Chronic PFOS exposures induce life stage–specific behavioral deficits in adult zebrafish and produce malformation and behavioral deficits in F1 offspring



Perfluorooctane sulfonic acid (PFOS) is an organic contaminant that is ubiquitous in the environment. Few studies have assessed the behavioral effects of chronic PFOS exposure in aquatic organisms. The present study defined the behavioral effects of varying life span chronic exposures to PFOS in zebrafish. Specifically, zebrafish were exposed to control or 0.5 µM PFOS during 1 to 20, 21 to 120, or 1 to 120 d postfertilization (dpf). Exposure to PFOS impaired the adult zebrafish behavior mode under the tapping stimulus. The movement speed of male and female fish exposed for 1 to 120 dpf was significantly increased compared with control before and after tapping, whereas in the groups exposed for 1 to 20 and 21 to 120 dpf, only the males exhibited elevated swim speed before tapping. Residues of PFOS in F1 embryos derived from parental exposure for 1 to 120 and 21 to 120 dpf were significantly higher than control, and F1 embryos in these two groups also showed high malformation and mortality. The F1 larvae of parental fish exposed to PFOS for 1 to 20 or 21 to 120 dpf exhibited a higher swimming speed than control larvae in a light-to-dark behavior assessment test. The F1 larvae derived from parental fish exposed to PFOS for 1 to 120 dpf showed a significantly lower speed in the light period and a higher speed in the dark period compared with controls. Although there was little PFOS residue in embryos derived from the 1- to 20-dpf parental PFOS-exposed group, the adverse behavioral effects on both adult and F1 larvae indicate that exposure during the first 21 dpf induces long-term neurobehaviorial toxicity. The authors' findings demonstrate that chronic PFOS exposure during different life stages adversely affects adult behavior and F1 offspring morphology, behavior, and survival. Environ. Toxicol. Chem. 2013;32:201–206. © 2012 SETAC


Fluorinated organic compounds such as perfluorooctane sulfonic acid (PFOS) have been manufactured for over 50 years and are used as components of fire retardants, lubricants, adhesives, paper coatings, pharmaceuticals, cosmetics, and insecticides 1–3. As a pollutant in the global ecosystem, the median value of PFOS in the serum of non-occupationally exposed humans is 18.8 ng/L, with a range of 8.1 to 150.7 ng/L 4. Although PFOS is generally found at low levels over a range of 0.1 to 100 ng/L in surface water, concentrations of up to 600 ng/L have been reported in downstream rivers of fluorochemical manufacturing facilities 5. Because PFOS is characterized by high bioaccumulation and negligible elimination, high concentrations of PFOS are detected in a variety of fish species in the aquatic environment. For example, PFOS concentrations in the liver of smallmouth bass (Micropterus dolomieu) and largemouth bass (Micropterus salmoides) from New York, USA, are present from 9 to 315 ng/g wet weight, and the average concentrations of PFOS in the fish are 8,850 times greater than those found in the surface water 6. To simulate this high bioaccumulation effect with a limited exposure period in experimental settings, we employed a dose range that is higher than the environmental background level for testing.

Perfluorooctane sulfonic acid has been shown to cause developmental toxicity, reproductive toxicity, hepatotoxicity, immunotoxicity, endocrine disruption, and neurotoxicity in mammalian and aquatic fish model species 7–11. In particular, the role of PFOS in the developing nervous system has received increasing attention 12, 13. For example, in vitro PFOS treatment can induce PC12 cells to differentiate into the acetylcholine neurotransmitter phenotype at the expense of the dopamine phenotype 14, 15. Treatment with PFOS has also been found to cause backward swimming and prolong the duration of high K+-induced backward swimming among paramecia 16.

The zebrafish is a good model for studying neurotoxicity using behavioral endpoints due to their predictable swimming habits in both larvae and adult fish 17. Our group has previously investigated behavioral changes in zebrafish larvae under acute PFOS exposure 18 or in F1 larvae under chronic whole-life–stage PFOS exposure 18, 19. One of our significant findings is that chronic PFOS exposure of parental fish during whole-life stage can induce behavioral deficits and primary motor neuron and muscle developmental malformations in F1 embryos. To further uncover the window of susceptibility of long-term effects of chronic PFOS exposure, we selected three time periods that cover different developmental life stages: 1 to 20 d postfertilization (dpf), 21 to 120 dpf, and 1 to 120 dpf. The 1- to 20-dpf period represents an early life stage from embryonic development to juvenile fish, which have acquired most adult characteristics along with the absence of sexual maturity 20. The 21- to 120-dpf life stage encompasses the sex differentiation period 21 and adulthood. The 1- to 120-dpf period represents the entire life span from embryonic development to adulthood. Our goal was to identify which developmental period is the most sensitive to PFOS-induced behavioral defects in adult zebrafish and their F1 offspring.


Fish husbandry and embryo collection

Adult zebrafish (Danio rerio) of the US-AB strain were raised at the Sinnhuber Aquatic Research Laboratory at Oregon State University, USA, under standard laboratory conditions of temperature 28 ± 0.5°C, pH 7.2 ± 0.2, and a 14:10 dark/light photoperiod, according to standard zebrafish breeding protocols 22. Water supplied to the system was filtered by reverse osmosis (pH 7.0–7.5), and Instant Ocean salt was added to the water to raise the conductivity to 450 to 1,000 µS/cm (system water). The adult fish were fed twice daily with live Artemia and dry flake food (Zeigler, Aquatic Habitats).

Zebrafish embryos were obtained from spawning adults in tanks with a sex ratio of 1:1. Embryos were collected within 1 h after spawning and rinsed in embryo medium 22. The fertilized embryos were inspected and staged using a stereomicroscope (Nikon) according to the procedure of Kimmel et al. 23.

PFOS stock solutions and chronic exposure protocol

The PFOS (CAS 1763-23-1, purity > 96%) was purchased from Sigma–Aldrich, and stock solutions (0.5 mM PFOS) were prepared by dissolving it in 100% dimethyl sulfoxide. High-quality, 8-h postfertilization embryos were divided into four treatment groups: vehicle control with 0.01% dimethyl sulfoxide (v/v, in fish water) and 0.5 µM PFOS-exposed groups at three different life stages, 1 to 20, 21 to 120, and 1 to 120 dpf. Starting at day 1, embryos were raised in a Petri dish (100 embryos/treatment, 150 ml/dish) for 4 d with daily media change; all embryos hatched and survived at this stage. At 5 dpf, the fish were transferred to 3.75-L stainless steel tanks for a period of 5 to 20 dpf with the water changed once every 3 d. At 21 dpf, 30 larvae in each group were separated into three replicate tanks with a total of 10 fish per tank. Exposure solutions were renewed (95% of volume) every 3 d. Each tank was checked daily for morbid fish and on each water change day for water quality, including pH, ammonia, and nitrite levels. Feeding was initiated on day 5. Between 5 and 14 dpf, fish were fed three times per day with zebrafish larval diet (Aquatic Habitats). During 15 to 120 dpf, they were fed on freshly hatched live Artemia three times per day during 15 to 95 dpf and two times per day after 96 dpf.

Adult behavior of F0 zebrafish exposed to PFOS

To evaluate sensory responses, computer-assisted video monitoring of swimming behavior was assessed by modifying the method of Eddins et al. 24. Adult behavior tests were carried out with males and females separated into two different test groups according to size and body type differences. For each test, individual fish (n = 12) from each treatment group were placed in individual 1.75-L tanks containing approximately 1.5 L of fish water (one fish per tank). Tanks were set inline on shelves with the broadside facing the camera and were separated by dividers. Tanks were backed with blank white paper and were evenly backlit. The room temperature was controlled at 28°C. Trials were recorded using a Sony HD camcorder (Sony Handycam HDR-SR11). The position of each fish in the tank was recorded every 0.2 s, from which distance moved was calculated, which was later averaged for each 30-s interval. Analysis of the recorded tracks was performed with Noldus Etho-Vision XT V 7.0 software. Startle stimulus to each fish was generated using an electromagnetic solenoid attached to the bottom of each tank, which was controlled by a manual switch to hit each tank simultaneously. Fish were fasted for the duration of the behavior trials and were given 2 h to acclimate prior to the test. Fish activity was recorded for a total of 30 min, of which the last 16 min were analyzed including 12 min of background activity followed by 4-min stimulated activity after a startle stimulus (tapping). Average movement speeds in the first 12.5 min before tapping and the first 4 min after tapping were used to evaluate the fish movement mode under tapping stimulus.

PFOS residues and developmental evaluation of F1 embryos and larvae

When chronic PFOS-exposed zebrafish began to spawn eggs (at approximately three months of age), breeding trials with a sex ratio of 1:1 were carried out to produce F1 offspring. The F1 embryos were monitored for developmental progression, and hatched larvae were assessed for malformations and mortality until 8 dpf, with feeding starting from 5 dpf in the absence of PFOS. Embryo monitoring was carried out in six-well plates containing embryo media (5 ml/well). The experiment was repeated five times using 20 embryos per group during each replication. The PFOS concentrations in whole body tissue of F1 embryos from separate trials were determined with ACQUITY ultra performance liquid chromatography combined Micromass Quattro Premier XE (Waters) based on a protocol described by So et al. 25. Samples were prepared using method described elsewhere 18. A total of 40 embryos were pooled as a single sample, and measurements were replicated four times for each group.

Larval behavior of F1 offspring derived from parents exposed to PFOS

Embryos derived from PFOS chronic exposed parents were allowed to develop in the absence of PFOS, and surviving larvae at 4 dpf with normal morphology were further subjected to a behavioral assessment in response to alternating light and dark stimulus as described earlier 26. The test was performed in 24-well plates with each plate containing six fish per treatment group plus controls inside a ZebraLab behavior monitoring station (ViewPoint Life Sciences). Briefly, the larvae were allowed to acclimatize for 20 min in the testing chamber before the test began. The lighting parameters used were alternating 10 min of light (visible light) and 10 min of dark (infrared light), with intervals repeated until 50 min of total elapsed time. Average activity of a total of 18 fish for each treatment group was calculated.

Data analysis

One-way analysis of variance was followed by least significant difference as a post hoc test to determine statistical significance among different treatment groups. All values are reported as means ± standard error. Statistical analysis was performed using SPSS 16.0 software, and α = 0.05 was set as the significance level.


Chronic PFOS exposure alters F0 adult behavior

Exposure water quality, including pH, ammonia, and nitrite, was measured throughout the whole study. The pH was maintained between 6.8 and 7.6. The ammonia concentration was between 0 and 2 mg/L. Before 60 dpf, the nitrite concentration was low, 0.073 ± 0.116 mg/L. After 60 dpf, its concentration increased and averaged 1.02 ± 0.651 mg/L. Chronic PFOS exposure impaired the adult zebrafish behavior mode as measured by the average distance moved in response to the tapping stimulus (Fig. 1). The movement speed of fish exposed to PFOS at the period of 1 to 120 dpf was significantly increased compared with control, whereas exposure groups of 1 to 20 and 21 to 120 dpf were not severely affected (Fig. 1A and B). After tapping, the movement speed was decreased for all groups. Fish in the 1- to 120-dpf group also had significantly (p < 0.05) greater movements compared with the other three groups (Fig. 1C and D). We also observed sex-specific effects between male and female fish. For example, female fish from the 1- to 120-dpf group appear to be more hyperactive than males either before or after tapping. Compared with control, swim speeds before tapping in male fish from the 1- to 20-dpf group were higher (p = 0.0001), although no differences were observed after tapping (p = 0.27; Fig. 1D).

Figure 1.

Behavior analysis of four-month-old adult zebrafish after chronic exposure to 0.5 µM perfluorooctane sulfonic acid (PFOS) during different life spans. Female and male fish were subjected to behavior test separately (n = 12). After subtracting the time for a rest period, fish activities were recorded continuously for 12 min (only 6 min are displayed) before the tap and 4 min after the tap (A and B, respectively). The average distance moved within the 12.5 min before and 4 min after tapping were recorded for both females (C) and males (D). Bars of the same pattern sharing the same letter indicate no significant difference. Error bars represent standard error.

Chronic exposure to PFOS affects PFOS residues in F1 embryos

Perfluorooctane sulfonic acid residues in F1 embryos derived from parental fish exposed to PFOS in the 21- to 120-dpf period were not different from those exposed in the 1- to 120-dpf period (p = 0.509; Fig. 2), but values in these two groups were significantly higher than those detected in the control or 1- to 20-dpf PFOS-treated group (p = 0.000; Fig. 2).

Figure 2.

Perfluorooctane sulfonic acid (PFOS) residues in F1 offspring derived from PFOS-exposed F0 parental zebrafish (n = 4 with 40 embryos per replicate). Bars sharing the same letter indicate no significant difference. Error bars represent standard error.

Chronic PFOS exposure affects F1 embryo development

Compared with control and fish exposed to PFOS at 1 to 20 dpf, fish exposed to PFOS in the 21- to 120- and 1- to 120-dpf groups produced a significantly higher percentage of malformation and mortality in F1 larvae. Percentage of malformation increased from less than 40% at 6 dpf to more than 90% at 8 dpf (Fig. 3A). Malformed larvae were characterized by uninflated swim bladders and bent spines (Fig. 3A). Percentage of mortality increased from less than 2% at 6 dpf to 70 and 93% at 8 dpf for the 21- to 120- and 1- to 120-dpf groups, respectively (Fig. 3B).

Figure 3.

Parental perfluorooctane sulfonic acid (PFOS) exposure-induced malformation (A) and mortality (B) for F1 embryos at 6, 7, and 8 d postfertilization (dpf). At the top are representative images of 6-dpf larvae showing normal morphology in control and 1- to 20-dpf exposed groups, as well as bent spine (*) and uninflated swim bladder (arrows) in 21- to 120- and 1- to 120-dpf exposed groups. Error bars represent standard error. [Color figure can be seen in the online version of this article, available at]

Chronic PFOS exposure affects the behavior of F1 larvae

Larvae at 4 dpf with normal morphology derived from all four treatment groups were subjected to the light-to-dark stimulation test. A rapid transition from light to dark resulted in a similar, brief burst of swimming among fish in all groups (Fig. 4A). Larvae from parents with PFOS exposure at 1 to 20 and 21 to 120 dpf elicited a higher basal swim rate (p < 0.05) than those from the control group in both the light and dark periods (Fig. 4B). Larvae derived from parents with PFOS exposure at 1 to 120 dpf showed lower swim speed in the light (p = 0.001) period and higher swim speed in the dark period (p = 0.028) compared with larvae from the control group (Fig. 4B). Overall, the average swim speed displayed an inverted U-shape among the control, 1- to 20-, 21- to 120-, and 1- to 120-dpf groups (Fig. 4B).

Figure 4.

The locomotion activity of F1 larvae at 4 d postfertilization (dpf) when subjected to a 50-min dark/light photoperiod stimulation was tested. Dots represent the mean speed at 60-s intervals of 10 larvae (A). Treatments refer to parental 0.5-µM perfluorooctane sulfonic acid (PFOS) exposure period. Average speed in light or dark period is shown (B). Bars of the same pattern sharing the same letter indicate no significant difference. Error bars represent standard error.


In the present study, we report for the first time that chronic exposure to PFOS during different life stage results in behavioral anomalies in adult zebrafish and their offspring. Additionally, compared with the control and 1- to 20-dpf groups, the F1 embryos derived from parents with PFOS exposure at 21 to 120 and 1 to 120 dpf developed almost 100% abnormality and high mortality at 8 dpf, which is possibly related to the high PFOS residues from the maternal transfer of PFOS to eggs. This finding is consistent with our previous study showing that decreased larval survival in F1 offspring was directly correlated with the PFOS body burden, and larval lethality and abnormality were due to maternal transfer of PFOS to the eggs 19.

Previously, prenatal or neonatal exposure to PFOS at high doses (21 µmol/kg/d) in rats or mice was shown to lead to abnormal spontaneous behavior, increased motor activities, and reduced habituation 9, 12, 13. In contrast, adult animals exposed to PFOS have shown no or only slight neurobehavioral effects 27, 28. In the present study, we did not observe a more sensitive response of adult fish with PFOS exposure during an early life stage (1 to 20 dpf) than those exposed at a later life stage (21 to 120 dpf), although there was a subtle difference in behavior of the 4-dpf larvae between these two groups. More pronounced differences between control and treatment groups were found mainly at the exposure periods of 21 to 120 and 1 to 120 dpf, indicating that the adverse behavior response mainly resulted from prolonged PFOS exposure or, more accurately, higher PFOS accumulation in body tissue, as evidenced by the high PFOS residues detected in embryos derived from these two groups. We suspect there is a threshold of PFOS concentration in body tissue that must be reached before organisms will exhibit PFOS-specific responses, and exposure during an early life stage in the present study with 0.5 µM may not be sufficient for juvenile fish to reach such a threshold. This may explain why early life stage exposure was not more toxic than later life stage exposure in the present study with chronic PFOS exposure; the opposite effect was found in rodents with high-dose acute PFOS exposure. Alternatively, age-specific biological processes may contribute to different sensitivities among various exposure windows. The fact that male and female fish exhibited sex-specific behavioral responses upon PFOS exposure further supports this alternative hypothesis.

Zebrafish have a biorhythm; larvae become active after exposure to sudden darkness and then slow down 29, 30. The 4-dpf F1 offspring derived from PFOS-exposed parents swam at a faster and then a slower speed among the 1- to 20-, 21- to 120-, and 1- to 120-dpf groups but elicited a higher basal activity in response to light-to-dark photoperiod stimulation. This trend is consistent with the PFOS residue burden: a lower dose of PFOS increases the speed and a higher dose decreases the speed, as found in a previous study 18. The mechanisms at the physiological or biochemical levels underlying locomotion behaviors in response to light-to-dark stimulation are still not completely understood. Clearly there is a significant involvement of motor neurons and skeletal muscle in overall locomotor behavior 31, 32. Histological examination of acute PFOS-exposed larvae showed disordered and loosened arrays in the muscle fibers 18. Parental PFOS exposure led to disordered and loosened arrays in the slow muscle fibers of F1 larvae, and the primary motor neurons also exhibited slower development 19. Thus, it is possible that behavioral changes in the F1 larvae may be due to direct impacts on vasculogenesis, or angiogenesis, or both within the trunk. The next step should be to examine the molecular events underlying the observed changes in behavior.

Although early-life PFOS exposure did not produce more pronounced changes in adult or F1 larvae behavior than in other PFOS-treated groups, it did induce behavioral changes that were different from the control fish. For example, the male adults from the 1- to 20-dpf PFOS-exposed group responded to tapping stimulus differently than the control adult males. Also, the behavior of F1 larvae was significantly affected in this early life stage PFOS-exposed group even though that these larvae were morphologically indistinguishably from the control group. These observations indicate that PFOS exposure during the first 21 d can induce long-term adverse behavioral effects on adults and F1 embryos. This is consistent with recent studies showing that prenatal or neonatal exposure to PFOS correlated with behavioral anomalies in rats or mice as well as delayed neuromotor maturation 9, 13, and these behavioral modifications appear to persist into adulthood 12.

It has long been known that PFOS acts as an endocrine disrupter in fish and mammals. In rats, PFOS was found to perturb the neuroendocrine system by increasing serum corticosterone levels as well as norepinephrine concentrations in the paraventricular nucleus of the hypothalamus 11. In zebrafish, PFOS exposure disrupted thyroid hormone levels by altering gene expression in the hypothalamic–pituitary–thyroid axis 33. Therefore, the adult behavior changes found in the present study may be due to hypothalamus-regulated changes in hormone levels following long-term exposure to PFOS.

In conclusion, our findings demonstrate that chronic exposure to PFOS at different life stages adversely impacts adult behavior, subsequent offspring malformation, and larval behavior.


The present study was supported in part by funding from the National Natural Science Foundation of China (grants 20977068 and 21277104), the National Environmental Protection Public Welfare Science and Technology Research Program of China (grant 200909089), the Research Program of the Department of Education of Zhejiang Province (grant Y201010056), the International Collaboration Project of the Wenzhou City Government (grant H20100062), the Natural Science Foundation of Zhejiang Province (grant Y2110659), and the U.S. National Institute of Environmental Health Sciences (grant P30 ES00210).