ent 384 (2007) 221–228www.elsevier.com/locate/scitotenv

Science of the Total Environm

Induction of CYP1A and DNA damage in the fathead minnow(Pimephales promelas) following exposure to biosolids

Constance Sullivan a, Carys L. Mitchelmore b,⁎, Robert C. Hale a, Peter A. Van Veld a

a Virginia Institute of Marine Science, P.O. Box 1346, Gloucester Point, VA 23062, United Statesb University of Maryland Center for Environmental Science, Chesapeake Biological Laboratory, 1, Williams Street, P.O. Box 38,

Solomons, MD 20688, United States

Received 28 March 2007; accepted 5 June 2007Available online 5 July 2007

Abstract

Biosolids (treated sewage sludge) are increasingly disposed of on land. Thus particle-sorbed and dissolved constituents havethe potential to enter nearby watersheds. Although organic contaminants are known to be present in biosolids these are notcurrently regulated and little data exist on their potential toxicity to aquatic organisms. We exposed Pimephales promelas to twoconcentrations of biosolids (0.5 and 2.5 g l− 1) for 28-days (static-renewal) and characterized contaminants present and the extentof CYP1A and DNA damage induction at various time points. Many organic contaminants were detected in the biosolids, withpolycyclic aromatic hydrocarbons (PAHs) being the dominant class. Substantial levels of polybrominated diphenyl ethers(PBDEs) and nonylphenols (NPs) were also present. Significant induction of hepatic CYP1A protein compared with controls(Pb0.05) was observed in both low (0.5 g l−1) and high (2.5 g l−1) exposed fish from Day 7. CYP1A levels peaked at Day 21with 21-fold and 8-fold inductions over controls in high and low dose fish respectively. Induction of DNA damage inhepatocytes (single strand breaks as measured using the COMET assay) was observed in both exposures compared with controlson Days 14 and 28 (Pb0.05). A significant correlation was found between CYP1A induction and DNA damage (Pearsoncorrelation index, Pb0.05). It is plausible that activation of PAHs may be responsible for the induction of CYP1A and resultingincrease in DNA damage. Our data show the potential for detrimental effects in the event of exposure of aquatic organisms tobiosolids and the need for further investigations of possible impacts due to constituents not covered by current guidelines.© 2007 Elsevier B.V. All rights reserved.

Keywords: Biosolids; Pimephales promelas; Genotoxicity; Cytochrome P450; PAHs

1. Introduction

Biosolids are produced from sewage sludge that hasundergone one or more additional stabilization treat-ments, such as lime addition, anaerobic/aerobic diges-

⁎ Corresponding author. Tel.: +1 410 326 7283; fax: +1 410 3267210.

E-mail address:mitchelmore@cbl.umces.edu (C.L. Mitchelmore).

0048-9697/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2007.06.006

tion, composting, heat drying, or pelletization (USEPA,1999). Class A biosolids have been subjected to furthertreatment to reduce pathogen concentrations. Excep-tional Quality (EQ) biosolids are Class A biosolids thathave undergone further processing to meet morestringent concentration limits of pollutants (USEPA,1999, 1994). Class B biosolids constitute the bulk ofthose produced and may exhibit higher burdens ofpathogens or chemical contaminants than Class A and

222 C. Sullivan et al. / Science of the Total Environment 384 (2007) 221–228

EQ (USEPA, 1999). In the US the majority (53% in1998) of biosolids are disposed of on land and representan attractive economical option for farmers versuscommercial inorganic fertilizers (USEPA, 1999). Thisactivity is expected to increase over time (USEPA,1999). Class A and B biosolids application are restrictedbased on the slope of land, proximity to flowing waterand lakes, proximity to groundwater sources, frequencyof application, and mode of dispersal (USEPA, 1994).EQ biosolids are marketed directly to the public forprivate use and the EPA makes no record of how theyare used (USEPA, 1999, 1994). There are no restrictionson application rates or where EQ biosolids are used(USEPA, 1999). Biosolids and associated contaminantsthus have the potential to run off into nearby watersheds.

Considering their inherent complexity and potentialto contain chemical pollutants, compositional studiesof biosolids are limited, especially with respect to orga-nic chemicals. Regulation is restricted to nine metals(USEPA 1999, 1994). Nonetheless, biosolids maycontain a variety of contaminants, including polybro-minated diphenyl ethers (PBDEs), polycyclic aro-matic hydrocarbons (PAHs), polychlorinated biphenyls(PCBs), organochlorine pesticides (e.g. DDE andchlordanes) and tributyltin (TBT) and pharmaceuticals(Hale et al., 2001; Hale and LaGuardia, 2002; Harrisonet al., 2006; Kinney et al., 2006). Other contaminantsfound in biosolids include nonylphenol polyethoxylates(NPEOs) and their degradation products (i.e. nonylphe-nols: NPs), synthetic musk compounds, and triclosan (LaGuardia et al., 2001, 2004; Hale and LaGuardia, 2002).

Relative to investigations of wastewater effluentimpacts on resident species, there have been very fewpublished studies on the effects of biosolids. Ciparis andHale (2005) measured bioaccumulation of PBDEs frombiosolids in the oligochaete, Lumbriculus variegates.Though this study did not address effects of biosolids onthe worms themselves, it demonstrated bioaccumulationof biosolids-associated contaminants and the potentialfor biomagnification. A terrestrial study by Paul et al.(2005) reported that sheep raised on pastures thatreceived biosolids as fertilizer demonstrated reducedparental and fetal body weights and alterations in malefetus hormones coupled with reduced testes size.

Our study was designed to evaluate potential bio-markers of exposure to biosolids in aquatic organismsdue to increasing use of these residuals as soil amend-ments, their potential for entry into local waterways,and the paucity of pertinent biological studies. The testorganism chosen was Pimephales promelas, a fresh-water oviparous fish in the family Cyprinidae. Its lifehistory and responses to pollution are well-studied.

P. promelas is widely used as a toxicity test organ-ism by the U.S. Environmental Protection Agency(USEPA, 2002). Past studies have utilized P. promelasfor study of CYP1A induction (Colavecchia et al.,2004; Lindstrom-Seppa et al., 1994) and DNA damage(Choi and Meier, 2000; Choi et al., 2000; Shugart,1988). We chose CYP1A induction and DNA damageas indicators of effects, as the biotransformation andactivation of procarcinogens by CYP1A can lead todetrimental effects to DNA. To determine the possiblecompounds responsible for changes in the biomarkers,we also analyzed the biosolids and exposure water forcompounds known to induce these biomarkers.

2. Materials and methods

2.1. Exposure

Male P. promelas (∼6 months old) were obtainedfrom Aquatic Biosystems (CO) and were maintained inwell water at 21±1 °C and fed tetra flake food daily.P. promelas underwent three exposure regimes to ClassA EQ biosolids: high dose (2.5 g l−1), low dose(0.5 g l−1), and a control dose (clean water). Biosolids(stabilized by anaerobic digestion) were donated by aNortheastern U.S. generator. Treatments were preparedin duplicate for a total of six tanks each with a totalvolume of 20 gallon. Biosolids were placed in 250 μmmesh bags to reduce turbidity. Twenty-four males wereinitially placed in each tank, with four fish sampled fromeach tank per time point. Sampling for CYP1A inductiontook place on days 0, 3, 7, 14, 21, and 28 during the 28-day exposure. The Comet Assay was performed onfish taken on days 0, 3, 7, 14, and 28. The exposure wasstatic-renewal, with new biosolid bags and 95% waterchanges carried out once a week on the day after sam-pling. To reduce stress, fish were kept in shaded tankswith structures for shelter. Outside interaction waskept to daily feeding, water changes, and dosing. Waterquality parameters (pH, dissolved oxygen and ammonia)were measured throughout the exposure period. Betweenwater changes, the pH in the control tank remained at 7.5,dissolved oxygen ranged from 9 to 10 mg l−1, andammonia remained below detectable levels for theduration. The pH in the low dose tank ranged from 7.0to 7.5, dissolved oxygen remained at 9 mg l−1, andammonia ranged from 0.00009 to 0.00186 mg l−1.Conditions in the high dose tanks were: pH 6.5–7.5;dissolved oxygen 6.0–8.5 mg l−1; ammonia 0.006–0.0495 mg l−1. Water clarity decreased over timebetween exposures as more biosolids particles werereleased from the mesh bags.

223C. Sullivan et al. / Science of the Total Environment 384 (2007) 221–228

2.2. Chemical analyses

Whole biosolids and water from exposure tanks wereanalyzed for a variety of organic contaminants. Controlwell water was also analyzed to determine the presenceof any pre-existing contaminants.

To obtain representative water for extraction, 2.5 g l−1

of biosolids were placed in 250 μm mesh bags in a tankcontaining well water. The biosolids were allowed toequilibrate for 24 h before sampling of water. Nosurrogate standard was added. A one liter distilled waterblank was run in conjunction with the extractions. Threeone liter water samples were analyzed and mimic con-centrations present in the tanks on Day 1. The pH of eachreplicate was adjusted to N11 using 6 N NaOH prior tosequential extraction in glass separatory funnels withthree aliquots of dichloromethane (DCM) totaling200 ml. Following the base/ neutral fraction, the waterwas adjusted to pHb2 using 3 N HCl (acid extract) andre-extracted with a total 200 ml of DCM. Emulsions thatformed were collected after extraction with the thirdaliquot of DCM. These were frozen to separate phasesand the DCM layer was collected and combined with thepreviously obtained acid DCM extract. Each extract wasconcentrated to 500 μl under a high purity nitrogen gasstream, solvent exchanged to toluene, and reduced to200 μl. Samples were then spiked with the internalstandard perinaphthenone and analyzed by gas chroma-tography/mass spectrometry (GC/MS).

The extraction procedure for the analysis for wholebiosolids was based on La Guardia et al. (2004). Thematerial was not freeze-dried prior to analysis, as itcontained only 2.88% water. BDE-166, PCB-30,-65,and -204, perinaphthenone, acenaphthene-d10, chry-sene-d12, 1,4-dichlorobenzene-d4, naphthalene-d8, per-ylene-d12, phenanthrene-d10, and 1,1-binaphthyl wereadded as surrogate standards. Two grams of biosolidssample were added to 33 g NaSO4. A NaSO4 blank wasrun in conjunction with the sample. Relatively nonpolarchemicals were extracted on a Dionex ASE 200, using60 ml of DCM. Conditions were: two extraction cycles,vessel pressure 1000 psi, temperature at 100 °C, heatfor 5 min, static for 5 min. A 60% vessel flush was used.Sample solvent volumes were reduced to 8 ml undernitrogen gas prior to cleanup on a size exclusion chro-matograph (Waters 717+Autosampler; Envirosep-ABC®, 350×21.1 mm column, Phenomenex). Thecolumn was eluted with DCM at 5 ml/ min. The first50 ml contained high molecular weight lipids and werediscarded. The next 60 ml contained the xenobioticcompounds of major interest. This fraction was solventexchanged to hexane, reduced to 500 μl under nitrogen

gas, and purified on a 2 g silica column. The columnwas sequentially eluted with 3.5 ml 100% hexane,6.5 ml 60:40 hexane:DCM, 5 ml 25:75 acetone:DCM,and 10 ml 100% acetone. The first elutent, containingaliphatic compounds, was discarded. The second frac-tion contained moderately nonpolar aromatic com-pounds (e.g., PBDEs, PCBs, and PAHs) and the thirdcontained moderately polar compounds such as NPs.The fourth fraction contained more polar compounds.The second (S2), third (S3), and fourth (S4) fractionswere reduced under a stream of nitrogen gas and ex-amined by GC/MS. An internal standard containingdecachlorodiphenyl ether, pentachlorobenzene, and p-terphenyl was then added.

Detection was by GC/mass spectrometry (MS)(Varian Saturn 4D GC/MS). The GC temperatureprogram used was: initial column setting 75 °C, hold1 min, ramp at 4 °C/ min, hold at 350 °C for 20.25 min.Total run time was 90 min, injector 320 °C, transfer line315 °C, MS manifold 280 °C. For acquisition segment 1on the MS, the mass range was 100–500 m/z+ at a rateof 0.670 s/scan for 49.50 min. For segment 2, the massrange was 100–650 m/z+ at a rate of 0.770 s/scan for40.50 min. The column used was a 60 m DB-5 with a0.25 μm film thickness and 0.32 mm inner diameter(J&W Scientific). The carrier gas was helium. Injectionswere made in splitless mode. Compounds of interestwere quantified using a five-point linear calibrationcurve using the internal standard and selected ions foreach targeted compound.

2.3. Biomarker endpoints

Fish were euthanized by an overdose of tricainemethane-sulfonate (MS-222). Livers were removed andplaced on ice with two thirds of each liver snap frozenfor CYP1A analyses (stored in liquid N2 until analysis)and the other portion used immediately for the CometAssay.

Hepatic CYP1Awas measured by Western blot usingthe monoclonal antibody Mab 1-12-3 (a gift from JohnStegeman, Woods Hole Oceanographic Institution) asdescribed previously, with modifications (Van Veldet al., 1997). Microsomes were prepared by homoge-nizing livers in 1 ml of stabilization buffer (100 mM KPcontaining 20% glycerol; 1 mM dithiothreitol; 1 mMEDTA; pH 7.4) containing 10 μl of 0.1 mM phenyl-methyl sulfonyl fluoride to minimize protein degrada-tion. The homogenate was centrifuged twice at 12,000 gfor 11 min, the pellets being discarded. Supernatantswere further centrifuged at 100,000 g for 63 min and themicrosomal pellets resuspended in 50 μl of stabilization

224 C. Sullivan et al. / Science of the Total Environment 384 (2007) 221–228

buffer. Total protein concentrations were determined bythe Bradford total protein assay using BSA as standard(Bradford, 1976). Proteins (20 μg) were separated byelectrophoresis through reducing, denaturing SDS-PAGE on 12% polyacrylamide gels and transferred tonitrocellulose. Blots were blocked with 5% casein intris-buffered saline (TBS) and incubated in the primarymonoclonal antibody Mab 1-12-3 followed by incuba-tion in IR-linked goat anti-mouse CY5 heavy and lightchain specific secondary antibody (Jackson ImmunoR-esearch Laboratories, Inc, PA). Before and after eachantibody incubation, blots were washed three times for10 min each in 0.1% Tween-20/ TBS. Images werecollected and analyzed using a LiCor Odyssey (LiCorBiosciences) and quantitated against pre-calibrated spot(Leiostomus xanthurus) microsomal standards as de-scribed previously (Van Veld et al., 1997).

The Comet Assay method performed was modifiedfrom Mitchelmore and Chipman (1998a). Hepatocyteswere used due to their role in the biotransformation andprocessing of accumulated contaminants. To obtainhepatocytes, the liver was placed in ice-cold aeratedHepes-buffered HBSS (1 mM Hepes; Ca2+ and Mg2+

free; pH 7.6), minced and filtered through a 70 μm filter.Cell viability was assessed using trypan blue exclusionand only samples with viability greater than 85% wereused. Microscope slides were coated with 1% normalmelting point agarose (NMA) in PBS and dried. Tenmicroliters of the cell suspension was added to 100 μl of0.6% low melting point agarose (LMPA) in aeratedHepes-buffered HBSS (1 mMHepes; Ca2+and Mg2+free;pH 7.6) at 37 °C and layered over the NMA. Afterpolymerization of the agarose, 100 μl of 0.6% LMPAat 37 °C was layered over the cell suspensions. Threeslides per sample were prepared. The slides were allowedto solidify before placement in lysing solution (10%DMSO, 1% Triton X-100, 2.5 MNaCl, 100 mMTris, 1%N-laurylsarcosine; pH 10.0) for at least 1 h in the dark at4 °C. Slides were drained, rinsed with milli-Q water andplaced in electrophoresis buffer (200 mM NaOH,100 mM EDTA; pHN12) in a horizontal electrophoresischamber for 15 min. Electrophoresis (25 V, 300 mA) wascarried out for 15 min after which the slides were rinsedthree times with neutralizing buffer (0.4 M Tris, pH 7.5)for 5 min. Slides were drained, placed in ice-coldmethanol for 5 min, dried in the dark at room temperature,and placed in a tightly closed container with desiccant inthe dark until analysis (Woods et al., 1999).

Slides were re-hydrated in aerated Hepes-bufferedHBSS (1 mM Hepes; Ca2+ and Mg2+ free; pH 7.6)containing 2 μg g−1 ethidium bromide (Woods et al.,1999). Slides were analyzed using an Olympus BX50

epifluorescence scope (x 200 magnification) with a QImaging Retiga 1300 camera. The Komet 6.0 (KineticImaging, Liverpool, UK) image analysis package wasused to score each sample. Fifty cells were scoredrandomly per slide using parameters of tail % DNA.

2.4. Statistical analysis

All statistical analyses were run using the statisticalpackage MiniTab (Version 14). Data were tested forsignificant differences between tanks using the standarddeviations; tanks were considered replicates if valueswere within one standard deviation of each other. If nosignificant difference was found, tanks were pooled for asample size of eight fish at each time point for eachtreatment. If data met assumptions for a parametric test,one-way analyses of variance (ANOVAs) with Tukey'spost-hoc test were used to determine variation betweendoses at each time point and doses over time. If data didnot meet assumptions of a parametric test, nonparametrictests were used. To test for significant differences overtime, a Kruskal–Wallis test was used. To test forsignificant differences at each time point, a Mann-Whitney test was used.

3. Results

3.1. Chemical analyses

No detectable levels of targeted organic contaminantswere found in the control well water. Extractions fromthe base/ neutral fraction of exposed water also showedno detectable levels of contaminants. The acid portion,contained nonylphenols (although quantitation of levelswas not carried out) but no other targeted contami-nants were detected. In the S2 fraction of the biosolidsextracts, PAHs were the dominant compounds (seeTable 1). The total amount of PAHs was 43, 000 ng g−1

(dry weight). PBDEs were also present at 720 ng g−1).Nonylphenols were detected in the S3 and S4 fractions ofthe biosolids extracts at 1, 520, 000 ng g−1; see Table 2)with mono-and di-nonylphenol polyethoxylates (NPEOs)totaling 51, 200 ng g−1. The personal care productstriclosan and tonalide were also detected at 4, 450 and1, 890 ng g−1, respectively.

3.2. Biomarker endpoints

CYP1A levels between the treatments at time zerowere compared and there were no significant differences(Kruskal–Wallis, pN0.05); therefore, these data werepooled. Differences between doses were tested pairwise

Table 2Organic contaminants detected in the biosolids S3 and S4 fractions

Compound Concentration (ng g−1 dry weight)

Nonylphenols (total) 1,470,000Nonylphenol 1-ethoxylate 48,200Nonylphenol 2-ethoxylate 3040Triclosan 4450Tonalide 1890

225C. Sullivan et al. / Science of the Total Environment 384 (2007) 221–228

at each time point using the Mann–Whitney test. Tocorrect for the data not being independent of each other,the alpha value was adjusted to 0.017. CYP1A wassignificantly induced (Pb0.01) in low dose fish afterday 21 and after day 3 in the high dose relative tocontrols (see Fig. 1). CYP1A levels peaked in the highdose on day 21 with a 21-fold induction when comparedto the control dose. The low dose also reached its peakinduction on day 21, with an eight-fold differenceobserved between dosed and control fish. There weresignificant differences between the low and high dosesfrom Day 3 (pb0.01).

Exposure of fish for 28 days to EQ biosolids resultedin significant increases in DNA damage in both the low

Table 1Organic contaminants detected in the biosolids S2 fraction

Compound Concentration (ng g−1 dry weight)

Naphthalene 1122-methyl naphthalene 2681-methyl naphthalene 286Biphenyl 262Diphenyl ether 382,6 and 2,7 dimethylnaphthalene 5201,3 dimethylnaphthalene 5991,6 dimethylnaphthalene 4631,4 and 2,3 dimethylnaphthalene 2751,5 dimethylnaphthalene 156Acenaphthylene 26.91,2 dimethylnaphthalene 3081,8 dimethylnaphthalene 554Acenaphthene 151Debenzofuran 1762,3,5-trimethylnaphthalene 1560Fluorene 379Dibenzothiophene 1830Phenanthrene 3690Anthracene 6622-methylphenanthrene 22401-methylphenanthrene 35303,6 dimethylphenanthrene 575Fluoranthene 4330Pyrene 5050Benz(a)anthracene 1400Chrysene 2210Benzo(b)fluoranthene 2200Benzo(k)fluoranthene 1990Benzo(e)pyrene 1370Benzo(a)pyrene 1650Perylene 549Indeno(1,2,3-cd)pyrene 1580Dibenzo(a,h)anthracene 362Benzo(g,h,i)perylene 1660BDE 47 308BDE 100 57.0BDE 99 294BDE 154 21.3BDE 153 21.6

and high dose hepatocytes (Fig. 2). There was nosignificant difference (pN0.05) between the treatmentson day zero, allowing these points to be pooled. Thecontrol dose had significant variance over time(pb0.001): day 28 was significantly lower than days0, 3, and 7. On days 3 and 7, no significant differenceswere found between doses (pN0.05); however, therewas a trend for increased strand breakage at the low andhigh doses compared to the control dose when using theTukey's post-hoc test. On day 14, there were significantdifferences between the doses (p=0.000). A Tukey'spost-hoc test showed the low and high doses weresignificantly higher than the control dose; however, thelow and high doses were not significantly different fromeach other. Day 28 also had significant differencesbetween the doses (pb0.05). A Tukey's post-hoc testshowed the control dose was significantly lower than thelow and high doses, though there was no differencebetween the low and high doses.

Fig. 1. Hepatic CYP1A in P. promelas exposed to 0, 0.5, or 2.5 g l−1

EQ biosolids for 28 days. Values shown are means±SD (n=8). ⁎;significantly different from controls at pb0.01. Inset: Western blot ofCYP1A in P. promelas from Day 21. Lanes 1 and 2 are fish from thecontrol dose; lanes 3–6 are Leiostomus xanthurus microsomalstandards ranging from 0.05–0.70 pmol CYP1A mg protein−1; lanes7 and 8 are from fish at the low dose; lanes 9 and 10 depict 2 fish fromthe high dose.

Fig. 2. Percentage Tail DNA (%)in hepatocytes of P. promelasexposed to 0, 0.5, or 2.5 g l−1 EQ biosolids for 28 days. Values shownare means±SD (n=8). ⁎Significantly different from controls atpb0.05.

226 C. Sullivan et al. / Science of the Total Environment 384 (2007) 221–228

A Pearson correlation index was used to test forcorrelations of CYP1A induction and DNA damage atan alpha value of 0.05. A significant positive correlationwas found between CYP1A induction and DNA damage(p=0.038).

4. Discussion

While an exhaustive analytical survey was not con-ducted, a number of contaminantswere detected in the EQbiosolids tested here. It is a reasonable expectation thatcontaminants would be at lower levels in the EQ bio-solids, than for the more commonly studied Class Bbiosolids. Certain stabilization processes, e.g. aerobiccomposting, used to generate EQ or Class A bio-solids have been reported to decrease burdens of NPs(La Guardia et al., 2001). Heat treatment or pelletizationwould be expected to volatilize or thermally degrade somecontaminants. Class B biosolids undergo less treatmentand have lower stringent metals standards. Nonetheless,NP concentrations in the EQ biosolids herein, weresimilar to, or higher than most published biosolids values(La Guardia et al., 2001). The PBDEs were also detectedin the EQ biosolids at 702 ng g−1. This is slightly less thanthose reported in a 2001 survey ofUS biosolids of varyingclass (Hale et al., 2001). The PBDE congener pattern wassimilar to the commercial penta-product used in NorthAmerica DE-71, with a modestly reduced BDE-99contribution. Deca-(predominantly BDE-209) was notassayed in this study, but is generally not observed inaquatic organisms. A diverse array of PAHs were presentin the EQ biosolid, including alkylated and nonalkylatedcompounds. The former are common in petroleum.ManyPAHs are vulnerable to aerobic degradation and were

likely reduced in concentration during sludge stabilizationprocesses.

The present study confirms the presence of contami-nants in biosolids that are capable of inducingCYP1A andDNA damage, e.g. PAHs. Not all PAHs induce CYP1A;the most potent are the higher molecular weight forms(Goksoyr and Forlin, 1992; Chaloupka et al., 1995), ofwhich there are many representatives in the biosolids usedin this exposure, including acenaphthylene, ace-naphthene, dibenzofuran, fluorene, phenanthrene, anthra-cene, fluoranthene, pyrene, benz(a)anthracene, chrysene,benzo(b)fluoranthene, benzo(k)fluoranthene, and benzo(a)pyrene. The model CYP1A-inducing compound isbenzo(a)pyrene (BaP), and it is used in many inductionstudies (Chaloupka et al., 1995; Van Veld et al., 1997;Stegeman et al., 1981). PBDEs have been postulated toinduce CYP1A due to their similarities to dioxin-likecompounds. Certain congeners (BDE-66, -85, -153, -183)induce CYP1A. However, the most abundant congenersfound in this study (BDE-47 and BDE-99) can act asinhibitors of other Ah-receptor agonists (Chen andBunce,2003). Other studies support the hypothesis that PBDEscontained in the biosolids used in the present study do notinduce CYP1A (Timme-Laragy et al., 2006; Tomy et al.,2004; Boon et al., 2002; Darnerud et al., 2001). In addi-tion, Tomy et al. (2004) and Stapleton et al. (2006) suggestPBDEs may be biotransformed in a manner similar tothyroid hormones and thus do not induce CYP1A activity.

As with CYP1A induction, the PAHs are the class ofcompounds most likely responsible for the observedincrease in tail % DNA. DNA damage has been demon-strated in organisms from a variety of PAH-contaminat-ed environments (Winter et al., 2004; Brown andSteinert, 2003; Roy et al., 2003; Anderson et al.,1999). Many PAHs are known genotoxicants/carcino-gens (e.g. B(a)P) and have been shown to induce DNAdamage, albeit by different mechanisms in differentspecies (Mitchelmore and Chipman, 1998a,b; Mitch-elmore et al., 1998a,b; Nacci et al., 1996; Shugart,1988). DNA damage results from the biotransformationof PAHs by CYP1A to harmful metabolites that can bindto DNA and form adducts (e.g. diol epoxides;Mitchelmore et al., 1998a; Buhler and Williams,1988). Some metabolites can also form reactive oxygenspecies and other radical species (e.g. via quinone redoxcycling), which cause DNA damage (Mitchelmore et al.,1998a; Anderson et al., 1994; Buhler and Williams,1988). Another class of compounds contained in thesebiosolids that may be contributing to the DNA damageis the PBDEs. BDE-47, a component of the Penta-BDEmixture, has been reported to produce a significantlyincreased amount of micronuclei in erythrocytes from

227C. Sullivan et al. / Science of the Total Environment 384 (2007) 221–228

Scophthalmus maximus after aqueous exposure (Bar-ioene et al., 2005). We found that there was a significantcorrelation between CYP1A levels and the extent of tail% DNA, which may indicate that the damage to DNA isoccurring from products of CYP1A induction, such asvia the PAHs.

In the present study we used mesh bags to exposeP. promelas to biosolids. As a result, the fish wereexposed to contaminants via several routes due to theirexposure to both fine sediment particles and aqueousexposure of dissolved contaminants. In the environmentexposure may be similar, albeit concentrations are likelyto be lower due to dilution and weathering processes.Concentrations of organic constituents likely decrease inthe laboratory exposures between dose renewals. VanVeld et al. (1997) were able to induce CYP1A in fishafter both dietary and aqueous exposures. In each typeof exposure, all fish in that study had CYP1A inductionin hepatocytes by 456 h. The primary sites of CYP1Ainduction from an aqueous exposure were the gill pillarcells, heart endothelium, and vascular elements in alltissues. The primary site of CYP1A induction from thedietary exposure was the gut. Though it is difficult todecipher the precise means by which the fish in thepresent study took up contaminants, it is thought theinduction in the present study is from both dietary (i.e.,sorption of compounds to food) and aqueous exposurevia desorption of contaminants from the biosolids. Thestudy organisms also may have taken up contaminantsfrom direct contact of sediment on the gills.

There was no significant difference in DNA damagebetween the low and high doses over time; however, therewas a trend for an increased effect in the low dosetreatment compared to the high dose. These results implythere was not a dose-related effect of biosolids in thisexposure regime on the amount of DNA damage. Typi-cally, there is a dose-dependent response of DNA damageto an exposure from compounds requiring biotransfor-mation up to a certain concentration (Mitchelmore andChipman, 1998a; Mitchelmore et al., 1998b). However,once a chemical reaches saturation kinetics, the DNAdamage from it may decrease (Mitchelmore et al., 1998b).The results of this study may indicate the levels of thecompounds in the biosolids may be so high that they havereached saturation.

5. Conclusions

Chemicals associated with biosolids have the capa-bility to be taken up by aquatic organisms, inducingCYP1A levels and causing DNA damage. There arecurrently no regulations on the distribution of EQ bio-

solids. Thus, in some situations, application of EQbiosolids to land could impact aquatic ecosystems. Thereare currently no regulations on the distribution of Class AEQ biosolids. The assumption is that their chemicalcontent is low enough that there would be no effecton either terrestrial or aquatic organisms and can thusbe spread anywhere by anyone. The sale is alsonot currently regulated, so no information exists on theamount purchased or applied. In addition, there is norestriction on the time between application and use of theland. Our data indicate the potential for effects on aquaticorganisms and the need for further study of theramifications of biosolids land applications.

Acknowledgements

This study was funded in part by a grant from theVirginia Water Resources Research Center to P.A.V.V.We thank Barb Rutan, Eilen Harvey and Elizabeth Bushfor technical assistance. This manuscript is contributionnumber 2839 from the Virginia Institute of MarineScience and contribution number 4101 from UMCES,Chesapeake Biological Laboratory.

References

Anderson D, Yu TW, Phillips BJ, Schmezer P. The effects of variousantioxidants and other modifying agents on oxygen-radical-generated DNA damage in human lymphocytes in the COMETassay. Mutat Res 1994;307:261–71.

Anderson JW, Jones JM, Steinert S, Sanders B, Means J, McMillin D,et al. Correlation of CYP1A induction, as measured by the P450RGS biomarker assay, with high molecular weight PAHs inmussels deployed at various sites in San Diego Bay in 1993 and1995. Mar Environ Res 1999;48:389–405.

Barioene J, Dedonyte V, Rybakovas A, Andreikenaite L, AndersenOK. Induction of micronuclei in Atlantic cod (Gadus morhua) andturbot (Scophthalmus maximus) after treatment with bisphenol A,diallyl phthalate and tetrabromodiphenyl ether-47. Ekologija2005;4:1–7.

Boon JP, van Zanden JJ, Lewis WE, Zegers BN, Goksoyr A, ArukweA. The expression of CYP1A, vitellogenin and zonaradiataproteins in Atlantic salmon (Salmo salar) after oral dosing withtwo commercial PBDE flame retardant mixtures: absence of short-term responses. Mar Environ Res 2002;54:719–24.

Bradford MM. A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248–54.

Brown JS, Steinert SA. DNA damage and biliary PAH metabolites inflatfish from Southern California bays and harbors, and theChannel Islands. Ecol Indic 2003;3:263–74.

Buhler DR, Williams DE. The role of biotransformation in the toxicityof chemicals. Aquat Toxicol 1988;11:19–28.

Chaloupka K, Steinberg M, Santostefano M, Rodriguez LV, GoldsteinL, Safe S. Induction of Cyp1a-1 and Cyp1a-2 gene expression by areconstituted mixture of polynuclear aromatic hydrocarbons inB6C3F1 mice. Chem Biol Interact 1995;96:207–21.

228 C. Sullivan et al. / Science of the Total Environment 384 (2007) 221–228

Chen G, Bunce NJ. Polybrominated diphenyl ethers as Ah receptoragonists and antagonists. Toxicol Sci 2003;76:310–20.

Choi K, Meier PG. Implications of chemical-based effluent regulationsin assessing DNA damage in fathead minnows (Pimephalespromelas) when exposed to metal plating wastewater. Bull EnvironContam Toxicol 2000;64:716–22.

Choi K, ZongM, Meier PG. Annual review: application of a fish DNAdamage assay as a biological toxicity screening tool for metalplating wastewater. Environ Toxicol Chem 2000;19(1):242–7.

Ciparis S, Hale RC. Bioavailability of polybrominated diphenyl etherflame retardants in biosolids and spiked sediment to the aquaticoligochaete, Lumbriculus variegatus. Environ Toxicol Chem2005;24(4):916–25.

Colavecchia MV, Backus SM, Hodson PV, Parrott JL. Toxicity of oilsands to early life stages of fathead minnows (Pimephalespromelas). Environ Toxicol Chem 2004;23(7):1709–18.

Darnerud PO, Eriksen GS, Johannesson T, Larsen PB, Viluksela M.Polybrominated diphenyl ethers: occurrence, dietary exposure, andtoxicology. Environ Health Perspect 2001;109(supplement 1):49–68.

Goksoyr A, Forlin L. The cytochrome P-450 system in fish, aqua-tic toxicology and environmental monitoring. Aquat Toxicol1992;22:287–311.

Hale RC, LaGuardia MJ. Have risks associated with the presence ofsynthetic organic contaminants in land-applied sewage sludgesbeen adequately assessed? New Solut 2002;12(4):371–86.

Hale RC, La Guardia MJ, Harvey EP, Gaylor MO, Mainor TM, DuffWH. Persistent pollutants in land-applied sludges. Nature2001;412:140–1.

Harrison EZ, Oakes SR, Husell M, Hay A. Organic chemicals insewage sludges. STOTEN 2006;367(2–3):481–97.

Kinney ChadA, Furlong Edward T, Zaugg StevenD, BurkhardtMark R,Werner Stephen L, Cahill Jeffery D, et al. Survey of organicwastewater contaminants in biosolids destined for land application.Environ Sci Technol 2006;40(23):7207–15.

La Guardia MJ, Hale RC, Harvey E, Mainor TM. Alkylphenolethoxylate degradation products in land-applied sewage sludge(biosolids). Environ Sci Technol 2001;35:4798–804.

La Guardia MJ, Hale RC, Harvey E, Bush EO, Mainor TM, GaylorMO. Organic contaminants of emerging concern in land-appliedsewage sludge (biosolids). J Residuals Sci Technol 2004;1(2):105–16.

Lindstrom-Seppa P, Korytko PJ, Hahn ME. Uptake of waterborne3,3',4,4'-tetrachlorobiphenyl and organ and cell-specificinduction of cytochrome P4501A in adult and larval fatheadminnow Pimephales promelas. Aquat Toxicol 1994;28:147–67.

Mitchelmore CL, Chipman JK. Detection of DNA strand breaks in browntrout (Salmo trutta) hepatocytes and blood cells using the single cellgel electrophoresis (comet) assay. Aquat Toxicol 1998a;41:161–82.

Mitchelmore CL, Chipman JK. DNA strand breakage in aquaticorganisms and the potential value of the comet assay inenvironmental monitoring. Mutat Res 1998b;399:135–47.

Mitchelmore CL, Birmelin C, Chipman JK, Livingstone DR. Evidencefor P-450 catalysis and free radical involvement in the productionof DNA strand breaks by benzo[a]pyrene and nitroaromatics inmussel (Mytilus edulis L.) digestive gland cells. Aquat Toxicol1998a;41:193–212.

Mitchelmore CL, Birmelin C, Livingstone DR, Chipman JK. Detectionof DNA strand breaks in isolated mussel (Mytilus edulis L.)digestive gland cells using the “comet” assay. Ecotoxicol EnvironSaf 1998b;41:51–8.

Nacci DE, Cayula S, Jackim E. Detection of DNA damage inindividual cells from marine organisms using the single cell gelassay. Aquat Toxicol 1996;35:197–210.

Paul C, Rhind SM, Kyle CE, Scott H, McKinnell C, Sharpe RM.Cellular and hormonal disruption of fetal testis development insheep reared on pasture treated with sewage sludge. EnvironHealth Perspect 2005;113(11):1580–7.

Roy LA, Armstrong JL, Sakamoto K, Steinert S, Perkins E, LomaxDP, et al. The relationships of biochemical endpoints to histo-pathology and population metrics in feral flatfish species collectednear the municipal wastewater outfall of Orange County,California, USA. Environ Toxicol Chem 2003;22(6):1309–17.

Shugart LR. An alkaline unwinding assay for the detection of DNAdamage in aquatic organisms. Mar Environ Res 1988;24:321–5.

Stapleton HM, Brazil B, Holbrook RD, Mitchelmore CL, Benedict R,Konstantinov A, Potter D. In vivo and in vitro debromination ofdecabromodiphenyl ether (BDE 209) by juvenile rainbow trout andcommon carp. Environ Sci Technol 2006;40(15):4653–8.

Stegeman JJ, Klotz AV, Woodin BR, Pajor AM. Induction of hepaticcytochrome P-450 in fish and the indication of environmental induc-tion in scup (Stenotomus chrysops). Aquat Toxicol 1981;1:197–212.

Timme-Laragy AR, Levin ED, Di Giulio RT. Developmental andbehavioral effects of embryonic exposure to the polybrominateddephenylether mixture DE-71 in the killifish (Fundulus hetero-clitus). Chemosphere 2006;62(7):1097–104.

Tomy GT, Palace VP, Halldorson T, Braekevelt E, Danell R, WautierK, et al. Bioaccumulation, biotransformation, and biochemi-cal effects of brominated diphenyl ethers in juvenile lake trout(Salvelinus namaycush). Environ Sci Technol 2004;38:1496–504.

USEPA. Land application of sewage sludge. A guide for land applierson the requirements of the federal standards for the use or disposalof sewage sludge, 40 CFR Part 503. EPA/831-B-93-002b.Washington, DC.: USEPA; 1994. 105 pp.

USEPA. Biosolids generation, use and disposal in the United States.Washington, DC.: USEPA 1999. EPA530-R 99-009. 74 pp.

USEPA. Methods for measuring the acute toxicity of effluents andreceiving waters to freshwater and marine organisms. 5th edition.2002.

Van Veld PA, Volgelbein WK, Cochran MK, Goksoyr A, Stegeman JJ.Route-specific cellular expression of cytochrome P4501A(CYP1A) in fish (Fundulus heteroclitus) following exposure toaqueous and dietary benzo (a) pyrene. Toxicol Appl Pharmacol1997;142:348–59.

Winter MJ, Day N, Hayes RA, Taylor EW, Butler PJ, Chipman JK.DNA strand breaks and adducts determined in feral and caged chub(Leuciscus cephalus) exposed to rivers exhibiting variable waterquality around Birmingham, UK. Mutat Res 2004;552:163–75.

Woods JA, O'Leary KA, McCarthy RP, O'Brien NM. Preservationof comet assay slides: comparison with fresh slides. Mutat Res1999;429:181–7.