ToxSci Advance Access originally published online on June 16, 2008
Toxicological Sciences 2008 105(1):86-96; doi:10.1093/toxsci/kfn113
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Evaluation of the Immune System in Rats and Mice Administered Linear Ammonium Perfluorooctanoate
DuPont Haskell Globel Centers for Health and Environmental Sciences, Newark, Delaware 19714
1 To whom correspondence should be addressed at DuPont Haskell Laboratory, P.O. Box 50, Elkton Rd, Newark, DE 19714-0050. Fax: (302) 366-6420. E-mail: scott.e.loveless{at}usa.dupont.com.
Received February 22, 2008; accepted June 2, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Repeated high doses of ammonium perfluorooctanoate (APFO) have been reported to affect immune system function in mice. To examine dose-response characteristics in both rats and mice, male CD rats and CD-1 mice were dosed by oral gavage with 0.3–30 mg/kg/day of linear APFO for 29 days. Anti-sheep red blood cell (SRBC) IgM levels, clinical signs, body weights, selected hematology, and lipid parameters, liver weights, spleen, and thymus weights and cell number, selected histopathology, and serum corticosterone concentrations were evaluated. In rats, linear APFO had no effect on production of anti-SRBC antibodies. Ten and 30 mg/kg/day resulted in systemic toxicity as evidenced by decreases in body weight gain to 74 and 37%, and increases in serum corticosterone levels to 135 and 196% of control, respectively. In mice dosed with 10 and 30 mg/kg/day, marked systemic toxicity and stress were observed, as evidenced by a loss in body weight of 3.8 and 6.6 g, respectively (despite a tripling of liver weight),
230% increase in serum corticosterone, and increases in absolute numbers of peripheral blood neutrophils and monocytes with an accompanying decrease in absolute lymphocyte numbers. Immune-related findings at 10 and 30 mg/kg/day that likely represent secondary responses to the systemic toxicity and stress observed at these doses include: decreased IgM antibody production at 10 (20% suppression) and 30 mg/kg/day (28% suppression); decreased spleen and thymus weights and cell numbers; microscopic depletion/atrophy of lymphoid tissue at 10 (thymus) and 30 mg/kg/day (spleen). In summary, no immune-related changes occurred in rats, even at doses causing systemic toxicity. In mice, immune-related changes occurred only at doses causing significant and profound systemic toxicity and stress. Key Words: APFO; PFOA; immunotoxicity; immunomodulation.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ammonium perfluorooctanoate (APFO; FC-143, C8, C7F15COO–NH4+, CAS Registry No. 3825-26-1) is a surfactant, manufactured by electrochemical fluorination, used as a processing aid in the production of fluoropolymers. Perfluorooctanoate (PFOA; C7F15COO–), the dissociation product of APFO, is not metabolized in rats or people (Vanden Heuvel et al., 1991
). PFOA is biopersistent and has been identified in (1) blood samples from exposed workers, the general population, various forms of wildlife, and (2) environmental media (Hansen et al., 2001; Lau et al., 2007; Olsen et al., 2003a, b; Taves et al., 1976).
Relatively few studies have examined effects on the immune system after exposure to APFO or PFOA. When spleen or thymus weights were evaluated in rodents following subchronic or chronic dietary exposure, minimal to no effect has been reported (Butenhoff et al., 2004; International Research and Development Corporation, 1978; Riker Laboratories, 1987).
In rhesus monkeys, APFO administration did not affect relative spleen weights or lymphoid or thymus histopathology at any dose up to 30 mg/kg/day in a 90-day toxicity study (Griffith and Long, 1980). In cynomolgus monkeys administered 0, 2, or 20 mg/kg/day APFO for 4 weeks or 0, 3, 10, or 20/30 mg/kg/day APFO for 6 months, nothing remarkable was observed in microscopic examinations of spleen, thymus, and mesenteric lymph nodes at any dose, including doses that resulted in significant increases in liver weights and peroxisome proliferation (Butenhoff et al., 2002).
In a series of papers by Yang et al. (2000), male C57BL/6 mice were fed PFOA at varying doses, resulting in liver hypertrophy (e.g., doubling of liver weight after 10 days on a diet containing 0.02% PFOA) and spleen and thymus atrophy. Liver hypertrophy was observed at a 10-fold lower dose and at an earlier time following fewer doses, compared with spleen and thymus effects (Yang et al., 2001). Administration of a single dose of 0.02% dietary PFOA (approximately 30 mg/kg/day) was reported to cause a significant suppression of the primary IgM antibody response to horse red blood cells in mice (Yang et al., 2002a). No histopathology of spleen, thymus, or lymph nodes was performed. Based on these findings a Science Advisory Board of U.S. Environmental Protection Agency (USEPA) identified the immune system as a potential target of toxicity (USEPA, 2005).
The primary objective of our study was to evaluate the potential of linear APFO to suppress the primary humoral immune response to sheep red blood cells (SRBCs) in a dose-response fashion in male rats and mice. These studies represent the first time that APFO has been administered by oral gavage and evaluated for effects on the immune system in both rats and mice, using a guideline protocol designed to detect immune effects after 29 days of exposure (USEPA, 1998). In addition, spleen, thymus, and lymph node histopathology and serum corticosterone were evaluated for further indication of immunosuppressive responses.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Materials.
APFO (linear) was provided by DuPont (Wilmington, DE) as a white to slightly opaque liquid in a 19.5% aqueous solution. The dosing solutions were prepared in NANOpure water on a daily basis. Although many of the previous toxicology studies evaluating APFO used a material with a composition of about 80% linear and 20% branched, the material for these studies was manufactured by electrochemical fluorination, resulting in a 100% linear molecule. Loveless et al. (2006)
showed that the 80 and 100% linear materials were very similar in their toxicology profile. Cyclophosphamide (CYP) was obtained from Sigma Chemical (St Louis, MO) and stored desiccated at approximately 4°C. CYP was prepared in sterile saline.
Animals and husbandry.
The research described in this study was conducted in a laboratory accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International and the investigators complied with the regulations and standards of the Animal Welfare Act and adhered to the principles of the Guide for the Care and Use of Laboratory Animals.
The study design complied with the USEPA, OPPTS 870.7800: Immunotoxicity, Health Effects Test Guidelines (1998).
Male Crl:CD(SD)IGS BR rats and Crl:CD-1(ICR)BR mice were obtained from Charles River Laboratories, Inc. (Raleigh, NC). Animal rooms were targeted at a temperature of 22–24°C and a relative humidity of 40–60%, and were artificially illuminated (fluorescent light) on a 12-h light/dark cycle. Throughout the test period, LLC Certified Rodent LabDiet 5002 (PMI Nutrition International, Inc., St Louis, MO) and tap water were available ad libitum. Prior to testing, rats and mice were evaluated by clinical observations and body weight determinations during a 4-day quarantine period to assure freedom from potential confounding variables.
Experimental design.
At approximately 8 weeks of age, rats and mice were randomized into six groups of 10 (rat) or 20 (mice) male animals. All animals were housed singly in stainless steel, wire-mesh cages suspended above cage boards. Each test group was administered daily doses of 0, 0.3, 1, 10, or 30 mg of test material/kg/day by oral gavage beginning on test day 0. Two groups per species received a dose of 30 mg/kg/day. On test day 23 (rat) or 24 (mice), all animals were injected intravenously in the lateral tail vein with 0.5 ml of 4 x 108 SRBC/ml (rat) or 0.2 ml of 1 x 109 SRBC/ml (mouse) (Covance, Denver, PA). One group/species that was dosed with 30 mg/kg/day of test substance through test day 23 (rat) or 24 (mice), were subsequently dosed with NANOpure water until sacrifice. These two groups are designated as receiving 30/0 mg/kg/day. All other groups continued to be dosed with APFO through test day 28 for a total of 29 days of dosing (test day 0 through test day 28). All animals were sacrificed on test day 29. Doses for rats and mice were selected ranging from those expected to be at or near no-observed effect levels to those expected to produce toxic effects. Dosages for this study were chosen based upon the results of a 14-day oral gavage study in male rats and mice (Loveless et al., 2006).
Rats and mice were dosed by intragastric intubation at a standard volume of 10 ml/kg of body weight. Control rats and mice received NANOpure water only. The animals were dosed at approximately the same time each day. Individual dose volumes were calculated using the body weight obtained prior to dosing.
Analysis of dosing preparations.
Samples of dosing preparations were analyzed for uniformity of mixing/concentration verification and stability. Concentrations of compound were measured by reversed-phase liquid chromatography interfaced to a triple quadrupole mass spectrometry (LC-MS/MS). The mobile phase used was acetonitrile/0.15% glacial acetic acid in water, and the column was octylsilyl-derivatized silica. The mass spectrometer was equipped with an electrospray interface operating in the negative ion mode. A stable isotope internal standard (dual 13C-PFOA, custom synthesized by DuPont) was used to aid in the quantification, and one precursor to fragment transition corresponding to (M-CO2) was monitored. Both quadrupoles were set to unit resolution to prevent cross talk of the transitions.
In vivo parameters.
Body weights for rats were recorded on test days 0, 3, 6–28, and daily for mice. Observations for clinical signs of toxicity and mortality were conducted at least twice daily throughout the study. Food consumption was measured twice weekly during the study. Body weight gains were calculated from test day 0 to test day 28.
Collection, processing, and analysis of blood and tissues.
Blood collection and necropsy procedures were conducted in a stratified order (e.g., first animal in group 1, first animal in group 2, etc.). Animals were maintained in a quiet area separate from the necropsy suite prior to collection of samples. Blood samples for immune and clinical pathology parameters were collected approximately 24 h after the last dose of APFO or vehicle. After final body weights were determined, animals were anesthetized by CO2 anesthesia, and blood was collected from the orbital sinus of each rat and mouse, immediately after anesthesia for hematology (mice only), clinical chemistry, and hormone measurements. Additionally, blood was collected from the abdominal vena cava of each rat for hematology. Sera collected for antibody quantitation was frozen. Animals were euthanized by exsanguination while under anesthesia. Liver, brain, spleen and thymus were removed and weighed and relative (to body weight) organ weights were calculated. In addition, weighed organs, popliteal, and mesenteric lymph nodes, bone marrow, femur/knee joint and sternum were collected and fixed in 10% buffered formalin. All tissues collected from control (0 mg/kg/day) and high-dose (30 and 30/0 mg/kg/day) rats and mice were processed to slides and evaluated microscopically. In addition, target tissues were examined from rats and mice in all dose groups to determine a no-observed-effect level for test-substance–related microscopic findings.
Complete blood counts, including reticulocytes, were determined on a Bayer Advia 120 hematology analyzer or determined from microscopic evaluation of the blood smear. Wright-Giemsa–stained blood smears from all animals were examined microscopically for confirmation of automated results and evaluation of cellular morphology.
To determine total cell counts for each spleen and thymus, each organ was weighed, cut in half and the weight of the halved spleen or thymus recorded. Each half was placed into Hank's Balanced Salt Solution (GIBCO, Carlsbad, CA), and cut into small pieces. Rat organ pieces were then poured into a Stomacher 80 Lab System bag, and placed into a Stomacher 80 Lab System on "high" setting for 120 s (spleen) or 60 s (thymus). Mouse organ pieces were gently "mashed" between the frosted ends of microscope slides to create a cell suspension. Cells were counted by hemacytometer and total cell count/organ calculated.
Serum lipids (cholesterol, high density lipoprotein (HDL) cholesterol, non-HDL cholesterol [calculated], and triglycerides) and protein (total protein, albumin, and globulin) were determined on an Olympus AU640 clinical chemistry analyzer (Melville, NY). Serum corticosterone was measured using a commercial radioimmunoassay (Diagnostic Products Corporation, Los Angeles, CA; Catalog #TKRC1).
Antibody assessment.
Humoral immune function was evaluated by examining sera from individual animals for SRBC-specific IgM levels with an enzyme-linked immunosorbent assay (ELISA) (Temple et al., 1993). To demonstrate that the assay functioned properly, sera was collected from rats and mice previously injected with SRBC and dosed i.p. for 6 days with 20 mg/kg/day (rats) or 5 days with 90 mg/kg/day (mice) of the known immunosuppressive agent CYP or vehicle. These sera were run concurrently with the study samples.
Statistical analysis.
Preliminary tests were conducted for homogeneity of variance (Levene, 1960) and normality (Shapiro and Wilk, 1965). If data were normally distributed and had homogeneity of variances, a one-way analysis of variance (Snedecor and Cochran, 1967) was conducted, followed by Dunnett's (1964) test. For data that did not show homogeneity of variances, a robust version of Dunnett's (1980) test was used. Non-normally distributed data was analyzed by non-parametric procedures (Kruskal and Wallis (1952) test followed by Dunn's (1964) test). For all statistical analyses, significance was judged at p < 0.05. Comparisons were made between the vehicle control group and the dosed or positive control groups. Comparisons were also made between the 30 and 30/0 mg/kg/day groups.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Data from the analysis of the samples during the study indicate that the test substance was at the targeted concentrations, mixed uniformly, and was stable under the conditions of the study. Test substance was not found in the 0 mg/ml samples (limit of detection = 10 ng/ml).
Rats
All rats dosed with APFO survived until the end of the study. Reductions in mean body weights and body weight gains were observed at 10, 30, and 30/0 mg/kg/day. Mean final body weights on day 28 were 10, 25, and 21% lower than the control group at 10, 30, and 30/0 mg/kg/day, respectively, as a result of reduced body weight gains (Fig. 1); overall body weight gains during test days 0–28 were 26, 63, and 50% lower for the same respective doses (data not shown). The magnitude and onset of the effects on body weight parameters were dose related in that the effects at 30 mg/kg/day were evident sooner and were more pronounced. There was no appreciable difference in the magnitude of the reduction between the 30 and 30/0 mg/kg/day.
|
Hemoglobin and hematocrit were mildly decreased in rats dosed with 10 or 30 mg/kg/day (means were 91–92% of the control group mean, respectively; statistically significant), but there were no effects on red blood cell counts (data not shown). Reticulocytes were moderately increased in rats dosed with 30/0 mg/kg/day (mean reticulocyte count was 197% of the control group mean), which correlated with histologic evidence of increased extramedullary hematopoiesis.
Total cholesterol was decreased in rats dosed with 0.3 or 1 mg/kg/day (Table 1). The decreases in cholesterol were due to decreases in both HDL and non-HDL cholesterol. Triglyceride was decreased in rats dosed with 
0.3 mg/kg/day with a flat dose-response. Triglycerides were still decreased in rats that were dosed with 30/0 mg/kg/day, indicating a lack of recovery for triglyceride concentrations.
|
Serum corticosterone was increased in a few rats dosed with 10 or 30 mg/kg/day (Fig. 2). Concentrations greater than 300 ng/ml (approximate upper limit for corticosterone concentration in non-stressed rats in this study) were observed in 2/10 and 4/10 rats dosed with 10 or 30 mg/kg/day, respectively, resulting in mean concentrations that were 135 and 196% of controls, respectively. In rats that were dosed with 30/0 mg/kg/day, serum corticosterone concentrations were generally similar to controls, indicating recovery.
|
No test-substance–related evidence of immunosuppression was observed at any dose; the IgM titers were generally comparable across all groups (Fig. 3). In contrast, the primary humoral immune response to SRBC was significantly decreased to 43% of control in sera from rats administered CYP.
|
Liver weights were significantly increased in rats dosed with
1 mg/kg/day (absolute) and 10 mg/kg/day (relative to body weight) (Table 2). The increased liver weights, at all dose levels, correlated with the microscopic finding of minimal to mild hepatocellular hypertrophy at 0.3 and 1 mg/kg/day and moderate hepatocellular hypertrophy at 10 mg/kg/day. Minimal focal necrosis was observed in 4 of 10 rats in the 30 mg/kg/day group and in 1 of 10 rats in the 10 mg and 30/0 mg/kg/day groups. Microscopic examination of lymphohematopoietic organs (spleen, thymus, bone marrow, lymph nodes) revealed increased hematopoiesis in the spleen of rats dosed with 30/0 mg/kg/day; the thymus, mesenteric lymph nodes and popliteal lymph nodes had no test-substance–related effects (Table 3).
|
|
No significant changes in total spleen cell or thymocyte number compared with control rats were noted in any groups treated with any dose of APFO (Table 4). For rats in the 30/0 mg/kg/day group, thymocyte number was increased compared with rats who continued to receive 30 mg/kg/day APFO, but not compared with vehicle control.
|
The no-observed-adverse-effect level (NOAEL) for APFO for systemic toxicity in male rats was 1 mg/kg/day, based upon focal liver necrosis at 10 mg/kg/day, whereas the NOAEL for immunotoxicity was 30 mg/kg/day, the highest dose tested.
Mice
In mice exposed to 10 and 30 mg/kg/day APFO, marked general systemic toxicity was observed, as evidenced by a loss in body weight of 3.8 and 6.6 g, respectively, compared with control mice. These doses resulted in decreases in body weight to 86 and 78% of control mice, respectively. In general, the test-substance–related effects on body weight parameters were dose related relative to the magnitude of the change and the onset of the reductions; effects on mice dosed at 30 mg/kg/day were more pronounced and evident sooner than effects at 10 mg/kg/day (Fig. 4). Test-substance–related increases in liver weights occurred (greater than threefold increases at 10 and 30 mg/kg/day compared with control; Table 2) and reduced the magnitude of the effects on body weight. To account for the potentially mitigating effects of the marked liver weight increases on the decreased body weights, an adjusted body weight was calculated for each animal by subtracting the weight of the liver from the final body weight. The means for the liver-weight-adjusted body weights were 28 and 35% lower than controls at 10 and 30 mg/kg/day, respectively (see Fig. 5).
|
|
Neutrophils and monocytes were increased in mice dosed with either 10 or 30 mg/kg/day. Treatment means were 236 and 296% (neutrophils) and 285 and 254% (monocytes) of the control group means, respectively. In these mice, individual mice with increased monocytes tended to have increased neutrophil counts. Lymphocytes were generally decreased in mice dosed with 30 mg/kg/day. Eosinophils were decreased in mice dosed with 10 or 30 mg/kg/day, with means of 57 and 64% of the control group mean, respectively (not statistically significant).
Total cholesterol was moderately decreased in mice dosed with 10 or 30 mg/kg/day (means were 69 and 51% of the control group means) due to decreases in HDL cholesterol and in non-HDL cholesterol (Table 1). Triglycerides were moderately decreased in mice dosed with 10 or 30 mg/kg/day (means were 47 and 32% of the control group mean, respectively).
Serum corticosterone was moderately increased in several mice dosed with 10 or 30 mg/kg/day (Fig. 6). Although serum corticosterone concentrations were between 0 and 400 ng/ml in all mice dosed with 
1 mg/kg/day, concentrations greater than 400 ng/ml were observed in 7/10 and 6/10 mice dosed with 10 or 30 mg/kg/day, respectively, resulting in mean concentrations that were 230 and 232% of the control group mean for mice for these two groups. In the 30/0 mg/kg/day group, mouse serum corticosterone concentrations were still mildly increased (6/10 were greater than 400 ng/ml, and the mean was 154% of the control group mean).
|
There was test-substance–related evidence of immunosuppression in mice at 10, 30, and 30/0 mg/kg/day. The anti-SRBC titers for these groups were reduced to 80, 72, and 70% of control group means (Fig. 7). There was no difference in mean primary humoral immune response between the 30 and 30/0 mg/kg/day, indicating that the shortened dosing and recovery period did not have an impact on this endpoint. In contrast, mice injected for 5 days with 90 mg/kg/day of the immunosuppressive material, CYP, demonstrated a decreased IgM antibody response to SRBC, which was 48% of control mice response.
|
When the serum corticosterone concentration of each mouse (excluding mice in the 30/0 group) was plotted against its corresponding antibody value, a correlation coefficient of –0.428 was obtained, which was significant at the p < 0.002 level, indicating that as corticosterone levels increased, the ability to make antibody to SRBC decreased.
Mean absolute and relative liver weights were increased in a dose responsive manner compared with the control values (Table 2). Mean relative (% body weight) liver weights were increased at all doses (33, 79, 292, 317, and 291%, respectively, compared with control means) with all increases statistically significant, except for those in the 0.3 mg/kg/day dose group. The increased liver weights correlated with the microscopic finding of mild hepatocellular hypertrophy at 0.3 mg/kg/day; moderate to severe hepatocellular hypertrophy with individual and focal necrosis at doses 1 mg/kg/day; and increased hepatocellular mitotic figures, hepatocellular fatty change, and bile duct hyperplasia at doses 10 mg/kg/day (Table 3).
Mean absolute spleen weights were decreased to 89, 56, 44, and 65% of control means in the 1, 10, 30, and 30/0 mg/kg/day dose groups, respectively (Table 2). Mean relative (% body weight) spleen weights were similarly decreased. Decreased spleen weights at 10 mg/kg/day correlated with the gross observation of small spleens and significant decreases in spleen cell number at doses 10 mg/kg/day. The greatest decrease in spleen cell number compared with vehicle-treated controls was observed at 30 mg/kg/day (27% of control group mean) (Table 4). In the 30/0 mg/kg/day group, partial recovery occurred (56% of control group mean), but this increase in cells compared with the 30 mg/kg/day was most likely due to increased extramedullary hematopoiesis observed in 15 of 19 mice in this treatment group (Table 3).
Mean absolute thymus weights were decreased to 50, 50, and 54% of control group mean in the 10, 30, and 30/0 mg/kg/day dose groups, respectively (Table 2). Mean relative (% body weight) thymus weights were similarly decreased. The decreased thymus weights, at 10 mg/kg/day, correlated with the microscopic finding of minimal to severe lymphoid depletion/atrophy in this organ and with the gross observation of small thymuses at doses 10 mg/kg/day (Table 3). Significant decreases in thymocyte number were noted in animals dosed with 10 mg/kg/day or greater, with the greatest suppression observed at 30 mg/kg/day (18% of control group mean) (Table 4). In the 30/0 mg/kg/day group, partial recovery occurred (49% of control group mean).
The NOAEL for APFO for systemic toxicity in male mice was 0.3 mg/kg/day based upon liver single cell and focal necrosis at 1 mg/kg/day, whereas the NOAEL for immunotoxicity was 1 mg/kg/day, based upon suppression of anti-SRBC IgM antibody levels observed at 10 mg/kg/day.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This study is the most comprehensive evaluation to date of the effects of APFO on the immune system. It is the first to evaluate the effects in two different rodent species, including the rat, which is the standard rodent species in toxicology, and employed oral gavage, which enables the most accurate assessment of administered dose. This study is the only one to our knowledge which evaluated immune cell function, immune organ weight, immune cell number, histopathology, and serum corticosterone levels all in the same experiment. These experiments are also the first to evaluate these parameters following a dosing duration of nearly a month, which is the worldwide standard for identifying a hazard to the immune system. In addition, because serum concentrations of PFOA were measured in an earlier study after two weeks of dosing at the same doses used in the current study (Loveless et al., 2006
), serum PFOA concentrations of workers and the general population can be put into perspective with regard to effects on the immune system.
Subchronic repeated dose studies conducted with APFO have previously shown no evidence of effects on the immune system. Rats fed from 10 to 1000 ppm (equivalent to approximately 0.7–70 mg/kg/day) for 90 days showed no histopathologic changes in either the spleen or the thymus and organ weights were normal (Griffith and Long, 1980). In a two-generation reproduction study in rats given from 1 to 30 mg/kg/day, no effects on spleen or thymus weights were seen in F0 or F1 parental animals or the F1 or F2 pups (Butenhoff et al., 2004). Two nonhuman primate studies also showed no effects on the immune system. In rhesus monkeys, spleen or thymus weights were unaffected following 90 days of oral treatment with up to 30 mg/kg/day. Histopathologic changes were only observed in lymphoid tissues at lethal doses of 30 and 100 mg/kg/day (Griffith and Long, 1980). In cynomolgus monkeys given from 3 to 30/20 mg/kg/day for 6 months, no microscopic pathologic findings were seen in the spleen, thymus, or mesenteric lymph nodes at any dose (Butenhoff et al., 2002). The absence of any effect in these tissues in repeated dose studies in both rats and primates suggests that the immune system is not a target for APFO in these two species.
In a series of studies published by Yang et al. (2000, 2001, 2002a), male C57B1/6 mice were typically fed PFOA at dietary concentrations of 0.02% (30 mg/kg/day based on daily intake) for up to 10 days, which resulted in weight loss of up to 17% of their original body weight. This dose also resulted in relative liver weights that were 236% of control mice, whereas relative thymus and spleen weights were 17 and 77% of control mice (Yang et al., 2000). In one experiment, mice fed 0.02% PFOA were immunized with horse red blood cells and the humoral immune response was measured by both the plaque-forming cell assay and determination of the antibody titer by ELISA (Yang et al., 2002a). PFOA treatment was reported to prevent both the increase in plaque-forming cells (anti-IgM and IgG) and increase in serum levels of IgM and IgG normally produced by such immunization. These results for IgM antibody suppression using the splenic plaque assay are very difficult to interpret, however, given that naïve control mice, never exposed to the horse red blood cell antigen, had, on average, 10,000 anti-horse red blood cell antibody-forming cells/spleen (vs. 27,500 ± 8699 in immunized control mice). An expected number would have been one or two. To our knowledge confirmation of these results (related to feeding of PFOA at any dose level) has not been reported in the literature. Following dietary administration of 0.02% PFOA, spleen cell proliferation in vitro in response to either concanavalin A or lipopolysaccharide was attenuated by about 20 and 50%, respectively, after 2 days in culture. In contrast, in vitro treatment of naïve mouse spleen cells with concentrations of PFOA ranging from 1 to 200 µM had no effect, suggesting an indirect mechanism of action, such as cytokine production or lipid regulation (Yang et al., 2002a). Similar effects were reported for thymocyte proliferation, with inhibition observed following in vivo exposure to PFOA, but no inhibition of naïve thymocytes exposed to varying concentrations of PFOA in vitro (Yang et al., 2000).
In this study, care was taken to decrease stress from experimental procedures. The test article was administered in an aqueous vehicle at 10 mg/kg body weight. Four-fold higher volumes of aqueous vehicles have been shown to be nonstressful in the rat (Brown et al., 2000). Animals were maintained in a area separate from the necropsy suite, and extraneous procedures were minimized. Blood was collected for hormonal measurements immediately after anesthesia with CO2 to minimize the effects of CO2 on corticosterone concentrations. Therefore, increased corticosterone observed consistently in a group of animals would be considered to be indicative of test article-related stress, rather than methodologically introduced bias.
In general, rats dosed with the test article did not show evidence of stress, serum corticosterone concentrations of most rats were within the range previously observed in this laboratory in blood collected from the retroorbital sinus under CO2 anesthesia or by decapitation (range from approximately 20 to 170 ng/ml; unpublished observations). In addition, corticosterone concentrations of most rats were within ranges reported in the literature using a variety of collection techniques (reviewed by Fomby et al., 2004) Although serum corticosterone was increased in a few rats dosed with 10 or 30 mg/kg/day, these rats did not differ from other rats in their groups using other parameters that often indicate stress such as body weight changes or stress-type leukocyte patterns (e.g., no increases in neutrophils and decreases in lymphocytes and eosinophils; Latimer and Prasse, 2003; Pruett et al., 2007).
The studies reported here confirm the results of previous studies, which demonstrated that the liver is the primary target organ of toxicity following exposure to APFO (Butenhoff et al., 2004; Kennedy et al., 2004; USEPA, 2002). In the rat following oral gavage, all doses resulted in an increase in liver weight, which correlated with an increase in hepatocellular hypertrophy. In mice, as in rats, the liver was the primary target organ of toxicity, as evidenced by hepatocellular hypertrophy (0.3 mg/kg/day and above), individual cell and focal necrosis (1 mg/kg/day and above), and increased mitotic figures and bile duct hyperplasia (10 mg/kg/day and above).
In the current studies, the immune system was a much less sensitive target than the liver for APFO-induced changes, and evidence of immunotoxicity in mice was primarily limited to doses that produced marked systemic toxicity. In rats, no inhibitory effect was observed on anti-SRBC IgM antibody production nor on spleen or thymus cellularity, even at a dose that caused focal hepatocellular necrosis in the liver (30 mg/kg/day). The absence of changes in these immune-related parameters occurred despite increases in serum corticosterone levels in some rats in the 10 and 30 mg/kg/day groups.
In mice, a number of findings occurred at 10 and 30 mg/kg/day in association with severe effects on body weight parameters that likely represent stereotypical secondary responses to the systemic toxicity and stress. These effects included decreased IgM antibody (20 and 28% suppression at 10 and 30 mg/kg/day, respectively); decreased spleen and thymus weights and cell numbers; microscopic depletion/atrophy of lymphoid tissue starting at 10 mg/kg/day in the thymus and 30 mg/kg/day in the spleen, and increased serum corticosterone levels. Leukogram changes were characterized by significant increases in absolute numbers of blood neutrophils and monocytes and decreases in absolute lymphocyte numbers and peripheral blood eosinophils (30 mg/kg/day). These leukocyte changes, especially the increases in neutrophils and decreases in lymphocytes, are typical of those observed in stressed animals of many species, including mice (Dhabhar et al., 1994; Latimer and Prasse, 2003). The finding that the antibody response to SRBC was not inhibited in rats, but was inhibited in mice at high doses is not surprising based on studies by Pruett et al. (1999, 2007), that show that immune parameters in mice are more sensitive to the effects of stress (as measured by corticosterone) compared with rats.
Consistent with the changes noted above for mice, serum corticosterone concentrations of greater than 400 ng/ml were observed in 7/10 and 6/10 mice dosed with 10 or 30 mg/kg/day, respectively, resulting in mean concentrations that were approximately 2.3-fold higher compared with control mice. In contrast, serum corticosterone concentrations were between 0 and 400 ng/ml in all mice dosed with 1 mg/kg/day.
The serum corticosterone concentrations associated with the two higher doses are in agreement with other published studies in mice evaluating the effect of corticosterone administration on antibody levels. A single dose of 100 mg hydrocortisone/kg body weight gave a peak serum concentration of corticosterone of 500 ng/ml after two days (Dracott and Smith, 1979a), which resulted in significant inhibition of the primary IgM response to SRBC (Dracott and Smith, 1979b). Pruett et al. (1999) injected 18 mg/kg of exogenous corticosterone in mice at 0 and 2 h and calculated the serum corticosterone area under the curve (AUC) value to be about 6400 ng/ml·h, with a peak corticosterone level of 1100 ng/ml at 2 h, falling to about 200 ng/ml at 12 h. This AUC occurred at a concentration required for 25% inhibition of a number of immune parameters, including antigen-specific antibody responses, natural killer cell activity, and spleen cell number (Pruett and Fan, 2001). In the current study, serum corticosterone was measured only once approximately 24 h after the last of 29 consecutive daily doses of APFO, so an AUC value was not able to be calculated. But the significant negative correlation between serum corticosterone levels and antibody levels observed in the current study supports the hypothesis that administration of 10 or 30 mg APFO/kg body weight resulted in serum concentrations of corticosterone sufficient to inhibit the IgM antibody response to SRBC.
Based on both body weight decreases and the evidence for significant stress in mice, these two higher doses exceeded the recommended dose level for immunotoxicity testing as given in the OPPTS 870.7800 Immunotoxicity testing guidance, which states: "The highest dose level should not produce significant stress, malnutrition, or fatalities, but ideally should produce some measurable sign of general toxicity (e.g., a ten percent loss of body weight)" (USEPA, 1998). In other words, had a goal of these studies not been to investigate stress endpoints at doses previously studied for immunotoxicity, the highest dose selected would have been less than 10 mg/kg/day.
Other high-dose effects on the immune system have been reported following exposure to PFOA. Fairley et al. (2007) published that dermal exposure to PFOA enhances total and ovalbumin-specific IgE antibody in inbred mice. These effects, however, were only observed using relatively high doses of PFOA. The lowest-observed-effect-level for enhanced IgE production at 14 days was 0.75% PFOA applied in acetone to the ear of mice for 4 days. This concentration of PFOA is equivalent to 18.75 mg/kg/day and resulted in almost a twofold increase in relative liver weight and a 75% decrease in spleen cellularity when measured ten days after the final application of PFOA. This dose also translates into 187,500 ng PFOA/cm2 of mouse skin. This amount far exceeds the potential exposure to humans from any source. Using treated carpets as a potential exposure source, the maximum concentration of PFO (anion of PFOA) that was able to be extracted from carpet treated with a chemical made using PFOA in the manufacturing process was less than or equal to 50 ng/cm2 (Washburn et al., 2005). Furthermore, rat skin, as well as skin from the mouse, rabbit, and guinea pig, has consistently been shown to be more permeable to topically applied compounds compared with human skin (USEPA, 1992). Indeed, when APFO was applied to rat and human skin in an in vitro dermal penetration assay, the penetration of APFO was 34-fold fold slower through human compared with rat skin (Fasano et al., 2005). Therefore, given the potential low exposure compared with an effect level plus the decreased absorption through human skin compared with rodent skin, the risk of heightened IgE production in humans due to exposure of PFOA appears quite small.
The role of the peroxisome proliferator activated receptor alpha (PPAR-
) on the immune system of PFOA-exposed rodents has been reported. Yang et al. (2002b) published a feeding study comparing the effects of PFOA in wild type and transgenic PPAR- null mice. These PPAR- "knockout" mice do not exhibit hepatic peroxisome proliferation after exposure to peroxisome proliferators. No significant decrease in spleen weight or splenocyte numbers or in vitro response to mitogens was observed in PPAR- null mice fed PFOA. Decreases in thymus weight and cell number were less prominent but not completely attenuated in mice fed 0.02% PFOA for 7 days. These results indicate that PPAR- plays a role in certain splenic and thymic effects observed in mice following administration of PFOA in the diet. To date there are no published studies on the role of PPAR- in modulating the effect of PFOA on antibody responses to an antigen. However, PPAR- knockout mice are being used to further evaluate the effects of PFOA on the immune system at other laboratories (personal communications) and these studies may help to further elucidate the relevance of this response to humans.
The results presented in the current study are quite similar to results reported by DeWitt et al. (2008). C57Bl/6 mice were exposed to APFO in drinking water for 15 days and the IgM antibody response to SRBCs was evaluated using an ELISA. The water concentration equivalent to a daily intake of 30 mg/kg/day resulted in a 29% suppression (compared with 28% suppression in this study at 30 mg/kg/day). Their LOAEL of 3.75 mg/kg/day resulted in 11 and 7% suppression in repeat experiments with a NOAEL of 1.88 mg/kg/day, compared with a nonstatistically significant 8% suppression in this study at a NOAEL of 1 mg/kg/day by oral gavage. Similarly, the ELISA data in the Yang et al. (2002a) paper indicates that 0.02% PFOA in chow (30 mg/kg/day) resulted in a maximum suppression of 30% of the IgM antibody response to horse red blood cells.
In mice, based on our work and others, the NOAEL for suppressing the primary immune response to SRBC is likely to fall between 1 and 2 mg/kg/day. With a LOAEL close to 4 mg/kg/day, the degree of suppression of the primary immune response at this concentration is around 10%, indicating that APFO has minimal effects on the ability to make IgM antibody in rodents. Given that serum concentrations of PFOA after 14 days of dosing with 3–3.75 mg/kg/day APFO have been reported to be 69–75 ppm (CD-1 mice, Loveless et al., 2006; C57Bl/6 mice, DeWitt et al., 2008), and the general U.S. population average serum concentration of PFOA is 4–5 ppb (Olsen et al., 2003c), and appears to be declining (Calafat et al., 2007, Olsen et al., 2008), the likelihood of immunosuppression in the human population due to PFOA seems highly unlikely.
In conclusion, these studies confirmed the results of previous studies in several species, which demonstrate that the liver is the most sensitive target following administration of APFO. This investigation advanced previous studies evaluating the effects of APFO and PFOA on the immune system by incorporating a second rodent species, dose-response evaluation after 29 days of dosing, histopathology, and serum corticosterone measurements. The immune system is a less sensitive target than the liver, with effects occurring in mice at or near doses that produce significant systemic toxicity.
|
NOTES |
|---|
2 Current address: 1201 Amgen Court West, Seattle, WA 98119.
|
ACKNOWLEDGMENTS |
|---|
We would like to thank Gerry Kennedy and Robert Rickard for their insightful comments during the preparation of this manuscript.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Brown AP, Dinger N, Levine BS. Stress produced by gavage administration in the rat. Contemp. Top. Lab. Anim. Sci. (2000) 39(1):17–21.[Web of Science][Medline]
Butenhoff J, Costa G, Elcombe C, Farrar D, Hansen K, Iwai H, Jung R, Kennedy G, Lieder P, Olsen G, et al. Toxicity of ammonium perfluorooctanoate in male cynomolgus monkeys after oral dosing for 6 months. Toxicol. Sci. (2002) 69:244–257.
Butenhoff JL, Gaylor DW, Moore JA, Olsen GW, Rodricks J, Mandel JH, Zobel LR. Characterization of risk for general population exposure to perfluorooctanoate. Regul. Toxicol. Pharmacol. (2004) 39:363–380.[CrossRef][Web of Science][Medline]
Calafat AM, Wong L-Y, Kuklenyik Z, Reidy JA, Needham LL. Polyfluoroalkyl chemicals in the U.S. population: Data from the National Health and Nutrition Examination Survey (NHANES) 2003–2004 and comparisons to NHANES 1999–2000. Environ. Health Perspect. (2007) 115:1578–83.[Web of Science][Medline]
DeWitt JC, Copeland CB, Strynar MJ, Luebke RW. Perfluorooctanoic acid-induced immunomodulation in adult C57BL/6J or C57BL/6N female mice. Environ. Health Perspect. (2008) 116:644–650.[Web of Science][Medline]
Dhabhar FS, Miller AH, Stein M, McEwen BS, Spencer RL. Diurnal and acute stress-induced changes in distribution of peripheral blood leukocyte subpopulations. Brain Behav. Immun. (1994) 8:66–79.[CrossRef][Web of Science][Medline]
Dracott BN, Smith CET. Hydrocortisone and the antibody response in mice. I. Correlations between serum cortisol levels and cell numbers in thymus, spleen, marrow and lymph nodes. Immunology (1979a) 38:429–435.[Web of Science][Medline]
Dracott BN, Smith CET. Hydrocortisone and the antibody response in mice. II. Correlations between serum antibody and PFC in thymus, spleen, marrow and lymph nodes. Immunology (1979b) 38:437–443.[Web of Science][Medline]
Dunn OJ. Multiple contrasts using rank sums. Technometrics (1964) 6:241–252.[CrossRef]
Dunnett CW. New tables for multiple comparisons with a control. Biometrics (1964) 20:482–491.[CrossRef][Web of Science]
Dunnett CW. Pairwise multiple comparisons in the unequal variance case. J. Am. Stat. Assoc. (1980) 75:796–800.[CrossRef][Web of Science]
Fairley KJ, Purdy R, Kearns S, Anderson SE, Meade BJ. Exposure to the immunosuppressant, perfluorooctanoic acid, enhances the murine IgE and airway hypersensitivity response to ovalbumin. Toxicol. Sci. (2007) 97:375–383.
Fasano WJ, Kennedy GL, Szostek B, Farrar DG, Ward RJ, Haroun L, Hinderliter PM. Penetration of ammonium perfluorooctanoate (APFO) through rat and human skin in vitro. Drug Chem. Toxicol. (2005) 28:79–90.[CrossRef][Web of Science][Medline]
Fomby LM, Wheat TM, Hatter DE, Tuttle RL, Black CA. Use of CO2/O2 anesthesia in the collection of samples for serum corticosterone analysis from Fischer 344 rats. Contemp. Top. Lab. Anim. Sci. (2004) 43:8–12.[Web of Science][Medline]
Griffith FD, Long JE. Animal toxicity studies with ammonium perfluorooctanoate. Am. Ind. Hyg. Assoc. J. (1980) 41:576–583.[Web of Science][Medline]
Hansen KJ, Clemen LA, Ellefson ME, Johnson HO. Compound-specific, quantitative characterization of organic fluorochemicals in biological matrices. Environ. Sci. Technol. (2001) 35:766–770.[Medline]
International Research and Development Corporation. Ninety day subacute rat toxicity study. US Environmental Protection Agency Administrative Record (1978) 226-0441.
Kennedy GL, Butenhoff JL, Olsen GW, O'Connor JC, Seacat AM, Perkins RG, Biegel LB, Murphy SR, Farrar DG. The toxicology of perfluorooctanoate. Crit. Rev. Toxicol. (2004) 34:351–384.[CrossRef][Web of Science][Medline]
Kruskal WH, Wallis WA. Use of ranks in one-criterion analysis of variance. J. Am. Stat. Assoc. (1952) 47:583–621.[CrossRef][Web of Science]
Latimer K, Prasse K. Leukocytes. In: Duncan & Prasse's Veterinary Laboratory Medicine: Clinical Pathology—Latimer KS, Mahaffey EA, Prasse KW, eds. (2003) 4th edn. Ames, IA: Blackwell Publishing. 46–79.
Lau C, Anitole K, Hodes C, Lai D, Pfahles-Hutchens A, Seed J. Perfluoroalkyl acids: A review of monitoring and toxicological findings. Toxicol. Sci. (2007) 99:366–394.
Levene H. Robust test for equality of variances. In: Contributions to Probability and Statistics—Olkin J, ed. (1960) Palo Alto, CA: Stanford University Press. 278–292.
Loveless SE, Finlay C, Everds NE, Frame SR, Gillies PJ, O'Connor JC, Powley CR, Kennedy GL. Comparative responses of rats and mice exposed to linear/branched, linear, or branched ammonium perfluorooctanoate (APFO). Toxicology (2006) 220:203–217.[CrossRef][Web of Science][Medline]
Olsen GW, Burris JM, Burlew MM, Mandel JH. Epidemiologic assessment of worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations and medical surveillance examinations. J. Occup. Environ. Med. (2003a) 45:260–70.[CrossRef][Web of Science][Medline]
Olsen GW, Hansen KJ, Stevenson LA, Burris JM, Mandel JH. Human donor liver and serum concentrations of perfluorooctanesulfonate and other perfluorochemicals. Environ. Sci. Technol. (2003b) 37:888–891.[Medline]
Olsen GW, Church TR, Miller JP, Burris JM, Hansen KJ, Lundberg JK, Armitage JB, Herron RM, Medhdizadehkashi Z, Nobiletti JB, et al. Perfluorooctanesulfonate and other fluorochemicals in the serum of American Red Cross adult blood donors. Environ. Health Perspect. (2003c) 111:1892–1901.[Web of Science][Medline]
Olsen GW, Mair DC, Church TR, Ellefson ME, Reagen WK, Boyd TM, Herron RM, Medhdizadehkashi Z, Nobiletti JB, Rios JA, et al. Decline in perfluorooctanesulfonate and other polyfluoroalkyl chemicals in American Red Cross adult blood donors, 2000–2006. Environ. Sci. Technol. (2008) doi:10.1021/es800071x.
Pruett SB, Collier S, Wu WJ, Fan R. Quantitative relationships between the suppression of selected immunological parameters and the area under the corticosterone concentration vs. time curve in B6C3F1 mice subjected to exogenous corticosterone or to restraint stress. Toxicol. Sci. (1999) 49:272–280.
Pruett SB, Fan R. Quantitative modeling of suppression of IgG1, IgG2a, IL-2, and IL-4 responses to antigen in mice treated with exogenous corticosterone or restraint stress. J. Toxicol. Environ. Health A (2001) 62:175–189.[CrossRef][Web of Science][Medline]
Pruett SB, Hebert P, Lapointe J, Reagan ML, Kawabata TT. Characterization of the action of drug-induced stress responses on the immune system: Evaluations of biomarkers for drug-induced stress in rats. J. Immunotoxicol. (2007) 4:25–38.[CrossRef][Medline]
Riker Laboratories. Two year oral (diet) toxicity/carcinogenicity study of FC-143 in rats. US Environmental Protection Agency Administrative Record (1987) 226-0437, 226-0438, 226-0439, pp. 226–440.
Shapiro SS, Wilk MB. An analysis of variance test for normality (complete samples). Biometrika (1965) 52:591–611.
Snedecor GW, Cochran WG. Statistical Methods (1967) 6th edn. Iowa: The Iowa State University Press.
Taves D, Guy W, Brey W. Organic fluorocarbons in human plasma: Prevalence and characterization. In: Biochemistry Involving Carbon-Fluorine Bonds—Filler R, ed. (1976) Washington, DC: American Chemical Society. 117–134.
Temple L, Kawabata TT, Munson AE, White KL Jr. Comparison of ELISA and plaque-forming cell assays for measuring the humoral immune response to SRBC in rats and mice treated with benzo[a]pyrene or cyclophosphamide. Fund. Appl. Toxicol. (1993) 21:412–419.[CrossRef][Web of Science][Medline]
USEPA. Dermal Exposure Assessment: Principles and Applications (1992).
USEPA. OPPTS 870.7800: Immunotoxicity. Health Effects Test Guidelines (1998).
USEPA. Revised draft. Hazard assessment of perfluorooctanoic acid and its salts. (2002) USEPA Public Docket AR-226.
USEPA. Draft risk assessment of the potential human health effects associated with exposure to perfluorooctanoic acid and its salts, 4 January 2005. (2005) Available at: http://www.epa.gov/opptintr/pfoa/pubs/pfoarisk.pdf.
Vanden Heuvel J, Kuslikis B, Van Rafelghem M. Tissue distribution, metabolism, and elimination of perfluorooctanoic acid in male and female rats. Biochem. Toxicol. (1991) 6:83–92.[CrossRef]
Washburn ST, Bingman TS, Braithwaite SK, Buck RC, Buxton LW, Clewell HJ, Haroun LA, Kester JE, Rickard RW, Shipp AM. Exposure assessment and risk characterization for perfluorooctanoate in selected consumer articles. Environ. Sci. Technol. (2005) 39:3904–3910.[Medline]
Yang Q, Abedi-Valugerdi M, Zie Y, Zhao X, Moller G, Nelson BD, DiPierre JW. Potent suppression of the adaptive immune response in mice upon dietary exposure to the potent peroxisome proliferator, perfluorooctanoic acid. Int. Immunopharmacol. (2002a) 2:389–397.[CrossRef][Web of Science][Medline]
Yang Q, Xie Y, DiPierre JW. Effects of peroxisome proliferators on the thymus and spleen of mice. Clin. Exp. Immunol. (2000) 122:219–226.[CrossRef][Web of Science][Medline]
Yang Q, Xie Y, Ericksson AM, Nelson BD, DiPierre JW. Further evidence for the involvement of inhibitors of cell proliferation and development in thymic and splenic atrophy induced by the peroxisome proliferator perfluorooctanoic acid in mice. Biochem. Pharmacol. (2001) 62:1133–1140.[CrossRef][Web of Science][Medline]
Yang Q, Zie Y, Alexson SEH, Nelson BD, DePierre JW. Involvement of the peroxisome proliferator activated receptor alpha in the immunomodulation caused by peroxisome proliferators in mice. Biochem. Pharmacol. (2002b) 63:1893–1900.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. C. DeWitt, C. B. Copeland, and R. W. Luebke Suppression of Humoral Immunity by Perfluorooctanoic Acid is Independent of Elevated Serum Corticosterone Concentration in Mice Toxicol. Sci., May 1, 2009; 109(1): 106 - 112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Fang, Y. Feng, Z. Shi, and J. Dai Alterations of Cytokines and MAPK Signaling Pathways are Related to the Immunotoxic Effect of Perfluorononanoic Acid Toxicol. Sci., April 1, 2009; 108(2): 367 - 376. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||