Document qm0yKD1vBDXvJNyB3D83NZqkk

"McCrea, Deborah" <mccrea@taftlaw .com> 12/28/2009 01 :55 PM To NCIC OPPT@EPA cc Bilott, Robert A." <bilott@taftlaw.com> bcc Subject 12128/2009 Letter To EPA Docket Center A P_ aa6 _ 34a.` Taft I Deborah McCrea / Legal Assistant Taft Stettinius & Hollister LLP 425 Walnut Street, Suite 1800 Cincinnati, Ohio 45202-3957 Tel : 513 .381 .2838 " Fax : 513 .381 .0205 c.7 www.taftlaw .com / mccrea@taftlaw .com ctt w Internal Revenue Service Circular 230 Disclosure: As provided for in Treasury regulations, advice (if any) relating to federal taxes that is contained in this communication (including attachments) is not intended or written to be used, and cannot be used, for the purpose of (1) avoiding penalties under the Internal Revenue Code or (2) promoting, marketing or recommending to another party any transaction or matter addressed herein . This message may contain inafroermnaottioanntihnatteinsdeatdtorrenceiyp-iecnlti,enutsperiavnildegdeids,caltotsourrneeoyfwtohirskmpersosduacgteorare otherwise confidential . If you prohibited . If you received this transmission in error, please notify the sender by reply e-mail and delete the message and any attachments . LJ 2612_001 .pdf p. 2 Taft/ Taft Stettinius &HolGster LLP 425 Walnut Street, Suite 1800 / Cincinnati, OH 45202-3957 / Tel: 513.381 .2838 / Fax: 513.381 .0205 / www.taftlaw.com Cincinnati / Cleveland / Columbus / Dayton / Indianapolis / Northern Kentucky / Phoenix / Beijing ROeRT A 80.o7T $13-357-9838 b8otletafUaw.corn December 28, 2009 FEDERAL EXPRESS EPA Docket Center, MC 2822T U.S. Environmental Protection Agency EPA West, Room 3334 1301 Constitution Avenue, NW Washington, D.C. 20004 Re: Submission to IRIS and AR-226 Database For PFOAIPFOS : EPA-HQORD-2003-0016 To IRIS Database for PFOAIPFOS : In response to the Notice issued by USEPA on February 23, 2006, regarding USEPA's efforts to consider perfluorooctanoic acid ("PFOA") and per5uorooctane sulfonate ("PFOS") within the Integrated Risk Information System ("IRIS"), 71 Fed . Reg. 9333-9336 (Feb . 23, 200fi), we are submitting the following additional information to USEPA for inclusion in that review, and for inclusion in the AR226 database: 1 . Lin, C.-Y., et, et. aL, "Investigation of the Associations Between LowDose Serum Perfiuorinated Chemicals and Liver Enzymes in US Adults," Am. J. Gastroenterology (doi: 10.10381ajg.2009.707) (Dec. 15, 2009) ; and 2. Guruge, K. S., at al., "Effect of Perfluorooctane Sulfonate (PFOS) on Influenza A Virus-Induced Mortality in Female B6C3F1 Mice," 34 J. Toxicol. Sci. 687-91 (2009) . RAB :rndm Enclosures Very truly yours, 4'6"1 C . Robert A. Bilott 11587425 .1 December 28, 2009 Page 2 cc: Gloria Post (NJDEP)(wl encl.) (via U.S. Mail) Helen Goeden (MDH)(wl encl.) (via U.S. Mail) Lora Wemer (ATSDR)(wl enci.) (via U.S. Mail) p. 4 p.5 , American Journal of Gastroenterology - Investigation of the Associations Between Low-Dose Se... Page l of 11 The American Journal (it fC_JAS'JFR(.~~l\` T F k('-)LOGY Liver and Biliary Tract Slibject Category: I3verand Biliarv'frect Am JGaspvenferoladvance online publication 1.5 December 2oog; doi: to .io38/aj8.2oo9.7o7 Investigation of the .Associations Between Low-Dose Serum Perfluorinated Chemicals and Liver Enzymes in US Adults Chien-Yu Lin MD, MPH-L2-=, Lian-Yu Lin MD, PhDd-7, Chih-Kang Chiang MD, PhDd, WeiJie Wang MDA, Yi-Ning Su MD, PhD, KuanYu Hung MD, PhD&$ and Pau-Chung Chen MD, PhDa 'Department of Internal Medicine, En (hu Kong Hospital, Taipei County, Taiwan 2lnstitute of occupational medicine andIndustrial Hygiene, National Taiwan University College of Public Health, Taipei, Taiwan aschool of Medicine, HL Jen Catholic University, Taipei County, Taiwan 4Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan SDepartntent of Internal Medicine, Taoyuan General Hospital, Taoyuan, Taiwan 6Depattment of Medical Genetics, National Taiwan University Hospital, Taipei, Taiwan 7't'hese authors contribute equally to this work eCotorrespondence . Corrrspondence: Pan-Chung Chen, MD, PhD, Institute of Occupational Medicine and Industrial Hygiene, National Taiwan University College of Public Health, ft SyuW'6ou Road, Taipei iooss, Taiwan . irmaa : gchen aQntu.edu.tw Received ib May2oog; Accepted 17 November 2oog; Published online is December 2oog. Abstract OB,TEGTIVES: PerHuorinated chemicals (PFCs) have been largely used for years in a variety of products worldwide. However, the toxic effect of PFCs on exposure to the liver in the general population has not yet been determined. MBTHODS: In this study, 2,2i6 adults (i8 years of age or older) were recruited in a National Health and Nutrition Bxamination Survey (NHANES) in i999-aooo and 2oogzoo4 to determine the relationship between serum level of PFCs and the levels of liver enzymes. The data were adjusted for all other confounding variants. RESIILT3 : Afterperforming mathematical analysis, we determined when serum logperfluorooctanoic acid (PFOA) increases in one unit, the serum alanine aminotransferase (ALT) concentration (1J/1) increases by 1.86 units (95% confidence interval (CI) ; i.24-2.48; P=o.oo5), and the serum log-Y-glutamyltransferase (GGT) concentration (U/1) is o.o8 unit higher (9596 CI, o.o5-o.ii; P=o.oi9). The association between PFOA and liver enzymes was more evident in obese subjects, as well as subjects with insulin resistance and/or metabolic syndromes. When dividing the serum PFOA into quartiles in the fully adjusted http://www.nature.com/aiatioumaUvaou/ncurrent/fulUais2009707a .htnil 12/22/2009 p. 6 Aanerican'lournal of Gastroenterology - Investigation ofthe Associations Between Low-Dose Se... Page 2 of 11 models in subjects with a body mass index ;?-'30 k9/m2, the ALT level trend across the serum PFOA quartiles was significant (P-o.oo3) . CONCLUSIONS : On the basis ofthese data, we conclude that a higher serum concentration of PFOA may cause liver enzymes to increase abnorrnaity In the general population, particularly in obese individuals. Further studies are wan-anted to clarify the casual relationship between PFCs and these Ifver.enzymes . INTRODUCTION Perfluorinated chemicals (PFCs) consist of a 4-i4 carbon backbone and a charged functional moiety (primarily carboxylate, sulfonate, orphosphonate). PFCs are man-madecontemporary chemicals that have only been used in the last half-century. Until recently, PFCs have been considered to be biologically inactive. Human and wildlife monitoring studies have identified PFCs worldwide. This finding led to efforts to better understand the hazards that may be inherent in these compounds, as well as the global distribution of PFCs W. The two most widely known Ms are perfluorooctanoic acid (PFOA) and perIIuorooctane sulfate (PFOS), which belong to the 8-carbon backbone subgroup (t). PFOA (primarily ammonium salt) can be used as a surfactant and an emulsifier in the production of polytetrafluoroethylene as well as other fluoropolymers and fluorpelastomers. N-allryl substituted derivatives of PFOS have been used in a wide variety of industrial and consumer products including protective coatings for carpets and apparel, insecticideformulations, and surfactants. Other PM, like perfl~rononanoic acid (~P used as~ surfactants in the production of fluoropolymer polyvinylidene fluoride. In contrast, perftuorohexane sulfonic acid (PFHxS) is mostly used in carpet-treatment applications (g}: Although the major manufacturer ofPFOS, gM, has phased out ofproduction since 2002, the potential risk ofPFCs to humans still needs to be continually evaluated (g) . The possible routes of human exposure to PF(:s are currentlybeing investigated. Potential routes include contaminated drinking water, dust, food, food packaging, and cookware. Animal studies have shown that PM arewell absorbed orally but are poorly eliminated. PFCs mainly distribute extraceliularly. PFCs have a binding affinity for fl-lipoproteins, as well as albumin and liver fatty acid-binding protein. PFCs are not metabolized anddistributed through enterohepatic circulation to the serum and the kidney. However, PFCs aremainly distributed to the liver with concentrations being several times higher than serum concentrations (g). The half-life of serum elimination of PFCS in humans seems to be long. The longer the carbon chain length, the longer PFCs persistin the body. For example, pertluorobutaue sulfonate (a 4-carbon PFC) is eliminated, on average, in a little over i month in humans, whereas PFOA and PFOS (8-carbon compounds, refened to as C8 compounds) are eliminated in 3.8 and 5.4 years, respectively. However, PFHxS (a 6-carbon compound) is an exception to the rule as it is eliminated in 8.5 years (4). Exposure to PFUs at relatively high concentrations is associated with damage to liver function in animal models (5.,v. The hepatotoxicity of PFOS and PFOA has been linked to the functions of these compounds as peroxisome proliferator-activated receptor or (PPARa) agonist and thus, their ability to alter the expression ofgenes involved in peroxisome proliferation, cell cycle control, and apoptosis ($,%iW . In addition, otherPFCs have alsobeen shown to act as strongperoxisomai P-oxidation inducers a(1r1e-4n9o)t~cIlneahrulyrndaefnibneeidn.gIsn, tohcecucpaautsiaolnbaliopcohpeumliactailonm,escehvaernailssmtsudoifehsehpaatviecfatiolxeicdittoy aefsttearbleixsphoasudrefeitnoitPeFCg association between exposure to PFCs and adverse health effects (y%14,15,IM. Afew cross-sectional and longitudinal occupational studies have proposed a positive correlation among PFOA, serum lipid, and liver enzymes levels (L7AM. In a non-worker population, examination of PFOA exposure through contaminated drinking water showed the serum PFOA concentration (W. an insignificant correlation between abnormal clinical markers and The relationship between the serum PFC levels and liver enzymesin a nationally representative survey of adults has never been performed. We hypothesized that PFCs might have adverse effects on liver chemistry in the general US population accordingto the large scale data set of PFCs (ag-) and liver erizyme profiles released by US National Health and Nutrition Examination Survey (NHANES) performed between i999-2ooo and 2oog-2oo4. htto://www.nature.com/ais/iournal/vaon/ncurrent/fulUaiu2009707a .html I2/'22/2009 p. 7 . American Journal ofGastroenterology - Investigation ofthe Associations Between Low-Dose Se... Page 3 of 11 METHODS Study design and population The data were adopted from i999-2ooo and 2oog-2oo4 NHANES (PFCs were not measured in 2ooi- 2002 NHANFS). NHANES is a population-based survey designed to collect information on the health ainnsdtintuuttriiotniaolniozefdtchieviUliSanhUouSsepohpoulldaptoipounl. a'It'ihoensuarnvdeytodoabttaaairneapurbelpriessheendtabtiiavnenusaalmlpyl. eDeotfatihleedncoonn-tents of the NHANES i999-2ooo and 2oog-2oo4 are available atthe NHANES website 21 . Ipnretghneain9t9o9r-n2uorsoionagnadt 2th0e0t3i-m2e0o0f4thNeHeAxNamEiSn,attihoen,paarntidcwipearnetsrwaenrdeomollydearstshiganne1d8tyoeraercseiovfeaegex,amwienraetinoonts (11=10,224). Individuals were excluded based on the following criteria : individuals who had fasted <6 h at the time of the examination (n=i,8o2); individuals that were hepatitis B virus or hepatitis C virus carriers by serology (n =168); individuals in which data were not available for body weight, body height, educational attainment, or smoking habits (n=ii3) ; and individuals without seram tests for PFCs, liver function or the five components of metabolic syndrome (n=5,925) . Atotal of 2,216 participants were left for final analysis. Aflowchart ofalgorithm is shown in 'F~ru ~e~. In NHANES, a subset ofthe participants who received a morning fasting examination (re=i,ii4) had blood fasting and insulin levels measured. Insulin resistance status was determined for this subset ofparticipants. Fi2ure i. , I'4ow chart algorithm. HBV, hepatitis B virus; HM hepatitis C virus ; PFCs, perfluorinated chemimis. Ea fig= wd irer.nd (i2sKl potential causes In accordance with of elevated liver earlier studies a2 enzymes ,2~), in addition to chronic hepatitis viral infection (which was emxacrlkuedresdoffriormonthsitsorsetsudays),powteenctoinaslicdaeurseedseoxfceelsesviavteeadllciovheorlecnoznysmuemspt. Aiosno,bessmiotky,inigns,ualinndrienscirsetaansceed, saenrdum mbeotdaybomlaiscssiynndderxo(mBeMIa)r,e sintsruolnignprreesdiiscttaonrcse,oafnidncmreetaasbeodlliicvesryenndzryommee aacstipvoitteyn(tima)l,cwonefaolusnodceornssiindelirveedr function tests. The data were collected at all study sites by trained personnel using standardized procedures . Sociodemographic information such as age, gender, race/ethnicity, history of medication, and education level was recorded during the high household interview. The education level school diploma and below. The degree of was categorized as either above alcohol intake was determined by high school a diploma or a questionnaire and categorized into the following four categories : <i2 drinks, <6o drinks, <24o drinks, and >24o drinks by a year) serum .cSotmionkiinenglesvtealtsuasnwdaasssduebsdcirviibdeeddoinnttohaenquaecsttivieonsnmaoikreer,2(a_f~.oSremrerumsmiorkoenra,nodrthotaasl never smoked iron binding capacity were measured by the Beckman /Coulter LX2o analyzer . The transferrin saturation value was calculated using the following equation: (iron/total iron binding capacity) x ioo96. Weight and height were measured by standard methods and digitally recorded . The BMIwas calculated as weight (in kg) divided by the square of height (in m) . Tglhuecohsoemeaonsdtianssiuslimnodleevlelassdsievsisdmeednbty(2H2O.5M)Ai)s roefgianrsdueldinarseasissitmapnlcee,(iHneOxMpAe-nsIiRv)e,ianndedxre(ltihaebplerosduurcrtogoaftbeasal http :/Iwww.nature .com/ajl;/journaVvaop/ncurrent/fulVajg2O097Q7ahtm! 12/22/2009 p. 8 American`7ournal of Gastroenterology - Investigation ofthe Associations Between LowDose Se... Page 4 of 11 measure ofinsulin resistanee (g). The swum glucose level and the plasmainsulin concentration were determined by thesheeaxto&l.dFn-as_e_ e.n_z-ymatfcal method and the immunoenzymometric assay, respectively. The National Cholesterol Sclum3rnoorFrograw Third AiiiurYi; eatment Panel (W) has established guidelines for metabolic syndrome with modifications for the different s. eam People who were 18 years old and above were defined as victims of metabolic syndrome if they meet at least three of the following criteria: a waist measurement >88 cm for women and >102 cm for men; serum trlglyceride >,i.69 mm; serum high-density lipoprotein choksterol < .0i3 mM in men and <1.29111M in women; systolic blood pressure >,t3o mmHg or diastolic blood pressure >,-8$ mmHg or self-report of anti-hypertensive medications; and fasting glucose level >.-6.iomM or self-report ofanti-hyperglycemic medications. Three, sometimes four, blood pressun determinations were taken with a mercury sphygmomanometer by a physician using the right arm unless otherwise specified. The averaged systolic and diastolic blood pressure was obtained. The level of driglyoerides was measured enzymatically. levels of high-density lipoprotein cholesterol were measured after precipitation of other lipoproteins using a Hitachi model 704 analyzer (Roche Diagnostics, Indianapolis, IN). Assmsment of liver enzyme parameters Total bilirubin, alanine aminotransferase (ALT), aspartabe aminotransferase, Y-glutaznyltransferase (GGT), and alkaline phosphataae were the liver enzymes parameters available from NHANF.S. Bilirubin is mostly derived from the metabolism of hemoglobin. Increases in bilirubin are highly specific for diseases of the liver or bile ducts (gM. Aspartate aminotransferase and ALT are enzymes presented in liver parm,nchymal cells. Both of these enzymes are elevated during acute liver damage. Increased ALT activity has been used as a surrogate measure for the presence of liver disease in earlier population-based studies (gg,g3g4j. GGTis found on the ceII surface of all cells. Particularly high concentrations ofGGT are found in the liver, the bile ducts, and the kidney. GGT increases occur earlier and persist longer than alkaline phosphatase in cholestatic disorders W . In the study, we used total bilirubin, ALT, and GGT as markers ofliver enzymes. Serum totalbilirabin, GGT, and ALT levels were measured by enzymatic methods through automated biochemical profiling Beckman Synchron L82o). Total bilirubin was calculated in micromolar (p3o and GGT and ALTwere calculated as units/litre (U/1) . Assessment of PIRCs concentration Thirteen kinds ofPFCs are available in NHANES . However, in nine, over 7o% ofthe PFCs are below the limitofdetection. Therefore, we used serum samples ofPFOA, PFOS, PFHxS, and PFNA for analysis in this study . A briefsummary ofthe PFCs assessment (JoJ is as follows : the serum was diluted with o.i M formic acid without protein precipitation and a ioo ul aliquot of the serum was injected into a commercial column switching system for the determination of the concentration of the analytes on a Ci8 solid-phase extraction column . This column was placed automatically in front ofa C8 analytical high-performance liquid chromatography column for chromatographic separation ofthe analybes . Detection and quantification were done by negative-ion 'IurbolonSpray ionization tandem inass spectrometry and isotope-labeled internal standards. The limit ofdetection for PFOAwas 0.2ng/ml in the NHANES i9992ooo and o.ing/mi in the NHANF.S 2oo3-2oo4 data sets. For PFOS, the limitof detection was 0.2 and o.4ng/ml in the NHANES i999-2ooo and 2003-2004, respectively. The limit ofdetection for PFHxS was o.i and o.3ng/ml in i999-20oo and 2003-2004, respectively. ForPFNA, the limit ofdetection was o.2ng/ml in i999-2ooo and o.ing/ml in 2003-2004. For concentrations below the detection limits (o.4% for PFOA, 0.1% for PFOS, 2.3% for PFHxS, and 1.7% for PFNA), a value equal to the detection limit divided by the square root of 2 was used (am). Statistics PFCs concentration was expressed as the geometric mean with a 95% confidence interval (CI). Log transformation was performed for HOMAIR, serum GGT, and PFCs levels with significant deviation from the normal distribution before further aaalyses. All the log-transformed .data inthe study had a normal distribution and no significant outliers were found For linear regression models, we used an extended model approach for covariates adjustment of potential eonfounders. Model i adjusted for age, gender, and race/ethnicity. Model 2 adjusted for age, gender, race/ethnicity, life style (smoldng status, drinldng status, and education level), and measurement data (BMI, HOMAIR, metabolic syndrome, and iron saturation status) . To avoid "model-dependent association," the association was considered http ://www.nature.com/ajRfioumal/vaop/ncurrent/full/ajg2009707ahtmi 12122/2009 p. 9 Amp,rican Journal of Gastroenterology - Investigation ofthe Associations Between Low-Dose Se... Page 5 of 11 significant onlywhen it remained statistically significant in all models . Each PFC was modeled separately. Sampling weights that accounted for unequal probabilities of selection, over-sampling, and non-response and variance estimation accounting for complex survey design were applied to all analyses by the complex sample Survey module of SPSS i3.o for Windows XP (SPSS Inc. Chicago, IL). A mobile . examination center weight variable was created by assigning half of the 2-year weight for 1999-2ooo and assigning half ofthe 2-year weight for 2003-2004. RESULTS The basic demographic ofthe sample population is outlined in Table i. The studysample consisted of i,o76 men and i,i4o women. In accordance with an earlier NHANES study ao , the results indicate that males have a higher average concentration of PFOS, PFOA, and PFHxS than females. Hispanic Americans have lower mean serum concentrations (ng/ml) for these three compounds than non-Hispanic whites and non-Hispanicblacks . In addition, the concentration ofPFOS and PFNAwas higher in the more highly educated cohort . The four PFCs were moderately correlated withone another. PFOA and PFOS were most strongly correlated, with a Spearman correlatiori coefficient ofo.68 ; PFHxS and PFNA were the least correlated at o.24. Table i - Basic demoeranhics fthe sant* _subjects including geometric means (s.e.) of PIC concentrations . Ful l table Unadjustedmean liver enzymes across quartiles of PFCs (ng/ml) are shown in Table 2. The Serum ALT levels (Uh) inareased across quartiles of PFOAand PFOS (Pvalue <o.ooi and 0.030, respectively). Similar to ALT, the serum level of GGT (U /1) also increased across quartiles of PFOA and PFOS (P value o.oi2 and.o.oio, respectively).The serum total b0irabin level W increased across quartiles ofPFHxS and PFNA (Pvalue <o.ooi and o.oi4, respectively). Table 2 rr*.~.a:Uat.+d 1;M gnomes (s e ) across quartiles of PFCs. M able MOM a A summary of the association between serum concentration of log-PFCs (ng/ml) and liver enzymes after theadjustment for other potential covariates is listed in 1_'Able ,3. When the four PFCs were entered into thefull regression models separately, one unit increase in serum log-PFOA concentration was associated with a 1.86 unit (95% CI, i.24-2.48; P=o.oo5) increase in serum ALT concentration (U/1), a o.o8 unit (9&% CI, o.o5-o.ii; P=o.oig) increase in serum log-GGT concentration (U/1). PFOS associated with ALT, whereasPFNA was associated with total bilirubin with borderline statistical significance . The PFHxS htfip://www.nature.com/aiu/iournal/vaou/ncurrent/fulUaig2009707a .html 12/22/2009 p. 10 Araf:rican'Joumal of Gastroenterology - Investigation of the Associations Between Low-Dose Se... Page 6 of 11 concentration was not associated with liver function tests. When the four PFCs were entered into the full regression models at the sametime, one unit increase in serum logPPOAconcentration was associated witha 2.19 unit (95%CI, i.4-2.98; P=o.oo9) increase in serum ALT concentration (U/1), a o.isunit (9g96 CI, o.i1-o.19; P-o.ooi) increase in serum log-GGT concentration (Uh). Oneunit increase in serum logPFOS concentration was negatively associated with a i.o6 unit (95% CI, -i.33 to -o.y9; P=o.oos) decrease in serum total bilirubin concentration (pM) and was negatively associated with log-GGT concentration with borderline statistical significance. One unit increase in serum log-PFNA and log PFHxS concentration were associated with a o.75unit (P=o.oo4 and o.ooi, separately) increase inserum total bilirubin concentration (IM). .'.lw;wq~ - ~ Full ~_.~. hinear regression coefficients (s.e.) ofblood analytes (ALT and GGT) witha unit increase in log-PFOA in the different subpopulations of the sample subjects are shown in Table d. Subjects with iron saturation above So9d were excluded because ofthe small sample size (. N=ag) The association between ALT and PFOA was significant in the following subgroups of non-Hispanic Caucasians, individuals with a lower education level, higher BMI, non-smoking, lower alcohol consumption, higher HOMAIR, and subjects diagnosed as having metabolic syndromes. On the other hand, the association between GGT and PFOA was significant in subgroups of non-Hispanic white, higher MR, lower alcohol consumption, and higher HOMAIR log-PFOA coneentrations (ng/ml) in subpopulations of the sample sub.iects. Full tabie when dividing serum PFOA into quartiles in the fully adjusted models in subjects with BMI3go kg/m2, the adjusted levels ofALT are shown in Figure z. The trend in ALT levels across quartilesofserum PFOA was significant (P=o.oo3), whereas the trend in log-GGT was not significant. Eigm 2~ Theadjusted yeometricmeans of ALTacross qnartila of theserum MA conoentrations . The data are fuom the MY Opined model(ase, gender. race/etlsnidty, smokingstatus, drinlkins status, education level, metabolicsyndtome. andmm sataration status) in snl;jects with a BMI ;*-3o kg/ms. The boends in ALT level acinss thequartiles of the mum PFOA were silpfficant (A"o.oo3). ALT, alanineaminotranafense; BbII,body mass indeq PFOA,perfluorooctanoic acid. httn_//wvwv.nature.comJaivfiournai/vaop/ncurrent/full/ajg2009707a.html 12/22/2009 p. 11 A.nlerican Journal ofGastroenterology - Investigation ofthe Associations Between LowDose Se. .. Page 7 of 11 DLSCUSSION To our knowledge, our report is the first to link serum PFC levels to liver enzymes in a nationally representative survey. In this study, we showed that increased serum PFOA concentrations are associated with elevated serumliver markers . PFOS and PFOA show liver tox;city in rodents and non-human primates (5,,a). Anumber of short-term studies focusing on rats and mice have shown that PFOS and PFOA are capable of inducing peroxdsome proliferation via activation of PPARo( (8,Q,io,ix). There is strong evidence to support the idea that PFOA-induced liver toxicity occurs through a PPARa agonistic mode of action in rodents (g). However, adverse hepatic effects of PFOA still exist in PPAR null mice (31). These findings imply that PFOA may exert its toxic effects through both PPARa as well as other alternative pathways. As the key events of hepatic toxicity induced by PFOS are not consistent with a PPARa-agonistic model, the relevance of PPARot-induced toxicity dependent on dose-response in humans is a current scientific debate (6,32). Some studies have used other PFCs that have been conducted in mice (ii,U). All of the compounds tested induced hepatomegaly and peroldsomal S-oxidase activity. The potency of PFCs to cause hepatomegaly and peroxisomal P-oxidase activity is shown as follows : PFNA>PFOA>perfluoroheptanoic acid>pertluorohexanoic acid. These results indicate that thelonger the perfluoroalkyl chain of the PFC, the higher the accumulation of the compound in the mouse liver. In some occupational epidemiology studies, the association between PFCs exposure and abnormal liver function tests could not be established (1&1-4,1&J16 . Only one cross-sectional occupational study (17 observed a positive relationship between serum PFOA and GGT. In the meantime, one longitudinal occupational study (LS) revealed that serum PFOAis related to total bilirubin (o.oo8mg/dl dedine/i,oooppb) and serum aspartate aminotransferase (o.35units increase/i,oooppb) . Astudywas conducted to testify whether PFOA affected the non-worldng in a community located near a fluoropolymer production facility using either hematologic or biochemical clinical markers (iq) . In this non-occupational population, the median serum PFOA was 354 ng/ml (an interquartile range, i84-5n ng/ml) . This serum concentration of PFOA is higher than the level in the general US population (median of 4.4 ng/ml). However, there was no significant correlation between abnormal clinical markers and . serum PFOA concentration in that study. In our study, serum PFOA was associated with a change in both ALT and GGT, but not other PFCs. However, the potential biological significance between PFOA and liver enzymes is small and subclinical in this general population . As PFOA are metabolically inert, it is difficult to detect the same metabolic effect in the low exposure group of the general population and in the occaipational studies presented with a high concentration level. Onepossible explanation is thebias of the "healthy worker effects," which states that severely ill or disabled people are more susceptible to PFOA exposure and excluded from employment. Another possible explanation is that the dose-response effects of pFOA on liver enzymes may notbe a linear relationship in humans . PFOA exerts the maximal effects at a lowdose already and no further consistent or potentially relevant clinical changes occur at an even higher level. In our study, the association between PFOA and liver enzymes was more evident in subjects suffering from obesity, insulin resistance, and the metabolic syndromes. However, there was no significant difference in regards to serum PFOA concentration between these groups. Studies of gene expression profiles in rat livers treated with PFOA showed that the largest category of induced genes are those involved in metabolism and transport of lipids, particularly fatty acids (3a,gg). An increase in lipid httn://www.nature.cotn/aig/ioucnal/vaon/ncurrent/full/aig2009707a .html 1 717.1717.009 p. 12 American-Journal of Gastroenteroiogy - Investigation ofthe Associations Between LowDose Se... Page 8 of I 1 .droplets due to alteration in lipoprotein metabolism after exposure to PFOA have also been observed (W. In obese individuals, a liver loaded with fat will increase liver enzymes and become insulin resistant (AM . Therefore, it is possible that PFOAmay further increase the 5ne fatty add accumulation in sabgec<s with hepatic steatosis. Furthermore, there might be some synergistic efiects on hepatotaacicity between PFOA and obesity. PFOA is primarily distributed to the liver in rats (g). In the liver, multiple proteins from the cytosol, nuclei, and mitochondria fractions are capable of specifically binding PFOA (3!7). Another possible explanation is that in obese subjecb with hepatic stestosis, intrahepatie PFOA accumulation might be higher than non-obese subjects despite the same serum level. With the higher accumulation of PFOA in the liver, the effect of PPOA on liver is more evident. We found that the association between PFOA and liver enzymes was more evident among non-smokers and those with lower alcohol consumption. One possible explanation is that the efiect of PFOA on fiver enzymes is much weaker than the affect of alcohol and tobacco smoke. When considering the hepatoto:dc effect of PFOA in the smoking or higher alcohol consumption population, the trend is too small to become statistic significant. Alternatively, it is also possible that the association between PFOA and smoking tobacco or drinking alcohol is opposite to the possible synergistic effect between PFOA and obesity mentioned above . There are several limitations of our study. F':rst, the cross-sectional design does not permit any causal inference . Second, we did not include other environmental chemicals, which maybe important covvariates or explanatory variables for the outcomes of our study. Third, we did not take into account any medications that may cause elevated ALT or GGT. Fourth, a common physiology could influence both serum PFCs and liver enzymes rather than exposure affecting outcome. Finally, the status of the liver tissue was not available to determine hepatic steatosis, inflammation, or fibrosis . In conclusion, using the NHANES data from the US adult population, we found that a higher serum concentration of PFOA was associated with elevated liver enzymes. These findings provide clues to the adverse effects of low-dose PFOA in humans. Although the potential biological significance between PFOA and liver enzymes is small and subclinical in the general US population, our data suggest that it would be prudent to monitor the liver enzymes of people with low level exposure of PFOA, particularly in subjects who are obese . Further studies are needed to confirm thesefindings and to clarify whether these associations are causal. CONFLICT OF INTEREST Guarantor of the article: Pau-Chung Chen, MD, PhD. Specific author contributions: Designed the study, drafted the article, and inte.rpretated the data: Chien-Yu Lin; designed the study's analytic strategy and helped conduct the literature review: ManYu lin; helped conduct the literature review: Chih-Kang Ghiang, Wei-die Wang, and Yi-Ning Su; helped conduct the literature review, approved the analytic strategy, and approval of the final version: Kuan-Yu Hung; contributed to the study's conception, reviewed the study, and final approval of the version to be published : Pau-Chung Chen. Financial support: This was an investigator-initiated unfunded study. All authors had access to the data and the statistical analysis report. Each author approved the final article and attested to the validity of the results. Potential competing interests : None. STUDY HIGHLIGHTS partner ofACORA, HINARI, Oxut$ n1A8P, QvnReJ and COUNTER hnp://www.nature.com/ajg/jouraaUvaop/ncurrent/fulUaig2009707a .html 12/22/2009 p. 13 American Journal of Gastroenterology - Investigation ofthe Associations Between LowDose Se... Page 9 ofl l References i. Houde M, Martin JW, Letcher RJ et al. Biological monitoring of polyfluoroalkyl substances : a review. Environ Sci Technol 2oo6 ;4o:3463-3473~ I Article I Pubm I Ch ernPort I 2. Prevedouros K, Cousins rr, Buck RC et a1. Sources, fate and transport of perfluorocarboxylates . ~ Environ Sci Technol 2oo6;40 :32-44~ I ~ I PuM r ChemPort I 3 . Lau C, Anitole K, Hodes C et al. Perfluoroalkyl acids: a review of monitoring and toxicological flndings. Toxicol Sci 2007;99:366--394 . I Arti1 I Pued I ChemPort I 4. Olsen GW, Burris JM, Ehresman DJ et al. Half-life ofsermn elimination of perfluorooctanesulfonate, perfluorohexanesulfonate, and pertluorooctanoate in retired fluorochemicai production workers . Environ Health Perspect 2oo7;1xS:i2981305 .1 Pu M I CheinPort I 5. Butenhoff J, Costa G, Elcombe C et al. Toxicity of ammonium perfiuorooctanoate in male cynomolgus monkeys afteroral dosing for 6 months . Toxicol Sci 2002;69:244257. I Article I lPubMed I ChemPort I 6. Seacat AM, Thomford PJ, Hansen KJ et al. Sub-chronic dietary toxicity ofpotassium perQuorooctanesulfonate in rats. Toxicology 2oo4;i83:u7-r3i . I Ardde 7. Seacat AM, Thomford PJ, Hansen KJ et al. Subchronic toxicity studies on perfluorooctanesulfonate potassium saltin cynomolgus monkeys. Toxiool Sci 2002;68 ".249264. I Article I IP~bMed I Chem Qrt I 8. Intrasuksri U, Rangwala SK O'Brien M et al. Mechanisms of peroxisome proliferation by perfluorooctanoic acid and 197. 1 Article I BmbMed I endogenous CbemPort fatty I acids. Gen Pharmacol 1998;31: 187- 9 . Sohlenius AK, Eriksson AM, Hogstrom C et al. Perfluorooctane sulfonic acid is a potent inducer of peroxisomal fatty acid b-oxidation and other activities known to be affected by peros;some proliferators in mouse liver. Pharmacol Toxicol 2003 ;72:90-93~ I Article io. Klaunig JE, Babich MA, Baetcke KP et al. PPAR agonist-induced rodent tumors : modes of action and human relevance. Crit Rev Toxico12oo3.33:655-780 " I Article I PubM I ChemPort I u. Kudo N, Bandi N, Suzuld E et al . Induction by perfluorinated fatty acids with different carbon chain length of peroxisomai beta-oxidation in the liver of rats. Chem Biol Interact 200o ;124 :ii9132~ ! Arfic-Is I b d I CheinPort I x2 . Kudo N, Suzuki-Naka3ima E, MitsumotoA et al. Responses of the liverto perfluorinated fatty acids with different carbon chain length in male and female mice: in relation to induction of hepatomegaiy, peroxisomal beta-oxidation and microsomal i-acylglycerophosphocholine acyltransferase. Biol Pharm Bu112oo6 ;29:i952-1957~ I Article I PubM I ChenzPort I r3 " Grice MM, Alexander BH, Hoffbeck R et al. Self-reported medical conditions in pertluorooctanesulfonyl fluoride manufacturing workers. JOccup Environ Med 2007 ;49:722- http://www.nature.com/ajg/journaUvaop/ncurrent/fulVaifz2009707a.html ' 12/22/2009 p. 14 AmericanJournai of Gastroenterology - Investigation ofthe Associations Between LowDose ... Page 10 of 11 729. I Ards~ I P!l~.4~ I CheunPort I 14. Olsea GW, Burris JM, Burlew mm et al. Epidemiologic assessment of worker serum perfluorooctaneaulfonate (PPOS) and perfluomoctanoate (PFOA) concentrations and medical surveillance examinations.J Occnp EnuironMod 2003;45:26o- 270.1 Arb-ch I PubMed I OLUDEQd I 1,5" Olsen GW, Burlew MM, Marshall JC et al. Analysis of episodes of care in a perEluorooctanesulfonyl fluoride production facility.JOccup EnvironMed 2004;46:837- 846.1 Article I P"t~t I MORE= I . 16. OpelrsfelnuGorWo,ocZtoabneolatLeL(pAFsOsAe)sscmoencnetnotfralitpiiodn,sheipnaftliuco,raonchdet,hmyircaolidprpoadruacmteitoenrwsowriktehrsse. rInutmArch Occup Environ Health 2oo7;8i :23i-246. I Artick I ~ I ChemPort I 17. Sakr CJ, Kreclanann IQi, Green JW et al. Cross-sectional studyofHpids and liver enzymes related to a serum biomarker of exposare (ammonium perfluorooctanoate or APPO) as part of a general health survey in a cohortof occupationally exposed workers . JOccup Enuiron Mod 2007:49:io86-io96. I ~. I P-UhM I CheYnPort I 18. Sianlwoo" rCJk,erLsewointahrdocRcCu,paKtrieocnallmieaxnpnosKuireettoala.mLmonogniituudminpaelrfsltuuodryooofdsaenroautmel.iOpicdcsuapnEdnlviiverroennMzeymdes 2007;49:872-879 . I Ardylg I ChexnPort I i9. Emmett EA, Zhang H, Shofer FS etal. Community exposure to perfluorooetanoate : relationships between serum levels and certain health parameters. J Occup 8nviron Mod 2006;48:771- 779" I Axdsds I PubM.ed I Chemport I 2o. CfarloamfatthAeMN,atWioonnalgHLeYa,lKtuhkalnednyNiuktrZiettioanl.EPxoalmyitlnuaotriooanlkSyulrcvheeym(icNaHlAsNiEn Sth)e2U0.0S3. p-o2pou0l4atainond: data comparisons with NHANF.S i999-2ooo . Environ Health Perspect 2oo7;iss:t,596- i6o2. 1 ~1 aemP011 I 21. CDC (Centers for Disease Control and Prevention) . National Health and Nutrition Examination Survey Homepage. 2003 . Available: hM :jfwww cdegav[nch$/nhanes-htm Accessed i March 2009 . 22. ClarkJM, Brancati FL, DiehlAM. The prevalence and etiology ofelevated aminotransferase levels in the United States .Am J Gastroenterol 2003:98 :96o- 967. I Ardck 1 I mI ~e~Qrt I 23. Ruhl CE, Everhart JE. Determinants ofthe association ofoverweight with elevated serum alanine aminotransferase activity in the United States. Gastroenterology 2oo3;i24:71- 79- I Miele I FyhM I W-1 I Chemport I 24 . Ioannou GN,Weiss NS, Boyko EJ et al. Contribution of metabolicfactors to alanine aminotransferase activity in persons with other causes of liver disease. Gastroenterology 2005;128:627-635. I Article I ~bM-ed I Chem~'o~ I 25. Weitzman M, Cook S, Auinger P et al. Tobacco smoke exposure is associated with the metabolic syndromein adolescents . Cir=lation 2oo5;iis:862-869. I Article I 1PubMed I ChenaPort I 26. Wallace TM, Levy JC, Matthews DR Use and abuse of HOMA modeling. Diabetes Care 2004;27:3487-1495~ I A19SIC I ~ I Uff 27. Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection . Evaluation, and treatment ofhighblood cholesterol in adults (Adult Treatment Panel III). JAlVtA 2ooi;288:2486-249y. I Article I PubM I M11 28. Sedlak TW, Snyder SH. Bilirubin benefits: cellular protection by abiliverdin reductase antioxidant cycle. Pediatrics 2oo4;i13 :i776-i782 . I Article I PabMed 29 . Whitfield JB. Gamma glutamyl transferase. Crit Rev Clin Lab Sci 2ooi ;38:26g- 355-IArfick I ~I ent4~ 1 30 . Kuklenyik Z, Needham LL, CalafatAM . Measurement of 18 perfluorinated organic acids and amides in human serum using on-line solid-phase extraction. Anal Chem 2oo5;77:6o856ogi. I Artide I ~I ChemPort I 31 . Yang Q, %ie Y, Alexson SE et al . Involvement of the peroxisome proliferator-activated receptor alpha in the immunomodulation caused by peroxisome proliferators in mice . Biochem Pharmacol 2002;63:i893-r9oo. I ~e_ I PubM-~ I ChemPort I http://www.nature.com/ajgljournaUvaop/ncurrent/fulVajg2009707a .html 12/2212009 AWrican Journal ofGastroenterology - Investigation of the Associations Between Low-Dose ... Page 11 of 11 32. Shipley JM, Hurst CH, Tanaka SS et al. Trans-activation of PPAR and induction of PPAR target genes by pertluorooctane-based c. hemicals Toxicol Sci 2004;8o:15ii6o. I A- r&le I PubMed I Che.mPort I 33" Guruge KS, Yeung LWY, Yamanaka N et al. Gene expression profiles in rat liver treated with perfluorooctaonic acid (PPOA). Toxicol Sei 2oo6;89 :93-io7- I Article I EWM.qd 34~ Rosen MB, Thibodeaux JR, WoodCRet al. Gene expression profiling in the lung and liver of PFOAexposed mouse fetuses. ToxEcology 2007.239:15-33~ I Article I PuMed I ChenaPort I 35~ Peters JM, Hennuyer N, Staels B et at. Alterations in lipoprotein metabolism in peroxisome proliferator-activated receptor alpha-deficient mice.JBiol Chern i99T272 :27307-273i2. I Article I Me t IX I ChemPort I 36 . 'I51g H, Moschen AR. Insulin resistance, inflammation, and non-alcoholic fatty liver disease. Trends Endocrinol Metab 2oo8;i9:371-379- I ~~e I Me I QLeniiPo~ I 37. Han X, Kemper RA, Jepson GW. Subcellular distribution and protein binding of pertiuorooctanoic acid in rat liver and kidney. Drug Chem Toxicol2oos;28 :i97-209.1 EUbMW I ChemPort I Acknowledgments We thank the many people who have contributed to the National Health and Nutrition Examination Survey data we have examined, including all of the anonymous participants in the study. We are particularly grateful to Antonia M Calafat, who carried out the laboratory assays of PFC concentration at the Division of Environmental Health Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention. htrip://www.nature.com/aig/ioumal/vaonlncuirentlfiilllai22009707sLhtml v)mnnno p.. 16 p. 17 The Journal ofToxicological Sciences (J. Toxicol. Sci.) Vol.34, No .6, 687-691, 2009 687 Letter Effect of perfluorooctane sulfonate (PFOS) on influenza A virus-induced mortality in female B6C3F9 mice Keerthi S. Guruge+, Hlrokazu Hikono2, Nobuaki Shimadal, Kenji Murakamis, Jun Hasegawa', Leo W.Y Yeung4, Noriko Yamanaka' and Nobuyoshi Yamashita" 'Safety Research Team "Research Teanr forAdvanced Biologicals, jResearrh Team for Viral Diseases, National Institute ofAninral hfeahh, Kannondai 3-1-5 Tsularba, Ibaraki 305-OSS6, Japan 4National Institute ofAdvanced Industrial Science and Technology, Onogaxsa 16-1, Tsuluba. Ibaraki 305-8569, Japan (Received July 23, 2009; Accepted August 16, 2A09) ABSTRACT - Recent studies showed that per$uorooctane sulfonate (PFOS) affects the mammalian immune system at levels reportedly found in the general human population. It has been demonstrated that exposure to imntunotoxic chemicals may diminish the host resistance of animals to various pathogenic challenges and enhance mortality. Therefore, the current study was carried out to characterize the effect of a 21 day pre-administration of zero, 5, or 25 kg PFOS/kg bw/day in female Li6C3Fl mice on host resistance to influenza A virus infection . At the end of PFOS exposure, body/organ weights did not significantly change whereas PFOS distribution in blood plasma, spleen, thymus and lung was dose-dependently increased . PFOS exposure in mice resulted a significant increase in emaciation and mortality in response to influenza A virus . The effective plasma concentrations in female mice were at least several fold lower than reported tnean blood PFOS levels from occupationally exposed humans, and fell in the upper range of blood concentrations of PFOS in the normal human population and in a wide range of wild animals. Hence . it should be important to clarify the precise mechanism(s) for excess mortality observed in the high dose group . Key Words : PFOS, liioaccumulation, Immunotoxic, Host resistance, Mortality INTRODUCTION Perftuorooctane sulfonate (PFOS) is one of the newly listed Persistent Organic Pollutants (POPs) under the Stockholm Convention. Recent studies have demonstrated that exposure to perfluorinated alkyl substances can modulate rodent humoral and cellular immune functions (Dewitt et al., 2008 ; Peden-Adams et al., 2008) . Suppression of the primary antibody response was reported in mice exposed to PFOS, at serum concentrations 14 times lower than the average concentrations of occupationally exposed workers, and in the upper range of levels reported for the general population (Peden-Adams et al ., 2008) . Exposure to xenobiotics and the resultant alteration of immune function may effect in a modification of an organism's ability to resist infectious disease . Several studies have shown that persistent pollutants enhance susceptibility to viral, bacterial, parasitic and neoplastic challenges (Burleson et al., 1996). However, no studies have been done to determine the level of susceptibility to pathogens that altered due to PFOS exposure. The rodent influenza virus model has been widely used to determine the itnmunotoxic effects of persistent chemicals such as dioxins on viral host resistance (8urleson et al., 1996 ; Nohara et al., 2(?02) . This study represents the first preliminary investigation of increase mortality caused by influenza A virus infection in mice pre-exposed to PFOS . MATERIALS AND METHODS Animals Female B6C3F1 mice were obtained from Japan SLC Inc . (Shizuoka, Japan) . They were acclimatized to the environment for at least t week before both the preliminary virus test (9- to 10-weeks old) and PFOS/virus administration (6- to 7-weeks old). Mice were housed in HEPA filtered disposable cages (Inocage ; Oriental giken, Tokyo, Japan) in a light- (12 hr light-dark cycle), tempera- Correspondence : Keerthi S. Guruge (E-mail: guruge@affrc.go.jp) Vol. 34 No . 6 p. 18 688 K .S . Guruge u al. ture- (22 * 2C) and humidity- (50 f l0/a) controlled Bio Safety Level-2 facility at the National Institute of Animal Health (NIiAH), Japan. They received food and water ad libiturn. Bedding, food, and water were changed weekly. All procedures used in this experiment were reviewed and approved by the biosafety, animal care and ethical committees of NIAIH, Japan. Influenza virus infection Mouse adapted influenza virus, A/PR18134 (H 1 N i ), (obtained from Dr. Hideki Hasegawa of National Institute of Infectious Diseases, Japan) was used as the infectious agent. Aliquots (0.1 ml) of 1 x 10' plaque forming units (pfu) of virus were prepared and stored in -80C . Virus dilution with phosphate buffer saline (PBS) was carried out immediately prior to use . Mice were anesthetized by intraperitoneal injection (i.p.) of Avertin and intranasally infected with 30 pt of virus suspension . As a preliminary test, susceptibility to the current passage of influenza A virus was examined using 9- to l0-week old female B6C3F1 mice in order to achieve a known lethal dose in control animals. Virus diluted at eight concentrations between 6.25 and 800 pfu was tested in 7 mice for each concentration. The general appearance and weight of infected mice was evaluated for twenty days. The preliminary test showed that between 100 and 200 pfu of virus inoculation caused approximately 40/. mortality (data not showing). No mortality was observed using lower virus regimes tested on female mice . Therefore, the effect of PFOS on the mortality was assessed using 21 days of PFOS exposure followed by 100 pfu of virus infection . PFOS exposure The potassium salt of PFOS (CAS number 2795-39-3) was acquired from Fluka Chemical (Steinltcim, Switzerland). Stock solution was prepared in Milli-Q water containing 0.5% Tween 20 at a concentration of 0.2 mg/ml. Dosing solutions were prepared weekly by serial dilution . Control mice received Milli-Q water containing only 0.5% 'ltween 20 . Mice were dosed by gavage for21 days with either vehicle, 5. or 25 jig PFOS/kg body weight/day prior to virus inoculation. Experimental design for measuring PFOS effect on host resistance Mice (6- to 7-wceks old) were randomly divided into 30 animals/dose group and 6 animals/cage . They were weighed weekly during the PFOS dosing period . At the end of a 21 day PFOS exposure, 3 animals from each group were randomly sacrificed by pentobarbital overdose, and blood samples were collected via cardiac punc- ture using heparinized syringes. Blood plasma was separated by centrifugation at 3,000 rpm for 10 min and kept at-20C until PFOS residue analysis . Liver, kidney, lung, spleen and thymus were collected and weighed. Organs were kept at -20C until PFOS residue analysis . The remaining mice were anesthetized by i.p. injection of Avertin and intranasally infected with 100 pfu (in 30 pl of PBS) influenza A virus suspension. Following virus inoculation, mice were observed for health conditions and mortality twice a day and weight was measured daily for twenty days. PFOS residue analysis The blood plasma samples were thawed at room tem- perature, fortified with 13C-PFOS (Wellington Laboratories, Guelph, Canada), and extracted. Analysis of PFOS in plasma (0 .2 ml) was carried out using the ion pairing method (tiuruge et al., 2005: Yamashita et al., 2004). The individual or pooled organ portions (approximately 0.1 g) were thawed, fortified with UC-PFOS and then 1 ml of Milli-Q water was added followed by homogenization with a Micro Homogenizing System (Tomy Seiko Inc., Tokyo, Japan) at 3,000 rpm for 3 min. The entire homogenate was used for the extraction similar to the procedure used with plasma. Matrix recoveries of "C-PFOS in plasma and organs were 93 t 8 (mean t S.D.) and 96 f 8% . r. espectively Thc limit of quantifications for plasma and organs was 0.2 ng/ml and 0.25 ng/g wet weight . The concentrations of PFOS in the samples were not corrected for recoveries. RESULTS At the end of the PFOS exposure, the measured organ masses were not significantly different among PFOStreated and control groups (Table 1).The distribution of PFOS in the body increased with repeated gavage (Table 2). At the end of PFOS exposure, mean plasma concentration was significantly higher in 25 Itg/kg exposure group compared to that in the 5 ttglkg exposure group. PFOSdistribution in the various samples ranked as follows: lung = plasma > spleen = thymus . Fig. 1 shows the change in body weight during PFOS gavage and virus infection. There was no significant weight change due to PFOS exposure alone. The mean body weights of both PFOS-treated groups have greater decreasing tendency compared with the control group from post-infection days 4 to 11, however, statistically significant differences were observed only at day 9 (P = 0.04 for 5 ltg/kg group and P = 0.02 for 25 pg/kg group, Dunnett's test). Vol. 34 No. 6 p. 19 689 PFOS effect on virus-induced mortality in mice Table 1. Mean body (g) and organ mass- of adult B6C3 F 1 female mice dosed with PFOS for 21 daysb Dose Body weight 5pleea_y Thymus b Liverb Kidneye Control 20 .4 * 0.95 0.34 t 0.04 0.36 * 0.07 3 .7 0.24 1 .1t0.04 5 ytg/kStbw/day 20 .1+0.45 0.31 f 0.03 0.32 t 0.05 3 .7 t 0_ 13 1 .1 +0.03 25 Kg/kg/bw/day 20_2 t 0.33 0.31 * 0.02 0.26 f 0.03 3 .9-+0.27 1 .2 * 0.03 "Organ mass = organ weight (g)/ body weight (g) x 100 . b Data are reported as mean } S.E. (n = 3) . Lung" 0.79t0 .13-~ 0.67 1- 0 .08 0.67 t 0 .03 The survival rate at each time point of female B6C3F1 mice exposed to PFOS followed by infection with virus is shown in Fig . 2. The mean survival time based on 20 days observation period was 14.1, 13 .2, and 11 .4 days in the control, 5 and 25 pg PFOSIkg bw/day groups, respectively. There was no significant difference in survival time among the three groups (P > 0 .05,1Caplan-Meier log-rank test) . The survival rate of the mice on day 20 after virus infection was 46%, 30% and 17% in the control, 5 and 25 Kg PFOSIkg bw/day groups, respectively. A significantly dose-dependent increase in mortality was observed with PFOS exposure (P = 0.014, Cochran-Amiitage trend test) . Additionally, the eventual survival rate differed significantly between 25 lig PFOS/kg bw/day group and control group (P = 0 .035, logistic regression Wald test). There was no statistically significant difference between 5 p.g PFOS/kg bw/day group and control group (P = 0 .28, logistic regression Wald test) . DISCUSSION Exposure to higher levels of PFOS often produces significant reduction in rodent body weight and this may influence their immunological outcome. Therefore, we have selected only two PFOS doses (5 and 25 jig PFOS/ kg bw/day) for this experiment to try to reproduce envi" ronnuntally existing blood levels in humans and wildlife . In our exposure regime, there was no significant change in body weight observed among all three mice groups at the end of PFOS exposure . Likewise, no significant change in organ mass was found for immune-responsive organs . The results for body and organ weight gain were similar in both male and female mice exposed to PFOS at compatrable exposure levels (Peden-Adams et al., 2008). hievertheless, the body weight reduction during virus infection was clearly greater in PFOS dosed mice compared to the controls. Therefore, PFOS accumulation is likely superimposed on viral illness in mice to increase mortality. The mean plasma levels of PFOS at the beginning of virus infection were 189 14 and 670 3: 47 ng/ml wet wt. in 5 and 25 jig PFOS/kg bw/day exposed groups, respec- Table2. Mean f S .D . concentration (wet weight) in , blood plasma and organs in adult B6C3F! _,. . .female mice exposed to PFOS Sample type ^y Dose Concentration* - Plasma control 2 .1 10.3 5 ug/kg bwrday 189 f 14 25 Ng/kg bw/day 670 f 47 Spleen b Control <0.25 5 ugikg bwiday 84 25 ltg/ltg bw/day 260 Thymus Control -------<0.25. . 5 ug/kg bw/day 60 t 5? 25 hg/kg bw/day 260 f 68 Lung Control 1 .2 * 0 .2 5 wg/kg bwlday 190+65 25 pg/kg bw/day 970 t 145 ,Concentrations in plasma are given in nglml, in other organs are given in ng/g. Three mice for each dose were analyzed . ~ Pooled samples. tively (Table 1) . Our data are similar to recently published serum levels in female B6C3Fi mice exposed to a comparable PFOS exposure regime (Peden-Adams et al., 2008). The PFOS concentrations found in mice plasma were within the range of those reported in sera for occu- pationally exposed workers in Decatur, Iowa, USA (range 145 - 3,490 ng/ml wet wt .), Antwerp, Belgium (mean : 1,480 ng/ml), adult donors (range 4 - 1,656 ng/ml wet wt .) in the USA and the upper level found in the normal population in China (Olsen et al., 1999, 2003 and 2007 ; lin et a1., 2007) . Several studies reported that concentrations of PFOS in the blood of wildlife were similar to our data for mice treated with 5 and 25 ttg PFOS/kg bw/day. In the late 1990's, mean PFOS concentrations found in blood of ringed seals and bottlenose dolphins were 242 and 143 nglml wet wt. in European waters (Kannan et al., Vol. 34 No. 6 690 K .S . Guruge er aL p. 20 Days Fig. 1 . Effect of PFOS on mean body weight in mice challenged with influenza A virus. Adult female B6C3FI mice dosed with 0, 5 and 25 Kg PFOSJItg bwJday for 21 days and infected intranasally. Number of animals at day 0 infection: control= 24, 5 hg PFOSIkg bw/day m 23. 25 pg PFOS/kg low/day = 24 . ' Significantly different (P < 0 .05, Dunnen's test) from the control group. Standard deviation of body weight during whole experiment was 0.30-3 .8 .0.25-5 .01 and 0.37-4 .29 for mice dosed with 0, 5 and 25 pg PFOSIkg bw/day, respectively. 100 80 > it 60 40 20 0 0 2 4 6 8 10 t2 14 16 18 20 Day after infection Fig. 2. Effcct of PFOS on host resistance to influenza A virus . Adult female B6C3F1 mice dosed with 0, 5 and 251tg PFOS/kg bw/ day for 21 days and infected intranasatly. Number of animals at day 0 infection : control = 24, 5 ltg PFOS/kg bw/day = 23, 25 pg PFOSJkg bwlday = 24. ' Significantly different (P < 0.05, logistic regression Wald test) from the control group . 2001, 2002) .The PFOS concentrations in plasma of bottienose dolphins from the Gulf of Mexico and the Atlantic Ocean were ranged from 46 to 3,073 ng/g wet wt (Houde et al ., 2005). Interestingly, a significant association between infectious disease and elevated PFOS con- centration in the livers of sea otters has been described (Kannan et al., 2006) . As shown in Fig . 2, we observed that PFOS administration resulted in a dose-dependent enhanced mortality. In particular, 25 ttg PFOS/kg bw/day exposed group exhibited a significantly reduced survival rate compared to the control group. It is important to note that female mice treated with similar PFOS levels to this study had suppressed plas- ma IgM antibody production (Peden-Adams et al., 2008). Numerous components of the inunune defense system such as cytotoxic T cells, NK cells and other hutnoral responses are activated and play an important role during a viral infection . Hence, any alteration of this mech- Vol. 34 No. 6 p. 21 691 PFOS eftect on virus-induced mortality in mice anism may contribute to suppression of viral clearance caused by exposure to PFOS . It has been suggested that slight alterations of several immunological functions may together result in significant immunosuppression that can be detected as an increased susceptibility to infectious diseases and measured by host resistance models (Burleson et al ., 1996). Therefore, the mechanism(s) of enhanced mortality may be related to several important immunological functions which response to viral clearance. In conclusion, this is the first study that shows the effect of PFOS on host resistance to a pathogen in laboratory animals. Our results suggest that PFOS accumulation may associated with mortality in influenza virus-infected female mice. It will be essential to examine numerous immunological endpoints before concluding that PFOS accumulation directly leads to an alteration in host resistance to pathogens in animals. ACKNOWLEDGMENTS This study was supported in part by a Grant-in-aid from the Global Environntent Conservation Research Fund to Dr. KSG by the Environmental Ministry of Japan (Year 2004-2008) . The authors thank Dr. Hideki Hasegawa of National Institute of Infectious Diseases, Japan for providing mouse-adapted influenza virus . Mr. Hitoshi Ohashi and Mr. Akira lino of NIAH, Japan are thanked for their invaluable assistance during the experiment. Prof. Paul Lam o City University of Hong Kong is acknowledged for providing postgraduate studentship to LWYY_ REFERENCES Burleson, G.R ., Lebrec, H., Yang, Y.G ., lbanes, J.D ., Pennlngton, K.N . and Bimbaum, L.S . (1996): Effect of2,3.7,8-tetrachlorodibenzo-p-dioxin (TCDD) on influenza virus host resistance in mice. Fundam . AppL Toxicol., 29, 40-47. Dewitt, J.C ., Copeland, C.B ., Strynar, M.J . and Luebke, R,W. (2008): Perfluorooetanoic acid-induced immunomodulation in adult C57BL/6J or C57BL/6N female mice . Envirou. Health Perspect ., 116,644-650. Guruge, K.S ., Taniyasu, S., Yamashita, N., Wijoratna, S., Mohotti, K.M., Seneviratne, H.R., Kannan, K., Yamanaka, N. and Miyazaki, S. (2005): Perfluorinatod organic compounds in human blood serum and seminal plasma : a study of urban and rural tea worker populations in Sri Lanka. J. Environ . Monit., 7, 371-377 . Houde, M., Wells, R.S., Fair, P,A., Bossart, G. D., Hohn,A.A., Rowles. T.K ., Sweeney, J .C., Solomon, K.R . and Muir, D.C . (2005) : Polyfluoroalkyl compounds in frce-ranging bottlenose dolphins (Tursiops lnancatus) from the Gulf of Mexico and the Atlantic Ocean. Environ . Sci. Teclmol., 39, 6591-6598. Jin, Y, Sailo, N., Harada, K.H ., Inouc, K. and Kointmi, A. (2007): Historical trends in human serum levels of perfluorooetanoate and perfluorooctane sulfonate in Shenyang, China. Tohoku J. Exp . Med ., 212,63-70. Kamum, K., Cotsolini, S., Falandysz, J., Oehttte, G., Pocardi, S. and Giesy, J.P (2002) : PerRuorooctanesulfonate and related fluorinated hydrocarbons in marine mammals, fishes, and birds from coasts of the Baltic and the Mediterranean Seas . Environ. Sci. Technol ., 36, 3210-3216. Kamian, K., Koistinen, J., Beckmen, K., Evans, T, Gotzelany, J.F., Hansen, K.1 ., Jones, PD., Relic. E., Nyman, M. and Giesy, J.P. (2001) : Accumulation of perfluorooctane sutfonate in marine mammals. Environ. Sci. Technol., 35, 1593-1598. Kannan, K., Petrotta, E. and Thomas, N.J . (2006) : Association between perfluorinated compounds and pathological conditions in southern sea otters . Enviroa Sci. Technol., 48,4943-4948 . Nohara, K_, Izumi, H., Tamura, S., Nagata, R. and Tohyama, C. (2002). Effect of low-dose 2,3,7,8-tettachlorodibenzo-p-dioxin (TCDD) on influenza A virus-induced mortality in micc. Toxicology. 170, 131-138. Olsen, G.W., Butris,l.M ., Ehttisman, DJ., Froehlich, J.1A!, Seacat, A.M ., Butenhotl; J.L . and Zobel, L.R. (2007): Half-IM of serum elimination of perlluotooctanesulfonate, perfluorohexanesulfonate, and per8uotooetanoate in retired fluormchemical produclion workers. Environ. Health Perspect 115, 1298-1305. Olsen, G.W., Burris, J.M ., Mandet, 1.H. and Zobel, L.R . (1999) : Serum perfluorooctane sulfonate and hepatic and lipid clinical chemistry tests in fluornchemical production employeos. J. Occup. Environ. Mod., 41, 799-806. Olsen, G.W., Church, T.R., Miller, J.P., Burris, 1.M ., Hansen, K .J ., Lundberg, J .K ., Anitage, J .B ., Herron, R.M ., Medhd'vadehkashi, Z., Nobiletti, J.B ., O'Neill, E.M., Mandel, J.H_ and Zobel, L.R. (2003): Perfluorooctanesulfonate and other fluorochemicais in the serum of American Red Cross adult blood donors. Environ. Health Perspect ., 111. 1892-19p1. Peden-Adams, M.M., Keller, J.M ., Eudaly,l.G., Bergar,J., Gilkeson, G.S . and Keil, D.E. (2008) : Suppression of humoral immunity in mice following exposure to pertluorooctane sulfonate. Toxicol . Sci., 104,144-154. Yamashita. N_, Kannan, K., Taniyasu, S. . Horii, Y, Okazawa, T. . Petrick, G. and Gamo, T., (2004): Analysis of perfluorinated acids at parts-per-quadrillion levels in seawater using liquid chromatography-tandem mass spectrometry . Environ. Sci. Technol ., 38, 5522-5528. ,ao 0N SN1d1NO0 Vol. 34 No. 6