescent probes are widely used to study transporter activity. BODIPY‐cholesterol (sterol C‐24) is a substrate for cholesterol efflux transporters ABCA1, ABCG1, ABCG5, and ABCG8. CMFDA (5‐chloromethylfluorescein diacetate) is transported by ABCB11 and multidrug resistance‐associated proteins ABCC2 and ABCC5. Fluorescein is a probe for OATPsand ABCC1 (MRP1). Rhodamine 123 is a classical substrate for P‐glycoprotein (ABCB1). Hoechst 33342 is transported by breast cancer resistance protein (ABCG2), ABCB1, and ABCC1.

Table 3.

Inhibitors used for transporters studies.

Inhibitor	Transporters Inhibited	References
Cyclosporin A	OATP1B1, OATP1B3, ABCB1, ABCG2, ABCC1, ABCA1	[38, 39, 40, 41]
Nicardipine	ABCB1, ABCC1, ABCG2	[42]
Rifampicin	OATP1B1, OATP1B3, OATP2B1	[43, 44]
PSC833	ABCB1	[31]
Ko143	ABCG2, ABCB1	[31, 45]

Note: The ones used are standard tools for transporter studies. Cyclosporin A inhibits OATP1B1/1B3 and several ABC transporters, including P‐glycoprotein (ABCB1), breast cancer resistance protein (ABCG2), MRP1 (ABCC1), and ABCA1. Nicardipine blocks ABCB1, ABCC1, and ABCG2; rifampicin inhibits OATP1B/1B3 and OATP2B1; PSC833 (valspodar) selectively targets ABCB1; and Ko143 is a potent ABCG2 inhibitor with some activity against ABCB1.

The inclusion of PFAS during both the pre‐exposure and fluorescent substrate incubation phases was intentional to maintain steady‐state conditions reflective of sustained environmental exposure. This approach allowed evaluation of transport activity under the same exposure context in which gene expression and protein regulation were altered, ensuring consistency between molecular and functional endpoints. As our goal was to assess integrated hepatic transport responses rather than acute inhibition or competition at single transporters, PFAS were maintained in the assay media throughout both uptake and efflux phases. Similar co‐exposure designs have been applied in previous HepaRG and transporter functional studies [25, 31].

The 45 min and 48 h exposure durations were selected to represent acute and adaptive cellular responses in differentiated HepaRG hepatocytes. The short‐term (45 min) exposure reflects immediate physicochemical interactions of PFAS with membrane transporters and aligns with established transporter probe protocols that assess efflux and uptake kinetics within 30–60 min [22, 31]. In contrast, the long‐term (48 h) exposure allows sufficient time for transcriptional regulation, protein synthesis, and cellular adaptation processes to occur, consistent with the known kinetics of transporter induction and metabolic adjustment in HepaRG cells [18, 30].

2.5. Statistical Analysis

All functional transport experiments were performed using four biological and four technical replicates each, and technical replicates were averaged for each plate. Prior to statistical analysis, our data were examined to ensure compliance with homoscedasticity and normality assumptions required for parametric tests. Differences between concentrations were evaluated using one‐way ANOVA and Bonferroni′s test, with statistical significance determined via GraphPad Prism version 10.00 for MacOS (GraphPad Software, San Diego, CA). A p‐value of < 0.05 was considered statistically significant unless otherwise noted. When significant differences were identified, Dunnett′s multiple range test was conducted to compare the effects of PFAS.

3. Results

3.1. Short‐Term PFAS Exposure and ABC Transport

Results are presented graphically in Figure 1 and numerically in Table 4. Overall, short‐term exposure to per‐ and polyfluoroalkyl substances (PFAS) did not significantly alter the membrane transport activity of the prototypical fluorescent substrates Hoechst 33342, Rhodamine 123, BODIPY‐cholesterol, or CMFDA. The only significant effect observed was following exposure to 1 µM PFBS, which resulted in a decrease in retained fluorescence for BODIPY‐cholesterol, with a relative retention of 0.7418 compared to the respective DMSO control (0.7418 ± 0.092, p = 0.0353). No other short‐term PFAS treatments produced statistically significant changes in substrate retention. Positive controls cyclosporin A and nicardipine, known inhibitors of ATP‐binding cassette (ABC) transporters including ABCA1, ABCB1, and ABCB4 through direct interactions with ligand‐binding sites [38, 46], significantly increased retained fluorescence for CMFDA and BODIPY‐cholesterol.

Figure 1.

Impact of short‐chain PFAS exposure on cellular efflux of ABC fluorescent substrates after short‐term (45 min) and long‐term (48 h) exposure. Long term exposure groups were pre‐treated with indicated PFAS concentrations for 48 h prior to transport assay. All plates were then treated with PFAS or 2 µM inhibitor and 10 µM fluorescent substrate to allow for uptake, washed, and allowed to efflux. Plates were read at appropriate excitation/emission for each substrate. All experiments had four biological replicates with four technical replicates each. Data are presented as relative RFU normalized to respective DMSO controls (DMSO control considered as 1). One way ANOVA and Tukey′s test performed on all data. Asterisks represent significant differences between treatment groups and DMSO controls (p < 0.05).

Table 4.

Impact of short‐chain PFAS exposure on cellular efflux of ABC and OATP fluorescent substrates after long‐term (48 h) and short‐term (45 min) exposure.

		1 nM				1 µM
	Control	PFBS	PFHxA	PFHxS	6:2 FTOH	PFBS	PFHxA	PFHxS	6:2 FTOH	CsA	NIC
BODIPY‐cholesterol (45 min)	1.00 ± 0.05	0.93 ± 0.08	0.81 ± 0.08	0.87 ± 0.04	0.84 ± 0.07	0.74 ± 0.09	0.86 ± 0.17	0.91 ± 0.14	0.89 ± 0.15	0.94 ± 0.09	1.38 ± 0.08
BODIPY‐cholesterol (48 h)	1.00 ± 0.08	0.42 ± 0.02	0.57 ± 0.21	0.70 ± 0.17	0.56 ± 0.16	0.46 ± 0.14	0.34 ± 0.08	0.40 ± 0.10	0.47 ± 0.09	0.94 ± 0.09	1.38 ± 0.08
										CsA	NIC
CMFDA (45 min)	1.00 ± 0.03	0.95 ± 0.11	1.02 ± 0.04	1.03 ± 0.10	0.94 ± 0.02	0.87 ± 0.08	0.98 ± 0.06	1.00 ± 0.03	0.94 ± 0.06	1.76 ± 0.21	1.21 ± 0.08
CMFDA (48 h)	1.00 ± 0.02	0.82 ± 0.00	0.85 ± 0.04	0.76 ± 0.07	0.65 ± 0.13	0.69 ± 0.01	0.79 ± 0.11	0.71 ± 0.04	0.69 ± 0.10	1.76 ± 0.21	1.21 ± 0.08
										NIC	PSC
Rhodamine123 (45 min)	1.00 ± 0.11	0.89 ± 0.18	1.17 ± 0.25	1.26 ± 0.24	0.98 ± 0.15	0.97 ± 0.35	1.13 ± 0.12	1.06 ± 0.15	0.93 ± 0.93	1.34 ± 0.09	1.28 ± 0.12
Rhodamine123 (48 h)	1.00 ± 0.04	0.62 ± 0.08	0.74 ± 0.16	0.70 ± 0.14	0.62 ± 0.08	0.71 ± 0.09	0.75 ± 0.10	0.81 ± 0.16	0.76 ± 0.15	1.34 ± 0.09	1.28 ± 0.12
										NIC	Ko143
Hoechst33342 (45 min)	1.00 ± 0.09	0.76 ± 0.14	0.97 ± 0.27	1.13 ± 0.21	0.95 ± 0.01	0.85 ± 0.26	0.84 ± 0.26	0.95 ± 0.18	1.02 ± 0.17	1.36 ± 0.35	1.00 ± 0.33
Hoechst33342 (48 h)	1.00 ± 0.02	0.49 ± 0.01	0.64 ± 0.05	0.68 ± 0.15	0.68 ± 0.11	0.47 ± 0.04	0.63 ± 0.12	0.57 ± 0.06	0.66 ± 0.10	1.36 ± 0.35	1.00 ± 0.33
										CsA	RIF
Fluorescein (45 min)	1.00 ± 0.06	0.91 ± 0.04	0.89 ± 0.01	0.97 ± 0.08	0.92 ± 0.11	0.86 ± 0.06	0.98 ± 0.01	0.99 ± 0.14	0.93 ± 0.11	1.00 ± 0.11	1.03 ± 0.07
Fluorescein (48 h)	1.00 ± 0.09	0.78 ± 0.17	0.83 ± 0.01	0.80 ± 0.02	0.88 ± 0.15	0.65 ± 0.07	0.70 ± 0.14	0.65 ± 0.07	0.85 ± 0.03	1.00 ± 0.11	1.03 ± 0.07

Note: All experiments had four biological replicates with four technical replicates each. Data are presented as relative RFU normalized to respective Control (as DMSO control considered as 1) and as mean ± SD.

Abbreviations: CsA, Cyclosporine A; NIC, Nicardipine; PSC, PSC833; RIF, Rifampin.

3.2. Long‐Term PFAS Exposure and ABC Transport

Short‐term PFAS treatments did not significantly alter transporter activity; however, 48‐h pre‐treatment with both 1 nM and 1 µM exposures resulted in notable decreases in retained fluorescence across multiple substrates compared to DMSO controls. For Hoechst 33342, a substrate of ABCG2, ABCB1, and ABCC1, retained fluorescence was significantly reduced following 1 nM and 1 µM PFBS (0.4722 ± 0.0429, p = 0.0038) and 1 µM PFHxS (0.5695 ± 0.0610, p = 0.0324). Rhodamine 123, transported by ABCB1, also showed decreased retention after 1 nM PFBS (p = 0.0025) and 1 nM 6:2 FTOH (p = 0.0030). For CMFDA, a substrate of ABCB11, ABCC2, and ABCC5, significant reductions were observed with PFBS (1 µM, p = 0.0031), PFHxS (1 nM, p = 0.0351; 1 µM, p = 0.0068), and 6:2 FTOH (1 nM, p = 0.0005; 1 µM, p = 0.0030). The most pronounced effects were observed with BODIPY‐cholesterol, transported by ABCA1, ABCG1, ABCG5, and ABCG8. Significant decreases were detected following both 1 nM and 1 µM PFBS (p < 0.0001), PFHxA (1 nM, p = 0.0012; 1 µM, p < 0.0001), PFHxS (1 nM, p = 0.0477; 1 µM, p < 0.0001), and 6:2 FTOH (1 nM, p = 0.0008; 1 µM, p < 0.0001). All data can be found in Table 4.

Collectively, these data suggest that prolonged, low‐dose PFAS exposure impairs efflux of multiple prototypical substrates, with the strongest effects observed on C‐24 cholesterol transport.

3.3. Short‐Term PFAS Exposure and Fluorescein Transport

No significant differences in relative internalized fluorescence were observed between treatment groups following exposure to either 1 nM or 1 µM PFAS, nor with the inclusion of the known transporter inhibitors cyclosporin A and rifampicin. Full numerical values are provided in Table 4, while the corresponding graphical representations are shown in Figure 2 for clarity and comparison across treatments.

Figure 2.

Impact of short‐chain PFAS exposure on cellular efflux of OATP fluorescent substrate, fluorescein, after short‐term (45 min) and long‐term (48 h) exposure. Long term exposure groups were pre‐treated with indicated PFAS concentrations for 48 h prior to transport assay. All plates were then treated with PFAS or 2 µM inhibitor and 10 µM fluorescent substrate. Following uptake, plates were washed and read at appropriate excitation/emission. Fluorescein is a substrate of OATP and ABCC1. All experiments had four biological replicates with four technical replicates each. Data are presented as relative RFU normalized to respective DMSO controls (DMSO control considered as 1). One way ANOVA and Tukey′s test performed on all data. Asterisks represent significant differences between DMSO controls and treatment groups (p < 0.05).

3.4. Long‐Term PFAS Exposure and Fluorescein Transport

Results presented in Table 4 and Figure 2 show that 48‐h exposure to selected 1 µM concentrations of PFAS significantly altered relative fluorescence levels. Specifically, treatment with 1 µM PFBS (0.6466 ± 0.0719, p = 0.0012), 1 µM PFHxA (0.7027 ± 0.1399, p = 0.0096), and 1 µM PFHxS (0.6524 ± 0.0703, p = 0.0015) all resulted in decreased fluorescence compared to DMSO controls. These reductions indicate that prolonged PFAS exposure may alter both ABC and OATP activity, decreasing uptake by OATP and increasing efflux by ABC, ultimately causing the observed effects. These data using fluorescein, a well‐established OATP substrate, suggest that OATP‐mediated transport is sensitive to certain PFAS, particularly at higher concentrations and longer exposure durations.

3.5. Gene Expression

ABC transporter gene expression following 1 µM PFAS exposure is summarized in the heatmap of Figure 3, and data are provided in Table 5. Each of the four tested PFAS (PFBS, PFHxA, PFHxS, and 6:2 FTOH) produced broad transcriptional changes across multiple ABC subfamilies. Significant increases (indicated by asterisks on the heatmap) were observed in several members of the ABCA family (e.g., ABCA3, ABCA4, ABCA5, and ABCA7), which are involved in lipid transport and membrane homeostasis. Relative to DMSO control (solvent control), the log fold change of ABCA4 expression was significant for all PFAS treatments (7.778 log‐fold change for PFBS, 5.947 for PFHxA, 6.687 for PFHxS, and 6.953 for 6:2 FTOH). Similarly, multiple ABCB transporters, including ABCB1 and ABCB4, showed elevated expression, suggesting potential modulation of efflux pathways relevant to xenobiotic clearance. Notably, ABCC family members (MRPs) such as ABCC1, ABCC2, ABCC3, and ABCC4 were consistently upregulated across treatments (ABCC2 upregulated from 5.459 to 5.965 log‐fold change depending on treatment group; ABCC3 from 3.943 to 6.544 log‐fold change; and ABCC4 from 5.618 to 6.591 log‐fold change), highlighting their responsiveness to PFAS exposure and possible roles in mediating altered cellular transport dynamics. Within the ABCD family, ABCD1 and ABCD2 displayed strong induction, particularly under PFHxS and 6:2 FTOH exposure, pointing toward effects on peroxisomal transport processes. Among the ABCG subfamily, ABCG2 and ABCG5/8 also demonstrated significant expression changes, which may indicate disruption of cholesterol and sterol transport. The expression profile reveals that PFAS exposure induces a coordinated response across multiple ABC transporters, with both efflux (ABCBs, ABCCs, ABCGs) and lipid‐handling transporters (ABCAs) significantly affected.

Figure 3.

Heatmap of ABC transporter gene expression following PFAS exposure. PFBS, PFHxA, PFHxS, and 6:2 FTOH induced broad transcriptional changes across multiple ABC subfamilies, with significant upregulation observed in ABCA (lipid transport), ABCB and ABCC (efflux), and ABCD (peroxisomal) transporters. Data are presented as Log‐fold normalized expression values of ABC genes relative to DMSO controls. All data were normalized using GUSB housekeeping gene and 0.1% v/v DMSO treated controls. Asterisks indicate statistically significant differences (p < 0.05) relative to controls. Data provided in Table 5.

Table 5.

Data table of log2 fold expression of ABC genes following 1 µM 48 h PFAS exposures.

	PFBS			PFHxA			PFHxS			6:2 FTOH
	Log2 fold change	SD	p value	Log2 fold change	SD	p value	Log2 fold change	SD	p value	Log2 fold change	SD	p value
ABCA1	3.7475	0.717	0.031	5.136	0.897	0.022	6.4791	0.274	0.006	4.4484	0.867	0.028
ABCA12	2.8027	0.714	0.051	3.0312	0.612	0.038	3.7638	0.635	0.026	3.4191	0.918	0.048
ABCA2	0.9917	0.718	0.314	3.2425	1.336	0.097	6.6054	0.86	0.015	2.0217	0.377	0.079
ABCA3	1.8204	1.114	0.217	1.8166	0.632	0.149	3.7506	1.071	0.064	2.5097	1.257	0.151
ABCA4	7.7784	1.036	0.012	5.9475	0.755	0.013	6.6874	0.977	0.014	6.9539	1.961	0.041
ABCA5	6.1667	0.713	0.017	5.2097	0.606	0.021	5.7334	0.404	0.015	5.3923	0.728	0.023
ABCA6	3.9692	0.713	0.06	2.7922	0.668	0.109	3.2836	0.468	0.074	2.7942	0.974	0.131
ABCA7	2.7723	0.747	0.079	4.2331	0.785	0.038	5.3464	0.363	0.016	3.6098	0.312	0.033
ABCA8	6.7601	0.725	0.019	4.6895	1.88	0.094	5.3237	2.575	0.116	4.3056	2.499	0.156
ABCB1	6.4137	0.752	0.016	5.5468	0.608	0.019	5.5257	0.248	0.014	5.575	0.905	0.025
ABCB10	4.1722	0.715	0.027	2.726	0.919	0.078	3.5513	1.141	0.064	2.7749	1.525	0.144
ABCB4	4.9041	0.723	0.032	5.2292	1.67	0.064	5.2807	1.089	0.038	5.6524	0.499	0.02
ABCB6	6.7228	0.727	0.01	5.5812	0.63	0.012	6.3535	0.416	0.007	5.905	0.395	0.007
ABCB7	6.5248	0.738	0.011	5.1403	0.625	0.016	5.8979	0.732	0.014	5.2438	0.838	0.02
ABCB8	4.3691	0.714	0.022	4.9145	0.791	0.02	5.9838	0.415	0.007	4.209	0.168	0.011
ABCB9	4.1295	0.72	0.048	4.76	0.607	0.034	6.665	0.916	0.023	4.204	0.878	0.053
ABCC1	4.3422	0.713	0.022	4.8956	0.619	0.015	6.1055	0.659	0.011	5.3088	1.061	0.025
ABCC10	5.504	0.905	0.038	5.249	0.635	0.035	6.1215	0.448	0.024	4.9537	0.499	0.037
ABCC11	3.8673	1.081	0.048	3.2306	1.121	0.07	3.4457	0.424	0.022	2.9667	0.184	0.022
ABCC2	5.9655	0.713	0.012	5.4592	0.926	0.02	5.8745	0.4	0.008	5.5756	0.191	0.006
ABCC3	3.9432	0.715	0.028	4.6546	0.987	0.03	6.5445	0.301	0.006	4.8121	0.359	0.011
ABCC4	6.5915	0.735	0.022	5.619	0.681	0.029	6.2803	0.859	0.027	5.2662	1.517	0.062
ABCC5	5.4391	1.268	0.032	4.1994	1.804	0.089	5.1181	1.782	0.061	2.5947	3.219	0.379
ABCC6	0.5106	0.87	0.61	4.1245	1.307	0.064	5.3379	0.444	0.015	3.9178	0.162	0.023
ABCC9	7.406	0.732	0.011	5.9532	1.151	0.028	6.453	1.416	0.031	5.3489	2.021	0.075
ABCD1	3.8965	0.721	0.047	5.6333	0.62	0.021	7.804	2.188	0.044	5.5819	1.59	0.052
ABCD2	10.7799	0.772	0.005	9.04	2.232	0.032	9.1545	2.266	0.032	8.6105	3.722	0.085
ABCD3	8.373	1.048	0.012	5.9135	1.371	0.034	6.8803	1.976	0.045	6.3771	2.296	0.066
ABCD4	5.9359	0.714	0.012	4.9172	0.604	0.015	5.4285	0.531	0.011	5.0788	0.787	0.019
ABCE1	7.0858	0.713	0.009	4.7407	1.318	0.044	5.2913	1.62	0.05	5.1226	1.703	0.057
ABCF1	8.0807	0.717	0.007	5.9224	1.274	0.028	6.3392	1.58	0.034	6.3439	1.71	0.039
ABCF2	7.4034	0.713	0.01	5.3186	1.077	0.029	5.8664	1.41	0.035	5.7511	1.439	0.037
ABCG2	3.1768	0.726	0.046	2.8951	0.763	0.058	2.7137	0.384	0.04	2.7384	0.578	0.05
ABCG5	4.1941	0.854	0.03	5.8864	3.149	0.122	5.8589	2.425	0.08	5.5768	2.462	0.09
ABCG8	2.9194	0.741	0.072	4.1326	1.181	0.06	4.1574	0.879	0.044	3.9286	0.183	0.027
ABCB2	6.8436	0.713	0.009	5.2787	0.75	0.017	5.9761	0.975	0.018	5.3614	1.367	0.037
ABCB3	5.5207	0.8	0.017	5.2941	0.884	0.021	6.3418	1.295	0.025	4.8807	1.639	0.06

Note: All experiments had three or more replicates and was normalized to GUSB housekeeping gene. Data display log2 fold change relative to DMSO controls ± SD (Standard Deviation). p values are provided with significant values < 0.05 noted in bold.

4. Discussion

In this study, we demonstrate that long‐term exposure to short‐chain PFAS at concentrations consistent with levels reported in human biomonitoring studies significantly alters hepatic transport activity of prototypical substrates. Importantly, we show that 1 µM exposure to PFBS, PFHxA, PFHxS, and 6:2 FTOH markedly increases the expression of numerous ABC transporters in HepaRG, providing a potential mechanistic explanation for the enhanced efflux activity observed in our functional assays. To our knowledge, this is the first investigation to comprehensively evaluate PFAS effects across a broad panel of ABC transporters and the first to directly link these transcriptional changes to functional outcomes in liver transport assays. These findings highlight a previously unrecognized pathway by which environmentally relevant PFAS exposures may disrupt hepatic transport processes and contribute to altered xenobiotic disposition.

Our functional assays were designed to evaluate the integrated response of endogenously expressed hepatic transporters under chronic PFAS exposure, rather than to isolate direct substrate competition. Differentiated HepaRG cells were selected because they retain physiologic polarization and transporter localization [18], providing a more representative system than vesicular or overexpressing models, which lack endogenous regulatory networks and inter‐transporter crosstalk. The co‐exposure of PFAS and fluorescent probes was therefore essential to capture transporter function as it occurs under continuous exposure conditions typical of human physiology. HepaRG are well characterized for their use in ABC and OATP transporter research, and provide the best possible immortalized in vitro liver model short of primary cells [20]. In HepaRG cells, transporter localization and polarization remain stable for up to 2 weeks following differentiation [18, 19]. Short‐term exposures ( < 1 h) are therefore well suited to assess direct, non‐genomic effects such as substrate competition or immediate alterations to membrane potential or transporter conformation. In contrast, 48 h exposures permit sufficient time for PFAS‐induced transcriptional regulation, post‐translational modification, and protein trafficking, which are key determinants of transporter function. This timescale is consistent with previous hepatocyte studies examining xenobiotic‐mediated transporter modulation [25, 31].

Fluorescent substrates were selected based on their established transporter associations and physiological relevance. ABC substrates were chosen based on their uptake through diffusion and established efflux by specific ABC. Hoechst 33342, a cell‐permeant dye, has long been used to evaluate ABC transporter function, particularly ABCG2 and ABCB1 [31, 45, 47, 48, 49, 50, 51]. Rhodamine 123 enters cells by diffusion, accumulates in mitochondria, and is effluxed by ABCB1 [25, 52]. CMFDA is taken up through diffusion and hydrolyzed intracellularly into a fluorescent metabolite that is effluxed by ABCC2, ABCC5, and ABCB11 [34, 49, 53]. BODIPY‐cholesterol (sterol C‐24) was included due to the central role of ABC transporters in hepatic cholesterol regulation; BODIPY conjugation enables fluorescence detection without altering transport properties [31, 42, 54, 55]. Fluorescein was unique in being transported into the cell by OATPs rather than through diffusion, though ABCC1 has also been shown to transport it [35, 56]. Using these substrates, we evaluated both short‐ and long‐term PFAS exposures. As predicted, short‐term (45 min) treatment with either low (1 nM) or high (1 µM) doses of PFAS largely did not alter retained fluorescence relative to DMSO controls. In contrast, 48 h exposure produced more widespread and significant decreases in substrate retention, consistent with increased efflux or decreased uptake activity. These findings are in line with earlier work demonstrating increased ABCG2 protein expression after 48‐h PFAS exposure, and decreased OATP1B3 gene expression [10]. Positive control inhibitors provided additional context. Cyclosporin A, nicardipine, and PSC833 increased retained fluorescence for several substrates, confirming inhibition of specific ABC transporters, though not all treatments reached significance. Variability in inhibitor effects may be attributable in part to their mechanisms of action. For example, while both Rhodamine 123 and Hoechst 33342 are effluxed by ABCB1, they bind distinct sites [42], meaning inhibition at one site may not affect transport of the other substrate. Prior studies of these inhibitors have typically used transfected cells or inverted vesicles [17, 31], whereas the present study provides a more integrated view of inhibitor and PFAS effects in a liver cell model.

We also evaluated OATP‐mediated transport using fluorescein as a model substrate. OATPs are known to facilitate the transport of several PFAS, including PFBS, PFHxS, and PFOS [17, 57], and have been implicated in sex‐specific differences in PFAS elimination [58]. Similar to the trends observed for ABC substrates, short‐term PFAS exposure produced minimal changes in fluorescein retention. In contrast, 48 h exposure to 1 µM PFBS, PFHxA, and PFHxS significantly reduced intracellular fluorescein relative to DMSO controls, suggesting decreased uptake activity by OATP transporters. However, contributions from ABCC1‐mediated efflux, which would likewise decrease retained fluorescein, cannot be excluded. It is important to note that these results do not necessarily indicate reduced activity or expression of OATP. While this work indicates increased ABC expression and efflux activity following PFAS exposure, and there is evidence for decreased OATP1B3 expression following PFAS exposure in a liver model [10], the ability of all 11 human OATP to transport fluorescein with variable affinities [36] and the lack of other OATP‐PFAS gene expression information, not to mention the possibility that ABCC1 or other ABC may play a larger role in fluorescein transport than currently known, together prevents a clear designation of OATP activity as solely responsible for these effects. However as 11 OATP uptake fluorescein, and only 1 ABC is known to efflux fluorescein, we believe the decreased retained fluorescence provides unique information suggestive of PFAS OATP activity impacts which are worthy or further inquiry.

Overall, our 48‐h treatment results for both ABC and OATP transporters indicate that PFAS exposure, particularly at 1 µM concentrations, enhances substrate efflux, while reduced fluorescein accumulation suggests impaired uptake may also play a role in the transport trends seen here. Taken together, these findings raise the possibility that with both uptake and efflux altered in hepatocytes, net transport trends may appear balanced; however, the underlying mechanisms are likely more complex. Importantly, the differences in substrate specificity, transporter localization, and physiological function between ABCs and OATPs should be considered when interpreting these results. OATPs primarily mediate uptake of steroids, drugs, and bile acids, whereas ABCs transport lipids and a broad range of pharmaceuticals [59]. This distinction implies that steroid hormone uptake could be impaired by OATP inhibition regardless of concurrent ABC activity. For bile salts and bile acids, which depend on coordinated uptake by OATPs and efflux by ABC transporters, a decrease in sinusoidal uptake through OATPs together with enhanced canalicular efflux via ABCs could collectively lower intracellular bile acid concentrations and ultimately reduce biliary excretion. Similarly, for dual‐handled substrates of transporters such as ABCA1, ABCC1, and ABCC3, concurrent reductions in uptake and increases in efflux across the hepatocyte‐sinusoid interface would be expected to minimize intracellular accumulation. In line with this interpretation, fluorescein, transported primarily by OATP1B1, OATP1B3, OATP2B1, and ABCC1, showed markedly reduced retention following 48 h exposures to 1 µM PFBS, PFHxA, and PFHxS, each producing decreases of 30% or greater relative to controls.

Another critical factor is the limited knowledge of transporter substrate specificity. While the substrates used in this study (summarized in Table 2) are established probes, the list is likely not exhaustive. The fluorescent substrates and inhibitors applied here are not strictly specific to single transporters. For instance, although ABCA1 and ABCA5 both transport cholesterol [59], only ABCA1 has been confirmed to transport BODIPY‐cholesterol [32]. Likewise, CMFDA, often used to quantify ABCB11 activity, is also transported by ABCC2 [22] and ABCC5 [53], limiting its use for isolating individual transporter activity. This complexity arises from the challenge of screening modulators against 48 distinct human ABC transporters, each with unique but sometimes overlapping substrate‐binding properties [60]. Thus, the strength of our approach lies in using functional transport assays in differentiated HepaRG hepatocytes, capturing integrated and physiologically relevant outcomes rather than focusing on isolated transporters in overexpression models. We believe this also impacted the inhibitors used in our transport assays, as the transporters which move specific substrates are largely not completely aligned with the transporters which are inhibited by a specific chemical. For example, fluorescein is transported by ABCC1 and all 11 human OATP, however common OATP inhibitors cyclosporin A and rifampicin only collectively inhibit the activity of OATP1B1, OATP1B3, and OATP2B1, potentially allowing ABCC1 and the other 8 OATP to continue to transport fluorescein unimpeded.

Among the findings, the broad upregulation of ABC transporters and the significant increases in BODIPY‐cholesterol efflux stand out as most relevant to human health. Lipids, including cholesterol, are known substrates of 18 human ABC transporters [59], and our results suggest that PFAS exposure may disrupt hepatic lipid transport pathways. Notably, significant alterations in transport activity were observed even at 1 nM PFAS, concentrations consistent with those detected in the U.S. population. This suggests that environmentally relevant PFAS exposures could influence hepatic handling of endogenous substrates such as cholesterol and bile acids, as well as commonly prescribed drugs known to interact with ABCs, including ibuprofen [61], levothyroxine, and statins [59]. Such changes could also modify susceptibility to drug–drug interactions or drug‐induced liver injury [21]. Generally, increased ABC expression and activity are associated with drug resistance in cancer, immunosuppression, and epilepsy treatments [13, 25], whereas reduced or dysfunctional ABC expression contributes to lipid accumulation disorders such as Tangier disease, Stargardt disease, and familial intrahepatic cholestasis [13, 62, 63, 64]. Beyond these clinical contexts, transporter variability contributes to demographic differences in drug outcomes: increased adverse drug reactions and higher rates of drug‐induced liver injury in women, differences in efficacy and toxicity among racial and ethnic groups, and reduced drug metabolism in the elderly [26]. Together, our findings highlight that short‐chain PFAS exposure disrupts transporter networks in ways that extend beyond simple uptake and efflux balance, with implications for lipid metabolism, pharmacokinetics, and interindividual variability in drug response.

5. Conclusion

This study provides critical insights at the interface of environmental toxicology and human pharmacology by demonstrating that common short‐chain PFAS, including PFBS, PFHxA, PFHxS, and 6:2 FTOH, can alter hepatic membrane transport activity at concentrations relevant to the U.S. population. Using a functional assay encompassing a broad panel of ABC and OATP transporters, we establish a human‐relevant model that not only captures transporter‐specific responses but also provides a framework for screening other environmental contaminants or pharmacological agents. Our findings show that prolonged PFAS exposure enhances efflux and may reduce uptake of prototypical substrates, and at higher concentrations, significantly upregulates ABC transporter gene expression. These results underscore the potential for PFAS to disrupt endogenous substrate handling and drug disposition, with implications for liver function and population‐level variability in pharmacological outcomes. Comprehensive integration of human serum PFAS measurements, environmental exposure data, transporter expression profiles, and identification of susceptible populations will be essential to predict drug responses more accurately, mitigate adverse outcomes, and guide regulatory and therapeutic strategies.

Although the present study did not employ transporter‐overexpressing systems or PFAS pre‐washout designs, these approaches will be valuable in future work to dissect direct binding or inhibition kinetics at individual transporters. The current experimental framework instead emphasizes the systemic and transcriptionally coupled responses of transporters within a physiologically relevant hepatic model, offering an integrated view of how sustained PFAS exposure may modulate overall drug disposition in vivo.

Author Contributions

Gracen E. Collier: conceptualization, formal analysis, investigation, methodology, validation, visualization, writing – original draft. Ramon Lavado: conceptualization, data curation, formal analysis, funding acquisition, project administration, supervision, validation, visualization, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

JBMT 3361030.R1_SI.

Collier G. E., and Lavado R., “Functional and Transcriptional Disruption of Hepatic ABC Transporters by Environmentally Relevant Short‐Chain Per‐ and Polyfluoroalkyl Substances (PFAS),” Journal of Biochemical and Molecular Toxicology 39 (2025): e70659. 10.1002/jbt.70659.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon request.