ecotoxicity of silver nanoparticles introduction on the...

Ecotoxicity of Silver Nanoparticleson the Soil NematodeCaenorhabditis elegans UsingFunctional EcotoxicogenomicsJ I - Y E O N R O H , † S A N G J U N S I M , ‡

J O N G H E O P Y I , § K W A N G S I K P A R K , |

K Y U H Y U C K C H U N G , ⊥

D O N G - Y O U N G R Y U , # A N DJ I N H E E C H O I * , †

Faculty of Environmental Engineering, College of UrbanScience, University of Seoul, 90 Jeonnong-dong,Dongdaemun-gu, Seoul 130-743, Korea, Department ofChemical Engineering, Sungkyunkwan University,Changan-gu, Suwon 440-746, South Korea, School ofChemical and Biological Engineering, Institute of ChemicalProcesses, Seoul National University, Shinlim-dong,Kwanak-ku, Seoul 151-744, Korea, College of Pharmacy,Dongduk Women’s University, 23-1, Wolgok-dong,Seongbuk-gu, Seoul 136-714, Korea, College of Pharmacy,Sungkyunkwan University, 300 Cheoncheon dong, Jangan-gu,Suwon, Gyeonggi-do 440-746, Korea, and College of VeterinaryMedicine, BK21 Program for Veterinary Science, SeoulNational University, Shinlim-dong, Kwanak-ku,Seoul, 151-742, Korea

Received December 8, 2008. Revised manuscript receivedFebruary 28, 2009. Accepted March 11, 2009.

In the present study, the ecotoxicity of silver nanoparticles(AgNPs) was investigated in Caenorhabditis elegans usingsurvival, growth, and reproduction, as the ecotoxicologicalendpoints, as well as stress response gene expression.Whole genome microarray was used to screen global changesin C. elegans transcription profiles after AgNPs exposure,followedbyquantitativeanalysisofselectedgenes.Theintegrationof gene expression with organism and population levelendpoints was investigated using C. elegans functionalgenomics tools, to test the ecotoxicological relevance of AgNPs-induced gene expression. AgNPs exerted considerabletoxicity in C. elegans, most clearly as dramatically decreasedreproduction potential. Increased expression of the superoxidedismutases-3 (sod-3) and abnormal dauer formation protein (daf-12) genes with 0.1 and 0.5 mg/L of AgNPs exposures occurredconcurrently with significant decreases in reproductionability. Overall results of functional genomic studies usingmutant analyses suggested that the sod-3 and daf-12 geneexpressions may have been related to the AgNPs-inducedreproductive failure in C. elegans and that oxidative stress mayhave been an important mechanism in AgNPs toxicity.

Introduction

Silver nanoparticles (AgNPs) have a wide range of currentand potential future applications, including spectrally selec-tive coatings for solar energy absorption (1), chemicalcatalysts (2), surface-enhanced Raman scattering for imaging(3), and in particular, antimicrobial sterilization (4), whichhas made them one of the most commonly used nanoma-terials (5). However, these same, effective, biocidal propertieshave the potential to adversely affect beneficial bacteria inthe environment, especially in the soil and water (6).Investigation is increasing into the potential toxic mecha-nisms and long-term effects by which these nanomaterialscould pose environmental risks through widespread produc-tion and use (7, 8). Widely used NPs, such as AgNPs, willmost likely enter the environment and may produce aphysiological response in certain organisms, possibly alteringtheir fitness, and ultimately might change their populationsor community densities. Research and literature regardingthe ecotoxicity of NPs is still emerging and gaps exist in ourknowledge of this area. Recent reports regarding NPs toxicitycome from mammalian studies of respiratory exposure or invitro assays with mammalian cells (9, 10) and ecotoxicologicalNP studies are increasing, with most of the available datainvolving freshwater species and species used for regulatorytoxicology studies, such as Daphnia magna, Oncorhynchusmykiss, and Chlorella kessleri (11-14). Unfortunately, fewstudies have involved terrestrial organisms (15), even fewerspecifically used AgNPs.

Caenorhabditis elegans, a free-living nematode mainlyfound in the liquid phase of soils, is the first multicellularorganism whose genome has been completely sequenced.The genome’s unexpectedly high level of conservation relativeto the vertebrate genome makes C. elegans an ideal systemfor biological studies in areas such as genetics, molecularbiology, and developmental biology, and functional genomictools (gene knock out and RNAi) are available to readily studysublethal effects on the animal’s metabolism and physiology(16). Recently, this species has also been used as an animalmodel for ecotoxicological studies due to its abundance insoil ecosystems, convenient handling in the laboratory, andsensitivity to various types of stresses, utilizing variousexposure media, including soil and water (17-21). In thisstudy, AgNPs ecotoxicity on C. elegans was investigated usingsurvival, growth, and reproduction, as well as, stress inducedgene expression, as toxic endpoints.

Global changes in the nematode transcription profilefollowing AgNPs exposure were detected using whole genomemicroarrays and, subsequently, selected genes analyzed byquantitative real-time PCR (qRT-PCR). DNA microarrayapplications for detecting changes in gene expression profilesfollowing pollutant exposure appear to provide a morecomprehensive, and sensitive insight into toxicity than typicalecotoxicological parameters, such as mortality, growth, orreproduction (22, 23). This “exotoxicogenomic” strategy aimsto develop new predictive models for identifying environ-mental or human health hazards and identify precise andrapid molecular biomarkers of chemical exposure. Microar-ray-based gene expression assays have the potential todescribe the interactions of suites of genes comprisingmultiple metabolic pathways, possibly revealing system-wideperturbations in functions associated with stress. They areparticularly interesting for the toxicity screening of newchemicals, such as NPs, with largely unknown modes ofaction.

* Corresponding author phone: 82-2-2210-5622; fax: 82-2-2244-2245; e-mail: [email protected].

† Faculty of Environmental Engineering, University of Seoul.‡ Department of Chemical Engineering, Sungkyunkwan University.§ School of Chemical and Biological Engineering, Seoul National

University.| Dongduk Women’s University.⊥ College of Pharmacy, Sungkyunkwan University.# College of Veterinary Medicine, Seoul National University.

Environ. Sci. Technol. 2009, 43, 3933–3940

10.1021/es803477u CCC: $40.75 2009 American Chemical Society VOL. 43, NO. 10, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 3933

Published on Web 04/21/2009

In this study, the integration of gene expression withorganismal and population level endpoints was undertakenusing the available C. elegans functional genomic tools, toexamine the ecotoxicological relevance of detected AgNPs-induced gene expression. Such functional genomic studiescan provide experimental evidence of the causal relationshipsbetween the observed altered physiological indicators andgene expression. C. elegans functional genomic tools can aidin the assessment of the physiological impact of up- ordownregulated gene expression and can provide indicatorsof the mode of action from the level of a single gene to thewhole organism (21). As such analyses can be validated usingin vivo mutational approaches (24, 25), here, the biologicalroles of AgNPs-responsive genes in the organisms’ defenseagainst AgNPs toxicity were studied using the selected mutantstrains, such as, mtl-2 (gk125), sod-3 (gk235), and daf-12(rh286). To compare the toxicity of AgNPs to that of Ag ions,toxicity of Ag ion was also investigated in C. elegans using thesame toxic endpoints used for AgNPs toxicity assay.

Materials and MethodsOrganisms. C. elegans were grown in Petri dishes onnematode growth medium (NGM) and fed OP50 strainEscherichia coli according to a standard protocol (26) andyoung adults (3 days) from an age-synchronized culture wereused in all the experiments. To produce age-synchronizedcultures, at 2 to 3 days, eggs from mature adults were isolatedusing a 10% hypochlorite solution, followed by a rinse withM9 buffer (27), and the eggs allowed to hatch on agar plateswith a food source, resulting in synchronized adult wormpopulations. In addition to wild type animal (N2 var. Bristol),mtl-2 (gk125), sod-3 (gk235), and daf-12 (rh286) mutantstrains were used. (Supporting Information Table 1). Wildtype and mutant strains were provided by the CaenorhabditisGenetics Center at the University of Minnesota.

AgNPs and Ag Ion Preparation and Exposure to C.elegans. AgNPs (size <100 nm, Sigma-Aldrich Chemical, St.Louis, MO) were homogenously dispersed in deionized waterby sonication for 13 h (Branson-5210 sonicator, Branson Inc.,Danbury, CT), stirring for 7 days, and filtering through acellulose membrane (pore size 100 nm, Advantec, Toyo ToshiKaisha, Japan) to remove NP aggregations. To compare thetoxicities of AgNPs and Ag ions, aqueous AgNO3 (AG002, NextChimica, Centurion, Republic of South Africa) in deionizedwater was used and the final concentrations of AgNPs andAgNO3 were estimated using multitype inductively coupledplasma emission spectrometer (ICPE-9000, Shimadzu, Tokyo,Japan). The concentrations for AgNPs and AgNO3 were in anequivalent Ag mass basis. From stock solutions (4 mg/L),experimental concentrations (0.05, 0.1, and 0.5 mg/L) ofAgNPs and AgNO3 were prepared in k-media (0.032 M KCland 0.051 M NaCl; ref 28). For toxicity tests, worms wereincubated at 20 °C for 24 h without a food source and thensubjected to each type of analysis. Only wild type C. eleganswere subjected to toxicity studies with AgNO3. An overviewof the experimental scheme is presented in SupportingInformation Table 2.

Characterization of AgNPs. Energy filtering transmissionelectron microscopy (TEM) was used to examine the particleshape and size of the AgNPs. Twenty µL of particle suspensionwere dried onto a 400 mesh carbon-coated copper grid andimaged with a LIBRA 120 TEM (Carl Zeiss, Oberkochen,Baden-Wurttemberg, Germany) at 80-120 kV. The sizedistribution of the AgNPs was evaluated using a Photaldynamic light scattering (DLS) spectrometer, DLS-7000(Otsuka Electronics Co., Inc., Osaka, Japan). AgNPs were alsoinvestigated in test media and in the exposed worms usingan inverted, dark-field microscope (1000× objective lens,CFI Plan Fluor ELWD, NA ) 0.6, Eclipse TE2000-U, Nikon,Tokyo, Japan). The incorporation of AgNPs was examined

by dispersing the transparent worm bodies on a glassmicroscope slide under an adjustable 100 W tungsten lampand a dark-field condenser (dry condenser, NA ) 0.80-0.95,Nikon).

Microarray and Quantitative Real-Time PCR. For mi-croarray analysis, C. elegans were exposed to 0.1 mg/L ofAgNPs for 24 h and the total RNA prepared from the exposedand control worms according to the standard protocol of theRNeasy Mini kit (Qiagen, Hilden, Germany). Five µg aliquotsof each total RNA product were used for reverse and in vitrotranscription followed by application to a GeneChip C. elegansGenome Array (Affymetrix, Santa Clara, CA), which contained22 500 probe sets against 22 150 unique C. elegans transcripts.For quantitative RT-PCR (qRT-PCR) analysis, C. elegans wereexposed to 0.05, 0.1, and 0.5 mg/L of AgNPs and AgNO3 for24 h, and analyzed in triplicate using an oligo(dT) primer(Bio-Rad laboratories, Hercules, CA) and the IQ SYBR GreenSuperMix (Bio-Rad). PCR was carried out on 26 selected genesusing a Chromo4 Real-Time PCR detection system (Bio-Rad).The primers were based on sequences retrieved from the C.elegans database (www.wormbase.org, Supporting Informa-tion Table 3) and to optimize the qRT-PCR conditions,efficiency and sensitivity tests performed for each gene, priorto the main experiment. Three replicates were conductedfor qRT-PCR analysis.

Survival and Growth (24 h Exposure Assay). The survivaland growth of the wild type and mutant worms were assessedafter 24 h exposures to different concentrations of AgNPsand AgNO3, as described previously (18). Briefly, survivalwas assessed by counting the number of live and deadindividuals under a dissecting microscope while probing theworms with a platinum wire. Growth was assessed in heatkilled samples by measurement with a reticle of the bodylength. Five replicates were conducted for survival and growthassays.

Reproduction (72 h Exposure Assay). The effects of AgNPsand AgNO3 on the reproduction of wild type and mutantstrains were investigated by 72 h exposures (29). After a youngadult was exposed to AgNPs or AgNO3 for 72 h, the numberof offspring at all stages beyond the egg were counted. Fivereplicates were conducted for reproduction assays.

Data Analysis. Statistical differences between the controland exposed worms were determined by a parametric t test,and a Pearson correlation test was performed for correlationanalysis, using the Statistical Package for the Social Sciences(SPSS, Chicago, IL).

ResultsCharacterization and Incorporation of AgNPs into C.elegans. AgNPs used for C. elegans toxicity studies werecharacterized by TEM, DLS and dark field microscope (Figure1), which showed the size of AgNPs at mainly about 20nm(Figure 1A) and existing in the mixture of single particles andaggregates (Figure 1A and B). The size distribution in the testmedium was investigated using a DLS method, which showedthat the main nanoparticle sizes distributed in the testmedium were about 14∼20 nm(Figure 1C). To investigatethe incorporation of AgNPs into the body of C. elegans, thescattered light from the AgNPs was measured using lightscattering dark field microscopy (Figure 2), and morescattered light assumed to suggest more AgNPs taken up bythe C. elegans body. Observed internal AgNPs always ap-peared as aggregates and were mainly distributed aroundthe uterine area, with only minor differences observed in thehead and tail areas between the control and exposed worms.

Microarray Analyses. AgNPs exposure-induced changesin transcription profiles investigated using C. elegans genomearrays (GSE14932) showed upregulated expression levels of415 gene probes and downregulation of 1217 following a24 h AgNPs exposure (changes > 2-fold; Supporting Infor-

3934 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 10, 2009

mation Table 4) and 26 of the upregulated and 685 of thedownregulated genes were annotated. The functional analysisof these genes was determined by submitting these mi-croarray results for interpretation using the Gene Ontology(GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)databases (Supporting Information Table 5). Thirteen GOcategories, such as nuclear signaling and transport pathways,were found to be significantly represented within theupregulated genes following the 24 h AgNPs exposure,whereas 149 GOs, including behavior-related pathways suchas response to stimulus, locomotive behavior, feedingbehavior, and reproductive behavior pathways were signifi-cantly represented within the downregulated genes. Ac-cording to the KEGG database, 15 up- and 44 downregulatedAgNPs-responsive C. elegans gene probes have been mappedto known metabolic pathways.

Quantitative Assessment of AgNPs-Responsive GeneExpression. qRT-PCR was performed for 16 genes chosenfrom the microarray data and for 10 genes chosen for theirknown involvement in stress response pathways (Figure 3A;Raw data, Supporting Information Tables 6). The 16 geneswere heat shock proteins (hsp-16.1, hsp-16.2, hsp-16.41, andhsp-70), collagens (col-131 and col-101), P-glycoprotein-related protein (pgp-14), ALG-1 interacting protein (ain-1),uncoordinated protein (unc-130), adaptin-related protein(apt-10), osmotic avoidance abnormal (osm-11), Abnormalcell LINeage (lin-10), conserved oligomeric golgi component(cogc-4), mammalian WNK-type protein kinase homologue(wnk-1), and two nonannotated genes, M162.5 and F02A9.4;and the 10 known stress-response genes were metallothio-neins (mtl-1 and mtl-2), superoxide dismutases (sod-1 andsod-3), glutathione-S-transferase (gst-1), catalases (ctl-2, andctl-3), aging alteration protein (age-1), and abnormal dauerformation proteins (daf-12 and daf-21). The qRT-PCR resultsshowed that the selected genes from the microarray resultstended to yield expression levels lower than those measuredusing the microarray and also revealed that, of the 26 testedgenes, the expression of four genes (M162.5, mtl-2, sod-3,and daf-12) was upregulated by AgNPs exposure. In par-

ticular, the increased expression of mtl-2 and sod-3 occurredin a AgNPs concentration dependent manner. Of the fourupregulated genes, functional analysis was carried out onthe three genes (mtl-2, sod-3, and daf-12) with an identifiedbiological role.

Functional Analysis of AgNPs-Responsive Genes UsingMutant Strains. Organismal and population-relevant toxicityendpoints were examined after short (24 h) and long-term(72 h) exposures to AgNPs in the wild type (Table 1; Rawdata, Supporting Information Table 6). Short-term survivaland growth experiments were compared with the sensitivityof physiological level responses of C. elegans to AgNPsexposure with those at the molecular level (gene expression).Short-term testing only provided a snapshot of the physi-ological status, thus longer term testing was conducted forthe effects on reproductive potential. Although AgNPsexposure did not affect the survival and growth of the wildtype, reproduction was seriously affected, with the numberof offspring per individual dramatically decreased (∼70% ofthe controls in 0.1 and 0.5 mg/L AgNPs). Increased expres-sions of sod-3 and daf-12 genes after a 24 h exposure to 0.1and 0.5 mg/L of AgNPs occurred concurrently with thedramatic decrease in reproduction. A Pearson correlationtest to identify any correlation between increased geneexpressions and higher level effects (Supporting InformationTable 7) showed that, indeed, the sod-3 and daf-12 geneexpressions significantly correlated with the observed effecton reproduction (r2 ) 0.999 and 0.993, respectively; p < 0.01).

C. elegans functional genomics tools were used toinvestigate the biological meaning and/or the ecophysi-ological consequences of the increased expression of mtl-2,sod-3, and daf-12 genes caused by AgNPs exposure. Thesurvival, growth, and reproduction of mtl-2 (gk125), sod-3(gk235), and daf-12 (rh286) mutant strains in response toAgNPs exposure were examined and compared with the wildtype (Table 1). The mutant strains’ survival and growthresponse were not different from the wild type, but thereproductive responses of the mtl-2 (gk125) and sod-3 (gk235)mutants were less sensitive (∼40-60% less at 0.1 mg/L and∼10% at 0.5 mg/L) to AgNPs exposure than the wild type,while the response of the daf-12 (rh286) mutant was similarto the wild type.

Expression of Stress Genes in Response to AgNPsExposure in Wild Type and Mutant C. elegans. To under-stand the mechanism of the different sensitivities of repro-ductive potential to AgNPs exposure in the mutant strains,the mRNA levels of 26 potential AgNPs-responsive genes weremeasured in the mutant strains (Figure 4; Raw data,Supporting Informaton Table 6) and compared to the wildtype (Figure 3A). Predictably, the mtl-2, sod-3, and daf-12genes were not detected in the mtl-2 (gk125), sod-3 (gk235),and daf-12 (rh286) mutants, respectively, but notably, themtl-2 gene was not detected in the sod-3 (gk235) mutant.

FIGURE 1. Characterization of AgNPs in test media using TEM (transmission electron microscopy, A); darkfiled microscope (B); andDLS (dynamic light scattering) spectrometer (C).

FIGURE 2. Images of AgNPs in C. elegans using darkfiledmicroscope. C. elegans was exposed to AgNPs for 24 h andincorporation of AgNPs into the transparent worm body wasinvestigated using an inverted, dark-field microscope.

VOL. 43, NO. 10, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 3935

Considerable differences in these genes’ expression patternstoward AgNPs exposure were found, most notably, in thelevels of hsps, M162.5, col-131, fgp-14, ain-1, apt-10, lin-10,cogc-1, and wnk-1, which were significantly increased in thesod-3 (gk235) mutant, a response distinctively different fromthe wild type. In contrast, the expression of only a smallnumber of genes was changed in the mtl-2 (gk125) and daf-12(rh286) mutants.

Comparison of Toxicity of AgNPs and Ag Ions. Thetoxicities of AgNPs and Ag ions in C. elegans were comparedusing the same toxic endpoints used to investigate AgNPstoxicity, which were stress-related gene expression (Figure3B; raw data, Supporting Information Table 6), survival,growth, and reproduction (Table 1; raw data, SupportingInformation Table 6). Ag ion exposure did not alter the worms’mortality rate or growth, but it seriously decreased repro-duction potential, similar to the effects of AgNPs exposure(Table 1). The degree of reproduction potential decrease,however, was more significant with AgNPs exposure thanwith Ag ions. Gene expression patterns from Ag ion exposurealso showed different tendencies compared to AgNPs ex-posure. Here, Ag ion exposure at the two highest concentra-tions (0.1 and 0.5 mg/L) induced the expression of hsp genegroups (hsp-16.1, hsp-16.2, hsp-16.41, and hsp-70) more than2-fold greater than the control.

Discussion

Preparation, dosing, and maintenance of NPs within the testmedium are important factors when investigating thepotential harmful effect of NP exposure in the environmentand, among the various physicochemical properties of NPs,the aggregation process is an important factor influencingtoxicity. Relatively uniform test solutions of NPs can beachieved by a chemical dispersant or physical methods (30),but these methods, particularly use of chemical dispersant,are ecologically irrelevant as NPs discharged to the environ-ment are not likely to accompany or encounter effectivedispersants. Thus, to be realistic and relevant, the presenttoxicity test design used only physical methods, such as,sonication, stirring, and filtration processes for creatinghomogeneous AgNPs dispersions; these methods wereselected from our previous work for optimizing aqueousAgNPs dispersions (31). As AgNPs exist both individually andin aggregates, their states in the test media, as well as in C.elegans were characterized by image analysis, using TEMand dark field microscopy, to determine the size and thestate of the AgNPs, as well as the uptake and the distributionof AgNPs in the C. elegans bodies (Figures 1 and 2). The TEMprovided information on the size and shape of nanoparticles;however, it could not provide information on whether thenanoparticles existed in single or aggregated forms in thetest medium, as the nanoparticles form aggregates when dried

FIGURE 3. Quantitative assessment of AgNPs-(A) and AgNO3- responsive gene expression in wild type C. elegans (B). The wild typeof C. elegans were exposed to AgNPs for 24 h and qRT-PCR was performed for gene expression analysis. Densitometric values ofstress-response gene expression were normalized using Actin mRNA. Data are presented in arbitrary unit compared to control(control ) 1; replicates number ) 3; mean ( standard error of the mean; *p < 0.05).


on the microscopic observation slide. The DLS result suggeststhat the AgNPs tended to exist as single particles in the testmedium. Clear observation of AgNPs, afforded by thetransparent C. elegans bodies, allowed measurement of NPsthrough light scattering without any pretreatment and thus,dark field microscopy can be used for the initial screeningindex for the study of nanotoxicity in a transparent animalmodel, such as C. elegans (32).

The microarray profile of gene expression in AgNPsexposed C. elegans provided an overview of the worms’molecular responses, revealing seriously affected transcrip-tional processes, with more genes downregulated thanupregulated following a 24 h AgNPs exposure (1217 down-versus 415 upregulated genes), and about 57% of thedownregulated and 17% of the upregulated genes wereannotated. The nonannotated genes may have representedexpressed sequence tags (ESTs) that contain mainly un-translated regions or they may have been C. elegans specificgenes. A major difficulty in analyzing expression data in thisstudy was the relatively limited amount of gene annotationavailable, making the placement of the results in a biologicalcontext challenging. As C. elegans genomics in the ecotoxi-cological context continues to grow, the ability to formulatemeaningful conclusions will be greatly enhanced by C. elegansfunctional genomic analyses. Here, GO and KEGG analysesrevealed that, for most of the AgNPs-responsive genes, GOcategories have not been assigned and the biochemicalpathways not mapped. Only 162 (∼10%) of the 1632 geneswhose expression significantly changed following a 24 hAgNPs exposure have been assigned GO terms and, similarly,only 59 (about 4%) of the gene probes have been mappedto known metabolic pathways included in the KEGG database.Although the effects of various physical and chemical stressorson individual gene expression has been intensively studiedin C. elegans, the biological consequences of global tran-scriptional changes have been largely unexplored, particularlynot changes relating to NP exposure. After exploring globaltranscriptional changes, further studies on the biologicalconsequences of individual gene expression changes causedby AgNPs are warranted.

qRT-PCR assays revealed that increased expression wasthe most pronounced in the mtl-2, sod-3, and daf-12 genes(Figure 3A). Being metallic NPs, the induction of the mtlgene by AgNPs was expected (33, 34) increased expressionof mtl-2, but not mtl-1 was also observed previously hereregarding cadmium exposure (18). Oxidative stress as a toxicmechanism of AgNPs was reported in microorganisms (35)and reactive oxygen species (ROS) generated by AgNPs or Agions was reported to be responsible for observed bactericidalactivity (36). Moreover, a recent study of the mechanisms ofAgNPs toxicity using stress-specific, bioluminescent bacteriademonstrated toxicity of AgNPs via oxidative damage (37).In the present study, increased sod-3 gene expression inexposed worms was considered to confirm the involvementof oxidative stress in AgNPs-induced toxicity in this testsystem, as the enzymes involved in the breakdown of ROSplay an important role in ROS related toxicity. Here, increasedexpression was observed in sod-3 (mitochondrial MnSOD),but not in sod-1 (cytosolic CuZnSOD) and the underlyingmechanism for this, as well as, the involvement of the sod-3gene in these worms’ reproductive failure, discussed below,merits further investigation. From egg through adults C.elegans has six life stages including an option for dauerformation, an alternative developmental stage of C. elegans,in which L1 and L2 stage animals have the option to diverttheir development under unfavorable environmental condi-tions. Signals for the dauer formation decision are integratedvia daf genes, with two types of daf gens investigated in thisstudy; daf-12 and daf-21. The former encodes a nuclearreceptor regulating dauer formation, apparently involved inselecting stage-appropriate developmental programs as daf-defective mutants bypass dauer formation (38, 39). The latterencodes a member of the hsp90 family of molecularchaperones and daf-21 activity is required for larval devel-opment (40, 41). Increased daf-12 expression observed heresuggested that AgNPs exposure may have acted as anenvironmental stressor that regulated C. elegans dauerformation, whereas unchanged daf-21 expression with AgNPsexposure suggested that AgNPs-induced alterations in re-production may have been due to a lack of daf-21 activity.

TABLE 1. Ecotoxicological indicators investigated after exposure to AgNPs and AgNO3in wild type(N2) and mutantstrains(mtl-2(gk125), sod-3(gk235) and daf-12(rh286)). Survival was investigated by counting the number of aliveindividuals compared to total introduced worms; growth was investigated by measuring the body length; reproduction wasinvestigated by counting the number of offspring per individual. Results were expressed as the mean value compared to control(control = 1; replicates number = 5; mean ± standard error of mean)

AgNPs (mg/L)

exposure duration parameters strains 0.05 0.1 0.5

24 h survival wild type(N2) 1.00 ( 0.000 1.00 ( 0.000 0.98 ( 0.026mtl-2(gk125) 1.00 ( 0.000 1.00 ( 0.000 0.98 ( 0.000sod-3(gk235) 1.00 ( 0.000 1.00 ( 0.000 0.98 ( 0.026daf-12(rh286) 1.00 ( 0.000 1.00 ( 0.000 1.00 ( 0.000

growth wild type(N2) 0.97 ( 0.018 0.96 ( 0.022 0.94 ( 0.011mtl-2(gk125) 0.98 ( 0.007 0.97 ( 0.015 1.04 ( 0.007sod-3(gk235) 1.00 ( 0.021 1.00 ( 0.018 1.02 ( 0.009daf-12(rh286) 0.99 ( 0.008 1.00 ( 0.006 0.97 ( 0.003

72 h reproduction wild type(N2) 0.83 ( 0.052a 0.32 ( 0.023b 0.32 ( 0.018b

mtl-2(gk125) 0.79 ( 0.019a 0.79 ( 0.087a 0.41 ( 0.015b

sod-3(gk235) 1.06 ( 0.053 0.88 ( 0.040a 0.40 ( 0.035b

daf-12(rh286) 0.54 ( 0.075b 0.35 ( 0.036b 0.25 ( 0.013b

AgNO3(mg/L)

exposure duration parameters strains 0.05 0.1 0.5

24 h survival wild type(N2) 1.00 ( 0.000 1.00 ( 0.000 1.00 ( 0.000growth wild type(N2) 0.99 ( 0.004 0.99 ( 0.000 0.97 ( 0.016

72 h reproduction wild type(N2) 0.84 ( 0.119 0.51 ( 0.175b 0.40 ( 0.122b

a p < 0.05. b p < 0.01.


Human health risk assessments use molecular biomarkersof human physiology (i.e., specific gene expression) to helpunderstand individual health, but in ecological risk assess-ment, biomarkers are useful only when they can predict theeffects on survival, growth, or reproduction. In most cases,survival, growth, or reproduction, which control popula-tion sizes of organisms, are widely accepted endpoints inecotoxicity monitoring and ecological risk assessment, eventhough their utility for assessing NPs toxicity has yet to be

convincingly demonstrated (42, 43). Linking molecularchanges with relevant ecological responses will greatlyimprove the predictive powers of tests based on molecularresponses and remains one of the great challenges inecotoxicology (44). Gene expression profiles and adverseoutcomes need to be conclusively linked at the individualand/or population level, but few NPs ecotoxicological studieshave sought to demonstrate direct experimental relationshipsbetween molecular/biochemical effects and the conse-

FIGURE 4. Quantitative assessment of AgNPs-reponsive gene expression in mutant types C. elegans (mtl-2(gk125), sod-3(gk235),daf-21(rh286)). The mutant types of C. elegans were exposed to AgNPs for 24 h and qRT-PCR was performed for gene expressionanalysis. Densitometric values of stress-response gene expression were normalized using Actin mRNA. Data are presented inarbitrary unit compared to control (control ) 1; replicates number ) 3; mean ( standard error of the mean; *p < 0.05).


quences at higher levels of biological organization. Usingbiochemical and organismal level endpoints, a recent ec-otoxicity study of the effects of titanium dioxide NPs onterrestrial isopods, Porcellio scaber, showed that the activitiesof antioxidant enzymes, such as catalase and glutathione-Stransferase, were affected, but higher level endpoints,including weight change and survival, were not affected (15).

In the present study, among the ecotoxicological param-eters tested (survival, growth, and reproduction), seriouseffects on reproduction were seen following 72 h AgNPsexposures (Table 1) and this method may provide anintegrated approach to predicting population responses ofnematodes to NPs. Environmental AgNPs exposure mayseriously affect C. elegans populations as reproductive failuremay induce significant disturbances in the population, mostprobably population density decreases. The present analysisexamining correlations between gene expression and higherlevel effects revealed that the expressions of the sod-3 anddaf-12 genes highly correlated with reproduction (SupportingInformation), but statistical analyses only indicated a cor-relation, not a causal relationship.

The most significant differences between the mutantstrain responses and the wild type toward AgNPs exposurewere in reproduction (Table 1). The less significant decreasesin reproduction in the exposed mtl-2 (gk125) and sod-3(gk235) mutants suggested that loss of gene function mayactually enhance the worms’ reproductive potential. Manygenes were upregulated in the sod-3 (gk235) mutant withAgNPs exposure (Figure 4), which may have indicated theimportance of this gene in AgNPs-induced toxicity. Inducedstress genes, including hsps, may have been a compensatorydefense mechanism against oxidative stress in the absenceof an important antioxidant enzyme gene, such as sod-3.Although the biocidal activity of AgNPs may be due tooxidative stress, but the toxic mechanism of AgNPs is largelyunknown. Further studies of the mechanism by which themitochondrial sod-3 gene is involved in these worms’reproductive pathway are warranted to better understandoxidative stress-related toxicity with AgNPs exposure. Fromthe present results, functional genomics using mutant strainsappears to be an ideal tool for biomarker discovery inecotoxicological research as it revealed the physiologicalmeaning or function of observed altered gene expressionsand thus helped in identifying the ecological relevance ofcertain molecular biomarkers. Moreover, in contrast to mostfunctional genomics studies using cells or tissues with a focuson biochemical, physilogical pathways, this work representsa more holistic usage of a functional genomics method toidentify biomarkers for enabling the monitoring of the overallfitness of an intact organism in an ecotoxicological context.

There have been discussions regarding the comparativetoxicity of AgNPs and Ag ions (45, 46), the latter’s bactericidalaction having been studied previously (47, 48). The extremelysmall size of NPs have properties different from Ag ions,largely due to their relatively large surface area and relatedhigher reactivity (49). Recently, Hwang et al. (37) demon-strated that AgNPs generate Ag ions and, subsequently,superoxide radicals, partially responsible for observed bio-cidal effect. The present study comparing the toxicity ofAgNPs and Ag ions (Table 1, Figure 3B) suggested that AgNPswere slightly more toxic than Ag ions in terms of reproductionpotential, and also it appeared that different mechanismsexerted toxicity with AgNPs and with Ag ions, as stress-relatedgene expression patterns were different between these twogroups. As it appeared that the biocidal effects of AgNPsmight be partially due to Ag ion generation, further studiesof this aspect of toxicity are required.

In conclusion, AgNPs exert considerable toxicity in C.elegans, particularly in reproduction potential. The resultsof functional genomics analyses suggested that the sod-3

and daf-12 gene expressions may have been related to AgNPs-induced reproductive failure in these worms and thatoxidative stress was an important mechanism in AgNPs-induced toxicity. This study also suggested that the inter-pretation of microarray and subsequent quantitative geneexpression data was greatly enhanced by linking them withorganism and population level experiments.

AcknowledgmentsThis work was accomplished through the generous supportof the Ministry of Environment as “The Eco-technopia 21project”.

Supporting Information AvailableTable 1 is a brief description on the mutant strains used inthis study. Table 2 is an overview of the experimental scheme.Table 3 is a lists of the primers used for gene expressionanalysis using qRT-PCR. Table 4 is a list of the genes thatwere up- or down-regulated (g2 fold) following 24 h AgNPsexposure. Table 5 is a list displaying the GO categoriessignificantly enriched with the up- or down-regulated genesand up- or down-regulated AgNPs-responsive genes that havebeen mapped to known metabolic pathways in the KEGGdatabase. Table 6 is table displaying raw values of qRT-PCR(Figures 3-4) and ecotoxicity indicators used for datanormalization (Table 1). Table 7 is a table displaying thepearson coefficient of correlation between gene expressionand organism/population level indivicators. This material isavailable free of charge via the Internet at http://pubs.acs.org.

Literature Cited(1) Rand, R. B.; Peumans, P.; Forrest, S. R. Long-range absorption

enhancement in organic tandem thin-film solar cells containingsilver nanoclusters. J. Appl. Phys. 2004, 96, 7519–7526.

(2) Zhai, H. J.; Sun, D. W.; Wang, H. S. Catalytic properties of silica/silver nanocomposites. J. Nanosci. Nanotechnol. 2006, 6, 1968–1972.

(3) Yamamoto, S.; Watarai, H. Surface-enhanced Raman spectros-copy of dodecanethiol-bound silver nanoparticles at the liquid/liquid interface. Langmuir. 2006, 22, 6562–6569.

(4) Pal, S.; Tak, Y. K.; Song, J. M. Does the antibacterial activity ofsilver nanoparticles depend on the shape of the nanoparticle?A study of the Gram-negative bacterium Escherichia coli. Appl.Environ. Microbiol. 2007, 73, 1712–1720.

(5) Maynard, A. D.; Michelson, E. the Nanotechnology ConsumerProducts Inventory; Woodrow Wilson International Center forScholars: Washington DC, 2006; online at http://www.nanotechproject.org/consumerproducts.

(6) Griffitt, R. J.; Luo, J.; Gao, J.; Bonzongo, J. C.; Barber, D. S. Effectsof particle composition and species on toxicity of metallicnanomaterials in aquatic organisms. Environ. Toxicol. Chem.2008, 27, 1972–1978.

(7) Hoet, P. H. M.; Bruske-Hohlfeld, I.; Salata, O. V. Nanoparticles- known and unknown health risks. J. Nanobiotechnol. 2004, 2,2–26.

(8) Handy, R. D.; Owen, R.; Valsami-Jones, E. The ecotoxicologyof nanoparticles and nanomaterials: current status, knowl-edge gaps, challenges, and future needs. Ecotoxicology 2008,17, 315–325.

(9) Donaldson, K. Resolving the nanoparticles paradox. Nanomed2006, 1, 229–234.

(10) Cha, K. E.; Myung, H. Cytotoxic effects of nanoparticles assessedin vitro and in vivo. J. Microbiol. Biotechnol. 2007, 17, 1573–1578.

(11) Hund-Rinke, K.; Simon, M. Ecotoxic effect of photocatalyticactive nanoparticles (TiO2) on algae and daphnids. Environ.Sci. Pollut. Res. Int. 2006, 13, 225–232.

(12) Lovern, S. B.; Klaper, R. Daphnia magna mortality when exposedto titanium dioxide and fullerene (C60) nanoparticles. Environ.Toxicol. Chem. 2006, 25, 1132–1137.

(13) Smith, C. J.; Shaw, B. J.; Handy, R. D. Toxicity of single walledcarbon nanotubes to rainbow trout, (Oncorhynchus mykiss):respiratory toxicity, organ pathologies, and other physiologicaleffects. Aquat. Toxicol. 2007, 82, 94–109.


(14) Fujiwara, K.; Suematsu, H.; Kiyomiya, E.; Aoki, M.; Sato, M.;Moritoki, N. Size-dependent toxicity of silica nano-particles toChlorella kessleri. J. Environ. Sci. Health., Part A. Toxic/Hazard.Subst. Environ. Eng. 2008, 43, 1167–1173.

(15) Jemec, A.; Drobne, D.; Remskar, M.; Sepcic, K.; Tisler, T. Effectsof ingested nano-sized titanium dioxide on terrestrial isopodsPorcellio scaber Environ. Toxicol. Chem. 2008, DOI: 10.1897/08-036.1.

(16) Menzel, R.; Yeo, H. L.; Rienau, S.; Li, S.; Steinberg, C. E.;Sturzenbaum, S. R. Cytochrome P450s and short-chain dehy-drogenases mediate the toxicogenomic response of PCB52 inthe nematode Caenorhabditis elegans. J. Mol. Biol. 2007, 370,1–3.

(17) Jonker, M. J.; Sweijen, R. A.; Kammenga, J. E. Toxicity of simplemixtures to the nematode Caenorhabditis elegans in relation tosoil sorption. Environ. Toxicol. Chem. 2004, 23, 480–488.

(18) Roh, J. Y.; Lee, J.; Choi, J. Assessment of stress-related geneexpression in the heavy metal-exposed nematode Caenorhab-ditis elegans: a potential biomarker for metal-induced toxicitymonitoring and environmental risk assessment. Environ. Toxi-col. Chem. 2006, 25, 2946–2956.

(19) Roh, J. Y.; Choi, J. Ecotoxicological evaluation of chlorpyrifosexposure on the nematode Caenorhabditis elegans. Ecotoxicol.Environ. Saf. 2008, 71, 483–489.

(20) Roh, J. Y.; Jung, I. H.; Lee, J. Y.; Choi, J. Toxic effects of di(2-ethylhexyl)phthalate on mortality, growth, reproduction andstress-related gene expression in the soil nematode Caenorhab-ditis elegans. Toxicology 2007, 237, 126–133.

(21) Leung, M. C.; Williams, P. L.; Benedetto, A.; Au, C.; Helmcke,K. J.; Aschner, M.; Meyer, J. N. Caenorhabditis elegansan: Anemerging model in biomedical and environmental toxicology.Toxicol. Sci. 2008, 106, 5–28.

(22) Hamadeh, H. K.; Bushel, P.; Paules, R.; Afshari, C. A. Discoveryin toxicology: mediation by gene expression array technology.J. Biochem. Mol. Toxicol. 2001, 15, 231–242.

(23) Menzel, M.; Rodel, J.; Kulas, J.; Steinberg, C. E. CYP35:Xenobiotically induced gene expression in the nematodeCaenorhabditis elegans. Arch. Biochem. Biophys. 2005, 438, 93–102.

(24) McElwee, J. J.; Schuster, E.; Blanc, E; Thomas, J. H.; Gems, D.Shared transcriptional signature in Caenorhabditis elegansDauer larvae and long-lived daf-2 mutants implicates detoxi-fication system in longevity assurance. J. Biol. Chem. 2004, 279,44533–44543.

(25) Yanase, S.; Ishii, N. Hyperoxia exposure induced hormesisdecreases mitochondrial superoxide radical levels via Ins/IGF-1signaling pathway in a long-lived age-1 mutant of Caenorhabditiselegans. J. Radiat. Res. 2008, 49, 211–218.

(26) Brenner, S. The genetics of Caenorhabditis elegans. Genetics1974, 77, 71–94.

(27) Hitchcock, D. R.; Law, S. E.; Wu, J.; Williams, P. L. Determiningtoxicity trends in the ozonation of synthetic dye wastewatersusing the nematode Caenorhabditis elegans. Arch. Environ.Contam. Toxicol. 1998, 34, 259–264.

(28) Williams, P. L.; Dusenbery, D. B. Aquatic toxicity testing usingthe nematode Caenorhabditis elegans. Environ. Toxicol. Chem.1990, 9, 1285–1290.

(29) Dhawan, R.; Dusenbery, D. B.; Williams, P. L. Comparison oflethality, reproduction, and behavior as toxicological endpointsin the nematode Caenorhabditis elegans. J. Toxicol. Environ.Health., Part Aq 1999, 58, 451–462.

(30) Crane, M.; Handy, R. D.; Garrod, J.; Owen, R. Ecotoxicity testmethods and environmental hazard assessment for engineerednanoparticles. Ecotoxicology 2008, 17, 421–437.

(31) Bae, E. J.; Lee, J. J.; Kim, Y. H.; Choi, K. H.; Yi, J. H. Physico-chemical Analysis of Silver Nanoparticle Suspension for Nano-toxicity Evaluation. J. Nanopart. Res. 2009, . submitted.

(32) Hu, M.; Novo, C.; Funston, A.; Wang, H.; Staleva, H.; Zou, S.;Mulvaney, P.; Xia, Y.; Hartland, G. V. Dark-field microscopystudies of single metal nanoparticles: understanding the factorsthat influence the linewidth of the localized surface plasmonresonance. J. Mater. Chem. 2008, 18, 1949–1960.

(33) Swain, S. C.; Keusekotten, K.; Baumeister, R.; Sturzenbaum, S. R.C. elegans metallothioneins: new insights into the phenotypiceffects of cadmium toxicosis. J. Mol. Biol. 2004, 341, 951–959.

(34) Williams, P. L.; Ma, H.; Glenn, T. C.; Jagoe, C. H.; Jones, K. L.A transgenic strain of the nematode Caenorhabditis elegans asa biomonitor for heavy metal contamination. Environ. Toxicol.Chem. 2009, DOI: 10.1897/08-496.1.

(35) Choi, O.; Hu, Z. Size dependent and reactive oxygen speciesrelated nanosilver toxicity to nitrifying bacteria. Environ. Sci.Technol. 2008, 42, 4583–4588.

(36) Kim, D. W.; Hong, G. H.; Lee, H. H.; Choi, S. H.; Chun, B. G.;Won, C. K.; Hwang, I. K.; Won, M. H. Effect of colloidal silveragainst the cytotoxicity of hydrogen peroxide and naphthazarinon primary cultured cortical astrocytes. Int. J. Neurosci 2007,117, 387–400.

(37) Hwang, E. T.; Lee, J. H.; Chae, Y. J.; Kim, Y. S.; Kim, B. C.; Sang,B. I.; Gu, M. B. Analysis of the toxic mode of action of silvernanoparticles using stress-specific bioluminescent bacteria.Small 2008, 4, 746–750.

(38) Riddle, D. L.; Albert, P. S. Genetic and Environmental Regulationof Dauer Larva Development in C. elegans II; Cold Spring HarborLaboratory Press.: New York, 1997; pp. 739-768.

(39) Antebi, A.; Culotti, J. G.; Hedgecock, E. M. daf-12 regulatesdevelopmental age and the dauer alternative in Caenorhabditiselegans. Development 1998, 125, 1191–1205.

(40) Birnby, D. A.; Link, E. L.; Vowels, J. J.; Tian, H. Z.; Colacurcio,P. L.; Thomas, J. H. A transmembrane guanylyl cyclase (DAF-11) and Hsp90 (DAF-21) regulate a common set of chemosensorybehaviors in Caenorhabditis elegans. Genetics 2000, 155, 85–104.

(41) Inoue, T.; Hirata, K.; Kuwana, Y.; Fujita, M.; Miwa, J.; Roy, R.;Yamaguchi, Y. Cell cycle control by daf-21/Hsp90 at the firstmeiotic prophase/metaphase boundary during oogenesis inCaenorhabditis elegans. Dev., Growth. Differ. 2006, 48, 25–32.

(42) Newman, M. C. Population Ecotoxicology; John Wiley and SonsInc: Chichester, UK, 2001.

(43) Crane, M.; Handy, R. D.; Garrod, J.; Owen, R. Ecotoxicity testmethods and environmental hazard assessment for engineerednanoparticles. Ecotoxicology 2008, 17, 421–437.

(44) Desalvo, M. K.; Voolstra, C. R.; Sunagawa, S.; Schwarz, J. A.;Stillman, J. H.; Coffroth, M. A.; Szmant, A. M.; Medina, M.Differential gene expression during thermal stress and bleachingin the Caribbean coral Montastraea faveolata. Mol. Ecol. 2008,17, 3952–3971.

(45) Ji, J. H.; Jung, J. H.; Kim, S. S.; Yoon, J. U.; Park, J. D.; Choi, B. S.;Chung, Y. H.; Kwon, I. H.; Jeong, J.; Han, B. S.; Shin, J. H.; Sung,J. H.; Song, K. S.; Yu, I. J. Twenty-eight-day inhalation toxicitystudy of silver nanoparticles in Sprague-Dawley rats. InhalationToxicol. 2007, 19, 857–871.

(46) Hidalgo, E.; Domınguez, C. Study of cytotoxicity mechanismsof silver nitrate in human dermal fibroblasts. Toxicol. Lett. 1998,98, 169–179.

(47) Clement, J. L.; Jarrett, P. S. Antibacterial silver. Met.-Based Drugs1994, 1, 467–482.

(48) Silver, S. Bacterial resistances to toxic metal ions-A review. Gene1996, 179, 9–19.

(49) Asharani, P. V.; Wu, Y. L.; Gong, Z.; Valiyaveettil, S. Toxicity ofsilver nanoparticles in zebrafish models. Nanotechnology 2008,19, 255102–255110.

ES803477U


ecotoxicity of silver nanoparticles introduction on the...

Documents