aquatic toxicology - mcgill universitybiology.mcgill.ca/faculty/cristescu/articles/chain et al....

Contents lists available at ScienceDirect

Aquatic Toxicology

journal homepage: www.elsevier.com/locate/aqtox

Variation in transcriptional responses to copper exposure across Daphniapulex lineages

Frédéric J.J. Chaina,b,⁎, Sarah Finlaysona, Teresa Creasec, Melania Cristescuaa Department of Biology, McGill University, QC, H3A 1B1, CanadabDepartment of Biological Sciences, University of Massachusetts Lowell, MA, 01854, USAc Department of Integrative Biology, University of Guelph, ON, N1G 2W1, Canada

A R T I C L E I N F O

Keywords:DaphniaCopperToxicityRNA-seqTranscriptomicsIntraspecific variationEcotoxicogenomics

A B S T R A C T

Copper pollution is pervasive in aquatic habitats and is particularly harmful to invertebrates sensitive to en-vironmental changes such as Daphnia pulex. Mechanisms of toxicity and tolerance to copper are not well un-derstood. We used RNA-sequencing to investigate these mechanisms in three genetically distinct D. pulex clonallineages with different histories of copper exposure. Upregulated genes after copper exposure were enrichedwith Gene Ontology (GO) categories involved in digestion, molting and growth, whereas downregulated genesafter copper exposure were enriched in the metal-regulatory system, immune response and epigenetic mod-ifications. The three D. pulex clones in our study show largely similar transcriptional patterns in response tocopper, with only a total of twenty genes differentially expressed in a single clonal lineages. We also detectedlower relative expression of some genes known to be important for copper tolerance, metallothionein and glu-tathione-S-transferase, in a sensitive lineage sampled from an uncontaminated habitat. Daphnia-specific genes(without orthologs outside the genus) and Daphnia-specific duplications (genes duplicated in the Daphnialineage) were overrepresented in differentially expressed genes, highlighting an important role for newlyemerged genes in tolerating environmental stressors. The results indicate that the D. pulex lineages tested in thisstudy generally respond to copper stress using the same major pathways, but that the more resistant clone withprevious copper exposure might be better able to regulate key genes. This finding highlights the importantnuances in gene expression among clones, shaped by historical exposure and influencing copper tolerance.

1. Introduction

Copper (Cu) pollution in aquatic ecosystems is a growing concernworldwide because of the ubiquitous use of Cu in industrial manu-facturing, agriculture and Cu mining practices (Järup, 2003; Zhu et al.,2014). Although Cu is essential for life and has an important function asa cofactor for many enzymes (Zhu et al., 2014), particularly those in-volved in hemoglobin synthesis (Lee et al., 1968), it is known to behighly toxic in high concentrations (Heugens et al., 2001). Cu can havesynergistic effects with other stressors (Heugens et al., 2001), canbioaccumulate (Zhou et al., 2008), and can have negative impacts ongrowth, reproduction and immune function of aquatic invertebrateseven at sublethal levels (De Schamphelaere et al., 2007; Poynton et al.,2007). These adverse effects cannot be easily mitigated, illustrating theneed for robust monitoring of aquatic ecosystems. In the genomic age,biomonitoring can take a bottom up approach, such as investigatingchanges in gene expression in response to environmental stressors toextrapolate mechanistic information.

Over the last decades, ecotoxicological approaches have been usedto assess the effects of excess Cu on a variety of aquatic species.Traditional studies have investigated survivorship and reproduction,however using these endpoints is inefficient for routine analysis(Robbens et al., 2007) and does not provide mechanistic information(Fedorenkova et al., 2010). The integration of genomic tools thatmeasure gene expression has proven to be valuable for ecotoxicology,as these tools provide information about the mechanism of toxicity on amolecular level that was missing previously (Robbens et al., 2007).Despite the growing interest in gene expression responses to Cu stress,the mechanisms underlying the toxic mode of action as well as themechanism of defense against Cu toxicity is still not well understood(Soetaert et al., 2007). Moreover, we do not understand how geneticbackground and history of exposure to Cu can influence the response ofan organism.

https://doi.org/10.1016/j.aquatox.2019.02.016Received 18 October 2018; Received in revised form 19 February 2019; Accepted 19 February 2019

⁎ Corresponding author.

$TXDWLF�7R[LFRORJ\��²��

$YDLODEOH�RQOLQH��)HEUXDU\��;��(OVHYLHU�%�9��$OO�ULJKWV�UHVHUYHG�

7

1.1. From gene expression studies to ecotoxicogenomics

There has been growing interest in using gene expression profilingto gain mechanistic information on Cu toxicity. Microarray gene ex-pression studies have shown that heavy metals produce distinct geneexpression profiles by metal, dosage, duration of exposure (Poyntonet al., 2007; Poynton et al., 2008a, b; Shaw et al., 2007), and life stage(Muyssen and Janssen, 2007). The best-known mechanism of toleranceis the production of Metallothionein proteins that bind to extraneousmetal ions. However, the mechanism of tolerance has been hypothe-sized to involve more than just a single important gene (Van Straalenet al., 2011; Janssens et al., 2009). Knowledge of genome-wide ex-pression patterns rather than select candidate genes allows us tobroaden our understanding of gene pathways involved in Cu tolerance.

Ecotoxicogenomic approaches have been used in many taxa un-dergoing metal exposure to investigate toxic modes of action, detectbiomarkers of exposure, and integrate this information into biomoni-toring programs (Jamers et al., 2013; Kim et al., 2013; Mussali-galanteet al., 2013; Koedrith et al., 2013). Mode of action assays have beenpaired with traditional toxicology tests to provide both genomic andphenotypic evidence for risk assessment reviews (Wilson et al., 2013;reviewed by Waters and Fostel, 2004). New genomic tools such astranscriptomic profiling offer great advantages that can be com-plementary to traditional toxicology endpoints such as survivorship andreproduction. These genomic methods provide faster, reproducible re-sponses and allow inferences about the specific pathways involved inthe mechanism of toxicity as well as pathways involved in coping me-chanisms (Robbens et al., 2007).

1.2. Daphnia for ecotoxicogenomics

Daphnia are micro-crustaceans found in lakes and ponds across theworld and are an integral part of freshwater ecosystems. As such,Daphnia is an important aquatic invertebrate for risk assessment and isconsidered to be a sentinel organism for environmental issues(Schindler, 1987). Daphnia species are widely studied in ecology (Sedaand Petrusek, 2011) and ecotoxicology (Altshuler et al., 2011), andhave been used to develop environmental regulations (Le et al., 2016).This provides a rich knowledge base for putting ecotoxicogenomicstudies into a broader ecological context. Ecotoxicogenomic studies onDaphnia pulex are made possible by the availability of a genome se-quence; D. pulex was the first crustacean to have its genome sequenced.D. pulex is considered to have an “ecologically responsive genome”;Colbourne et al. (2011) suggested that about one third of all D. pulexgenes are sensitive to environmental changes and many of these geneshave no homologs in closely related species. This allows for discovery offunctions in responsive genes such as the newly annotated metallothio-nein genes (Shaw et al., 2007). Subsequent follow-up functional assayscan also be used to discover novel functions. Another reason Daphniaare often used in ecotoxicology studies is their clonal mode of re-production, effectively minimizing variation among individuals com-pared to other species (Haap and Köhler, 2009). However, given thewide range of toxic stress tolerance between Daphnia clones in the samespecies (Barata et al., 2002), it is essential to consider genetic variationwhen carrying out ecotoxicogenomic studies.

1.3. Implications of genetic variation in toxicology

Understanding the link between genetic variation and gene ex-pression response to a stressor has been minimally explored. Differencesin stressor response between phenotypes have been shown in ar-thropods, such as differences in insecticide tolerance (McKenzie andYen, 1995), which suggest different genetic mechanisms of tolerance.However, there has been little investigation using gene expressionanalysis. In Daphnia, the use of genetically different clonal lineageslocally adapted to different lakes or ponds leads to much variation in

toxicity test results between laboratories (Barata et al., 2000) as well asfield studies (Haap and Köhler, 2009). Given the widespread use ofDaphnia in ecotoxicology, and the broad applications of these studies inrisk assessment and environmental regulation, it is important to addressthe effect of genetic variation and gene expression response (Baird andBarata, 1998). Moreover, understanding the interaction between ge-netic variation and gene expression has applications beyond ecotox-icogenomics and can advance areas such as laboratory use of micro-biota strains (Kvitek et al., 2008) and human health (Aardema andMacGregor, 2002).

The degree to which gene expression patterns under metal exposurevary across Daphnia clonal lineages (interclonal variation) remains littleunderstood. Genetic variation between clones has been studied mainlyin terms of fitness and biokinetic parameters (Muyssen et al., 2010; DeConinck et al., 2013). These studies document differences in physiolo-gical responses (growth, ingestion rates, energy reserves and electrontransport activity) between clones and thus provide a basis to in-vestigate underlying gene expression. Gene expression studies in-corporating different Daphnia magna lineages have mainly focused oninterclonal differences in expression of a few genes before and afteracquired tolerance (multi-generational) to cadmium (Cd) stress (Haapand Köhler, 2009; Haap et al., 2016). A recent exposure study foundgenotype by environment interactions for numerous environmentalstressors in D. magna, which was largely driven by lineage-specificgenes only found in crustaceans (Orsini et al., 2018). However, studiesare lacking on the genetic response of Daphnia to Cu stress, and whetherthis involves common stress-response pathways or lineage-specificgenes. Similar to the work on metal homeostasis and tolerance con-ducted on hyper-tolerant plants (reviewed in Clemens, 2001), there is aneed to integrate genetic variation among Daphnia clones with geneexpression studies to identify genes and gene networks that can helpdisentangle complex mechanisms of toxicity and tolerance to Cu stress.

Despite the interest, the exact mechanism of toxicity has yet to befully understood in Daphnia (Soetaert et al., 2007), although the tran-scriptional effects of Cu toxicity have been predicted to include genesinvolved in digestion suppression, oxidative stress, immune suppressionand disruption of vital exoskeleton processes (Poynton et al., 2007).Our study uses RNA-sequencing to investigate differences in the wholetranscriptome of phenotypically and genotypically different D. pulexclones exposed to Cu. To explore how differences in response betweenclones may be important for Cu tolerance, we first determine biologicalpathways responsive to Cu exposure across all clones. In particular, wediscuss how our dataset fits with previously proposed effects of toxicityand candidate genes that are involved in metal tolerance. We furtherinvestigate how expression patterns differ between clonal lineages tomake inferences about Cu tolerance relative to genetic backgrounds andprevious exposures. We provide a list of genes that can subsequently beinvestigated in targeted mode of action studies.

2. Materials and methods

2.1. Daphnia clones

Daphnia pulex clones were isolated from three habitats (Table S1);Dump Pond (D) in Illinois and Solomon Pond (S) in Michigan, USA,which are undisrupted habitats not known to be contaminated withmetals, and from Kelley Lake (K) near Sudbury Ontario, Canada. Cucontamination in this lake was historically very high at 2500 μg/L in1965 according to the Ontario Waters Resource Commission (Johnsonand Owen, 1966), and has dropped to 100 μg/L 40 years later (City ofGreater Sudbury, 2001; Keller et al., 2004). Kelley Lake currently hasthe highest level of Cu contamination in the region. Daphnia cloneswere established from single-wild caught individuals and cultured inthe lab in FLAMES medium (Celis-Salgado et al., 2008) at 18 °C with a16:8 light-dark cycle for about one year prior to use in toxicity testing.

F.J.J. Chain, et al. $TXDWLF�7R[LFRORJ\��²��

��

2.2. Toxicity test

Preliminary toxicity assays were conducted to determine the re-lative tolerance of Daphnia clones (based on LC50) and to determine theconcentration to use in the subsequent Cu exposures for gene expressionanalysis. After an initial range-finding test indicated an average LC50above 150 μg/L, we performed 48 h toxicity tests by adding total Cuconcentrations from 150 to 186 μg/L in 6 μg/L increments to FLAMESmedium (contains 0.2842 μg/mL of copper; Celis-Salgado et al., 2008).Source Daphnia were first reared in control conditions (FLAMESmedium with no added Cu) for at least 3 generations to control formaternal effects. For each Cu concentration, three replicates of 5 neo-nates from each lineage were randomly selected for exposure to Cu, andthese were all run at the same time. The exposure involved the transferto an intermediate vial with 30mL of media containing the targetconcentration of Cu, and then transferred from the intermediate to anexperimental vial also containing 30mL of media at the target Cuconcentration. The intermediate vial was used to minimize the transferof the original Cu-free medium to the experimental vials. The neonates,each less than 48 h old at the time of exposure, were not fed and theirsurvivorship was observed after 48 h. Neonates were considered to bedead if they were immobile and no movement was observed after lightagitation with a pipette. The Lethal Concentration at 50% mortality(LC50) was calculated with a Probit analysis (Finney, 1952) using the“MASS” package (Venables and Ripley, 2002) in R with the dose.pfunction.

2.3. Copper exposures for gene expression analysis

An acute Cu exposure experiment was carried out for gene expres-sion analysis. The experiment involved exposing adult primiparous D.pulex from three clonal lineages (D, S, and K) for 24 h to 90 μg/L of Cu.Five replicates of each clone were then placed in 50mL falcon tubessuspended from a plexiglass scaffold in each of three 8 L polypropylenetanks containing FLAMES medium with Cu added to a final con-centration of 90 μg/L, or in one 8 L tank containing only FLAMESmedium as a control. The bottom of each falcon tube was covered with300 μm nylon mesh to allow for free flow of media from the tank. Thelarge tanks simulated a homogenous environmental condition with in-dividual clones isolated in tubes but all exposed to the same conditions.The pH of each tank was monitored throughout the experiment. Tankswere placed on stir plates and constantly stirred at a rate of 120 rpm.Tanks were covered with a plexiglass lid to reduce evaporation andcross contamination.

Three species of algae, Ankistrodesmus, Pseudokirchneriella,Scenedesmus, were grown in Bold’s Basal Medium (Stein, 1979). Algaecultures were centrifuged, and the pellet was resuspended in FLAMESand added to the tanks to a final concentration of 20,000 cells/mL. Thealgae for the Cu treatment tanks were suspended in FLAMES containingCu. Tanks were kept in a controlled temperature chamber at 18 °C with16:8 light-dark cycle. The tissue collection was done by using a wide-bore pipette to transfer Daphnia individuals and a small amount ofmedium into a 1.5 mL microcentrifuge tube. The medium was thenremoved and 300 μl of RNAlater® (Qiagen) was added to the tube.Samples were placed at 4 °C for 24 h and then transferred to −80 °C forstorage.

2.4. RNA extraction

Replicates from the three clonal lineages were sequenced to eval-uate interclonal variation while optimizing sequencing depth persample, and included two control replicates from one tank and five Cureplicates spread across three tanks (Fig. 1). Whole Daphnia sampleswere stored in RNAlater® (up to 4 adult primiparous individuals persample in 300 μl) at −80 °C prior to extraction. To extract RNA, sam-ples were thawed on ice and the animals were transferred from the

RNAlater® to the column extraction buffer. Total RNA was then purifiedusing the RNeasy Plus Universal Mini Kit (Qiagen) as per the protocol.Modifications include homogenization of Daphnia with a sterile pestleand motor mixer for 2min, or until no particles remained visible. RNAwas aliquoted into 3 separate tubes for short term and long term storagein the freezer. Ethanol precipitation was performed to clean each RNAsample of unwanted salt contaminants leftover from extraction buffers.The following were added to a sample of 10 μl total RNA: 1 μl of thecarrier glycogen, 1.1 μl sodium acetate, and 28 μl 100% ethanol. Thiswas mixed gently by pipette and stored overnight at −80 °C. Sampleswere then centrifuged at 4 °C for 30min at 12,000 x g. The supernatantwas carefully discarded to avoid disturbing the pellet, then 500 μl offreshly made 75% ethanol was added and the pellet was washed byinverting the tube once. Samples were spun for 15min at 4 °C at 12,000x g and then allowed to stand at room temperature for 15min to dis-solve co-precipitated salts. Samples were centrifuged for 5min and thesupernatant was discarded. This washing was repeated twice. Pelletswere air-dried for 5min and then resuspended in 30 μl RNase-free waterfor 15min to ensure complete solubilization. The concentration of eachsample was determined by analyzing 1.5 μl of each sample on a Na-noDrop Spectrophotometer.

2.5. RNA sequencing

RNA Sequencing libraries were prepared from the three clones (D, S,K) using the NEXTflex™ Rapid Directional mRNA-Seq Kit (BiooScientific), which is compatible with the Illumina sequencing system.Two lanes of sequencing were performed on the Illumina HiSeq 2000 atGenome Quebec using Paired-End (PE) 100 base-pair reads, aftermultiplexing samples using unique sequencing barcodes (NEXTflex™RNA-Seq Barcodes). For the first lane of sequencing, 100 ng of totalRNA per sample was used as starting material. mRNA was selected fromeach sample using NEXTflex™ poly(A) Beads, and libraries were pre-pared as per the manufacturer's protocol. For the second lane of se-quencing, 200 ng of total RNA per sample was used to obtain a higherconcentration in the final libraries. Libraries were quantified at GenomeQuebec using an Agilent Bioanalyzer 2100. The transcriptomic se-quence data for all samples have been deposited in the EuropeanNucleotide Archive (ENA) under accession PRJEB28650.

2.6. Bioinformatics

Each FASTQ file was first checked for sequencing quality usingFastQC (Andrews, 2010; http://www.bioinformatics.babraham.ac.uk/projects/fastqc). The GenPipes RNA-Seq Pipeline developed at McGillUniversity and Génome Québec Innovation Centre (https://bitbucket.org/mugqic/mugqic_pipelines/src/master/pipelines/rnaseq) was usedto perform subsequent analyses. This included Trimmomatic (Bolgeret al., 2014) to trim low quality sequences and adaptors (ILLUMINA-CLIP:TruSeq3-PE.fa:2:30:15 TRAILING:30 MINLEN:32), STAR (Dobinet al., 2013) to map sequences to the Daphnia reference genome(Colbourne et al., 2011) using default parameters, Picard (http://broadinstitute.github.io/picard) to mark read duplicates using defaultparameters, RNA-SeQC (Deluca et al., 2012) to evaluate alignmentmetrics (-n 1000 and using the Ensembl v30 GTF file), the cufflinkspackage (Trapnell et al., 2010) to assemble the transcriptome usingdefault parameters, and the software packages EdgeR (Anders andHuber, 2010) and DESeq (Robinson et al., 2010) to perform differentialexpression analysis. Differential expression was first conducted com-paring all control samples versus all copper-exposed samples (regard-less of genotype), followed by three separate analyses of each clonallineage between treatments. A gene was determined to be differentiallyexpressed if the adjusted p-value after Benjamini-Hochberg correctionwas less than 0.05 in both DESeq and EdgeR analyses. Orthologs andparalogs were acquired from the EnsemblCompara framework (Vilellaet al., 2009) to determine recently duplicated genes and genes with


��

little conservation outside the Daphnia genus. Significant enrichment ofthese putatively recently emerged genes was determined using Pear-son’s Chi-squared test with Yates’ continuity correction in R.

Gene ontology (GO) enrichment analysis was performed with topGO(Alexa and Rahnenfuhrer, 2006) using GO terms acquired from En-sembl (v30). GO terms were identified as significantly enriched com-pared to the genomic background using a False Discovery Rate (FDR)-corrected weighted p-value<0.05 in topGO. Candidate genes involvedin Cu toxicity were identified based on previous studies on Daphniaexposed to Cu and other metals (Table S2). We also explored co-ex-pression of genes to generate modules of highly correlated genes (usingPearson correlation) by applying a weighted gene co-expression net-work analysis using the WGCNA package for R (Langfelder andHorvath, 2008). We used differentially expressed genes with adjusted p-values below 0.1 in at least one of DESeq or EdgeR for any of our fourcontrast analyses. One of the six control samples (T1-07S; Table S3) wasan outlier in the hierarchical clustering analysis with WGCNA, and wastherefore excluded from the network analysis following recommenda-tions in the WGCNA manual. However, this sample (T1-07S) follows thesame expression trends as the other control samples, and retaining thesample in the network analysis gives highly similar modules and net-work relationships (Figure S1). A heatmap showing the relative geneexpression levels of DE genes across samples was created in R usingheatmap.2 in the gplots package.

3. Results

3.1. Toxicity tests

Daphnia pulex clones used in the toxicity experiment included twolineages from non-contaminated environments (D and S clones) and onefrom a habitat with a long history of Cu contamination (K clone). Ourcopper toxicity tests show that the K clone has an LC50 of 181 μg/L Cucompared to 153 μg/L in D and 169 μg/L in S, suggesting that prior Cuexposure in K has led to increased Cu tolerance (Table S1). The final Cuconcentration used in the exposure experiment to measure gene ex-pression responses was 90 μg/L, corresponding to ˜49%, 53% and 59%of the LC50 for clones K, S and D, respectively.

3.2. Sequencing and mapping

A total of 245,671,301 raw paired-end reads were obtained from 21sequenced RNA libraries, which included 6 control samples (two fromeach clonal lineage from one tank) and 15 Cu-exposed samples (5 fromeach clone, spread among 3 tanks; Fig. 1). The average percentage ofreads that mapped to the reference genome was 89% (ranging from 84to 92%). The mean number of reads per gene was 493 (ranging from319 to 643). This depth is considered acceptable for finding differen-tially expressed genes that are not relatively highly expressed (Lei et al.,2015). The average number of expressed genes in a sample was 17,128(ranging from 15,134 to 17,869). Summaries are found in Table S3.

3.3. Differential gene expression under Cu exposure

Comparing all Cu exposed samples (n= 15) with all of the controlsamples (n= 6) resulted in 206 significantly differentially expressed(DE) genes (Figure S2); 41 were downregulated and 166 were upre-gulated under Cu exposure (Table S4). Six genes were differentiallyexpressed at p < 0.01 and had a fold change four times that of thecontrol group, highlighting them as strongly responsive to acute Cuexposure (Table 1). We identified gene ontology (GO) annotations formore than half (135) of the 206 DE genes. The major GO categories inwhich there are more DE genes than expected include metabolism,metal binding and transport, and peptidase activity (Fig. 2). Genesupregulated under Cu exposure were significantly enriched with thefunctional categories proteolysis, serine-type endopeptidase activity,chitin binding and metabolism, and metallocarboxypeptidase activity.Downregulated genes were enriched with the GO terms carbonate de-hydratase activity, extracellular matrix structural constituent, struc-tural constituent of cuticle, carbohydrate binding, and one-carbonmetabolic process (Table 2).

A co-expression analysis was performed using a more liberal dif-ferential expression threshold (p-value< 0.1; see Methods) to clustergenes into modules with similar expression profiles across samples. Thisanalysis grouped 600 protein-coding genes into 5 modules, three ofwhich (M1, M2 and M3) consisted of genes generally upregulated in Cu-exposed samples compared to control samples, and two of which (M4and M5) included genes that were mainly downregulated in Cu-exposedsamples (Fig. 3A). Module M1 corresponded to the significantly upre-gulated DE genes (p < 0.05), module M4 corresponded to the sig-nificantly downregulated genes, whereas most of the genes with higheradjusted p-values (> 0.05) were found in module M5 (Fig. 3B). The co-expression analysis concurred with the overall patterns of differentialgene expression and functional enrichment analyses. The gene ontologyenrichment of genes in module M1 (upregulated genes) suggests thatgenes important in the structure of the exoskeleton (chitin) and proteincatabolism (proteolysis, serine-type endopeptidase activity, metallo-carboxypeptidase activity) potentially interact in response to Cu ex-posure (Table S5). The gene ontology enrichment analysis of moduleM4 reflected the results of downregulated genes (see Table 2 and TableS5), whereas there were no enriched functional categories in the otherthree modules (M2, M3 and M5).

Using orthology and paralogy information from each expressedgene, we found an enrichment of Daphnia-specific genes (those in D.pulex and/or in D. magna with no other ortholog) and of recently du-plicated genes (inferred to have been duplicated in Daphnia since thedivergence with non-Daphnia species) among DE genes; 45% of the DEgenes had no detectable orthologs outside of Daphnia (p < 0.001) and49% of the DE genes had Daphnia-specific duplicate genes (p < 0.001).Genes with no detectable orthologs in non-Daphnia genomes contributeto both upregulated genes (45%) and downregulated (43%) genes inCu-exposed samples, whereas Daphnia-specific duplications pre-ferentially contribute to downregulated genes (64%) compared to

Fig. 1. The sampling design for RNA sequen-cing of Daphnia pulex clones exposed tocopper. Isolates from each of the three clones(D, K and S) were placed in individual tubeswithin one of four tanks: one control tank (T1)and three copper exposure tanks (T2 throughT4). The bottom of each tube was coveredwith mesh to allow for free flow of media.Differential gene expression analyses wereperformed by grouping all copper treatmentstogether compared to all controls together, aswell as each clone (5 samples) compared to itsrespective control (2 samples) separately, re-sulting in four analyses: all clones, clone D,clone K and clone S.


��

upregulated genes (45%). Daphnia-specific duplications make up 70%of the 106 genes in module M4 (p < 0.0001), the main module con-taining downregulated genes.

3.4. Clone specific patterns of gene expression

Given its long-term historical exposure to Cu, it is possible to in-vestigate whether clone K responds differently to Cu exposure thanclones D and S at the transcriptional level. To this end, differentialexpression analysis using DESeq and EdgeR was also performed for eachclone individually (Cu treatment n= 5, control n= 2 per clone), re-sulting in 2 uniquely upregulated genes in the K clone (out of 37 totalDE genes for the K clone), 14 (8 downregulated, 6 upregulated) in the Sclone (out of 43 total DE genes for the S clone), and 4 (3 downregulated,1 upregulated) in the D clone (out of 33 total DE genes for the D clone)(Figure S3). More than two thirds of all DE genes from individual cloneanalyses overlapped with the global comparison between Cu exposedand controls suggesting most responses to Cu are shared among clones.DE genes specific to clone S appear to be mainly driven by clonal dif-ferences in controls rather than in copper exposed individuals (Fig. 4;Figure S4).

Each one of the 20 clone-specific DE genes was either a Daphnia-specific gene or a Daphnia-specific duplicate gene. The functions ofmany of the genes are not known, including both DE genes unique tothe K clone (Table S7). The unique DE genes in the D clone include onedownregulated gene related to hemoglobin activity, one downregulatedgene encoding an ATPase activity involved in ion transport, as well asan upregulated metallothionein gene (Table S8). Nine genes were dif-ferentially expressed (6 downregulated, 3 upregulated) in each in-dividual clone as well as in all clones grouped together (Table 3). The Kclone shared 65% of its DE genes with the other clones, compared to58% for D and 37% for S (Figure S3).

3.5. Candidate genes for copper tolerance

Metallothionein proteins function to mediate metal homeostasis anddetoxification in invertebrates (Janssens et al., 2009). Of the five an-notated metallothionein genes in D. pulex (Shaw et al., 2007; Asselmanet al., 2012), only the metallothionein 1 (mt1) gene family wasTa

ble1

Elev

endifferen

tially

expressedge

neswith

sign

ficant

(p<

0.01

)logfold

chan

ge(log

|FC|

>4)

inDap

hnia

pulexclon

esexpo

sedto

copp

er.C

olum

nsindicate

Ensembl

gene

ID,log

fold

chan

geforallc

lone

sgrou

ped

toge

ther,P

-value

foran

alysis

usingDES

eqan

dEd

geR,

func

tiona

lann

otations

basedon

GO,InterProScan

(predicted

func

tion;

integrated

sign

atures)an

dBL

AST

p(for

thosethat

didno

tha

vean

yGO

anno

tatio

n).

Gen

eID

Allclon

esgrou

ped

GO

anno

tatio

nInterPro

scan

NCB

IBL

AST

p

LogFC

DES

eqEd

geR

2250

09−4.74

6.80

E-30

7.20

E-31

Carboh

ydrate

bind

ing

C-type

lectin

1041

67−5.77

59.40

E-27

3.90

E-36

Carboh

ydrate

bind

ing

C-type

lectin

3193

41−5.91

81.50

E-05

3.10

E-21

Carboh

ydrate

bind

ing

C-type

lectin

1099

80−4.82

60.00

094

1.00

E-12

Proteinbind

ing

Caprin,C

1qdo

main,

Tumou

rne

crosis

factor-like

domain

1014

726.51

79.60

E-13

1.20

E-15

Growth

factor

activ

ity;P

ositive

regu

latio

nof

celldivision

;Mem

bran

eCy

stine-kn

otcytokine

,PDGF/

VEG

Fdo

main

2474

414.04

8.10

E-05

0.00

011

G-protein

coup

ledreceptor

sign

allin

gpa

thway

;Integ

ralc

ompo

nent

ofmem

bran

eG

protein-co

upledreceptor

(rho

dopsin-like

),GPC

R,rhod

opsin-

like,

7TM

2225

296.87

88.00

E-06

3.60

E-10

NA

Uninteg

ratedsign

atures

(Non

e)Simila

rto

D.p

ulex

hypo

thetical

proteins,n

okn

own

func

tion

3003

827.72

80.00

662.00

E-07

NA

Uninteg

ratedsign

atures

(Non

e)Simila

rto

D.p

ulex

hypo

thetical

proteins,n

okn

own

func

tion

3134

285.00

15.80

E-06

1.50

E-10

NA

Uninteg

ratedsign

atures

(Non

e)Ch

itina

se,P

utativeglycosyl

hydrolases

3173

484.73

10.00

272.20

E-08

NA

Uninteg

ratedsign

atures

(Non

e)Simila

rto

D.p

ulex

hypo

thetical

proteins,n

okn

own

func

tion

3321

195.33

0.00

40.00

017

NA

Retrotranspo

songa

gdo

main;

Simila

rto

D.p

ulex

hypo

thetical

proteins,n

okn

own

func

tion

Fig. 2. Functional enrichment for 206 differentially expressed (DE) genes in the“all clones” grouping of Daphnia pulex clones exposed to copper. All gene on-tology (GO) categories shown were observed among DE genes at least threetimes more frequently than expected based on the functional annotations of the17,508 expressed genes in the dataset.


��

significantly upregulated under copper exposure when grouping allclones (Fig. 5A and 5B); mt1a was significantly upregulated when allclones were grouped together (pvalDESeq = 0.00056, pvaledgeR = 1.80E-06) but not in clone-specific DE analyses, and mt1b was significantlyupregulated in the D clone (pvalDESeq = 0.0013, pvaledgeR = 0.00033)and in the K clone in the EdgeR analysis (pvalDESeq = 0.13; pvaledgeR =0.027), but not in the S clone (Fig. 5B). The metallothionein genes mt2and mt4, which are known to respond to chronic Cu exposure inDaphnia, were not differentially expressed in any DE analysis (Fig. 5C).

Proteins that have been documented to interact withMetallothioneins include the antioxidant, Glutathione transferase andthe metalloprotein Metallocarboxypeptidase (Yadav, 2010). The

glutathione transferase (gst) gene was significantly upregulated when allclones were grouped together (pvalDESeq = 8.20E-16, pvaledgeR =1.20E-14), in the D clone (pvalDESeq = 8.90E-05, pvaledgeR = 0.0027)and the K clone (pvalDESeq = 4.00E-10, pvaledgeR = 2.60E-08), but notin the S clone (pvalDESeq = 0.73, pvaledgeR = 0.18) (Fig. 5D). Me-tallocarboxypeptidase is encoded by two Daphnia genes (DAPPUDR-AFT_117945 and DAPPUDRAFT_195011). Both genes were significantlyupregulated when all clones were grouped together (117945: pvalDESeq= 0.027, pvaledgeR = 0.0019, 195011: pvalDESeq = 0.00053, pvaledgeR= 0.00028), but not in any of the individual clones (Fig. 5E and 5 F).

Table 2Gene Ontology (GO) enriched terms of downregulated and upregulated genes found using topGO enrichment analysis of 206 differentially expressed genes identifiedwhen comparing all copper-exposed Daphnia pulex samples to all controls. The p-value for each GO term was corrected (COR) for multiple testing using the falsediscovery rate (FDR) for both Fisher’s exact test and the weight test statistics implemented in topGO.

UpregulatedGO ID Term Annotated Significant Expected Fisher COR Weight COR0006508 Proteolysis 775 48 13.65 3.12E-12 1.57E-100004252 Serine-type endopeptidase activity 264 23 3.61 3.28E-10 1.64E-090008061 Chitin binding 113 15 1.55 6.29E-09 2.20E-080006030 Chitin metabolic process 121 15 2.13 1.56E-06 1.28E-050004181 Metallocarboxypeptidase activity 38 7 0.52 8.59E-05 0.00031512DownregulatedGO ID Term Annotated Significant Expected Fisher COR Weight COR0042302 Structural constituent of cuticle 303 10 1.68 0.00404999 0.008099990004089 Carbonate dehydratase activity 31 4 0.17 0.01107419 0.016611280006730 One-carbon metabolic process 32 4 0.15 0.02565011 0.025650110030246 Carbohydrate binding 127 6 0.71 0.02422175 0.032295670005201 Extracellular matrix structural constituent 47 4 0.26 0.02734256 0.04404503

Fig. 3. Gene expression patterns between con-trol and copper exposed Daphnia pulex clones.(A) Gene co-expression network graph withfive modules: modules M1 (green), M2 (red)and M3 (yellow) mostly consist of genes upre-gulated in copper-exposed samples comparedto control samples, whereas modules M4 (blue)and M5 (grey) are genes that were down-regulated in copper-exposed samples or notdifferentially expressed. M1 consists of sig-nificantly upregulated genes, and M4 consistsof significantly downregulated genes. (B)Boxplots show the expression distribution ofgenes in the M1 and M4 modules by replicateand by clone (D is red, K is green, S is blue).The dashed line indicates the mean expressionin control and copper treatments.


��

4. Discussion

We investigated patterns of gene expression in response to highlevels of Cu exposure among three Daphnia pulex clonal lineages thatvary in historical Cu exposure. We found that the majority of differ-entially expressed genes were shared across clonal lineages, and involveseveral biological pathways generally associated with metal toxicity.Overall, there was an enrichment of genes with no known orthologs

outside of Daphnia as well as Daphnia-specific gene duplications amongdifferentially expressed genes. Below we discuss the relevance of clonalvariation and how our results compare with previously proposed toxiceffects of Cu.

4.1. Interclonal variation in gene expression response to Cu

Interclonal differences in toxicology have not been well studied in

Fig. 4. Heatmap of DE gene expression acrosssamples. Each row represents the relative ex-pression (Z-score) of a DE gene in each sample.Black boxes indicate whether the DE gene isclone-specific or found in the global compar-ison and at least one clone-specific compar-ison. Gene IDs and their co-expression modulesare shown on the right in green (M1), red (M2),yellow (M3), blue (M4), or grey (M5).

Table 3Gene Ontology annotations for genes differentially expressed (DE) in all four Daphnia pulex clone and group comparisons. Nine genes were DE in all comparisons (allclones together and each clone individually). Columns indicate Ensembl gene ID, GO annotation ID, Functional description and the log fold change of each gene foreach clone separately.

Gene ID GO ID GO Description logFC D logFC S logFC K

229607 0016998 Cell wall macromolecule catabolic process 3.592 2.89 3.3210009253 Peptidoglycan catabolic process0003796 Lysozyme activity

220200 0004869 Cysteine-type endopeptidase inhibitor activity 4.068 3.942 3.230043086 Negative regulation of catalytic activity0010951 Negative regulation of endopeptidase activity0010466 Negative regulation of peptidase activity0050790 Regulation of catalytic activity0052548 Regulation of endopeptidase activity0052547 Regulation of peptidase activity

319612 NA NA 2.817 3.46 4.123226732 0071918 Urea transmembrane transport −3.046 −2.491 −3.278

0015204 Urea transmembrane transporter activity0016021 Integral component of membrane

104167 0030246 Carbohydrate binding −5.672 −5.642 −5.976221339 0030246 Carbohydrate binding −3.456 −3.482 −3.545225009 0030246 Carbohydrate binding −5.378 −4.457 −4.686319341 0030246 Carbohydrate binding −5.076 −6.052 −6.153303054 NA NA −3.145 −3.099 −3.053


��

Fig. 5. Box plots of expression level for metallothionein, glutathione-S-transferase and metallocarboxypeptidase in three Daphnia pulex clones. The y-axis is expression infragments per kilobase of transcript per million mapped reads (FPKM) for control (n= 2) and copper treatment (n= 5) samples from each clone (D in red, K in green,and S in blue). The bold line in each box indicates the median value, box edges are the 25th and 75th percentile, whiskers represent 1.5 x IQR (inter-quartile range),and values beyond the whiskers are marked as individual points. Asterisks above a single box represents differential expression (P < 0.05) between copper andcontrol samples for that particular clonal lineage, and asterisks above a bracket signifies differential expression between all copper samples and control samples. (A,B) The two duplicate genes encoding Metallothionein 1 (mt1): mt1a and mt1b are both differentially expressed when all clones are grouped together, and mt1b is alsosignificant for clone D. (C) The copper-specific metallothionein (mt2) gene does not show differential expression. (D) The gene encoding Glutathione-S-transferase wasdifferentially expressed in the D and K clones and when all clones were grouped together, but not in the S clone. (E, F) The two genes encodingMetallocarboxypeptidase (DAPPUDRAFT_117945 and DAPPUDRAFT_195011) were differentially expressed only when grouping all clones together.


��

Daphnia despite the implications for risk assessment (Baird and Barata,1998). A comparison of gene expression profiles between Daphniaclones using whole transcriptome sequencing has only recently beenconducted (Orsini et al., 2018), although previous studies have lookedat variation of candidate genes (Haap and Köhler, 2009; Haap et al.,2016) or of methylation patterns (Asselman et al., 2015). The fact thatall three D. pulex clones in our study, which vary in prior Cu exposure,show largely similar patterns of gene expression (upregulation of genesinvolved in digestion and molting, oxidative stress response and metalbinding; downregulation of certain immune and oxygen transportgenes) shows that the major Cu stress response pathways act similarlyin all clones, as might be expected for lineages of the same species.However, differences in expression levels of certain genes may be veryimportant overall if they have central roles in tolerating Cu stress and ifcertain clones have adapted to Cu due to long-term exposure. For ex-ample, the glutathione–S-transferase (gst) gene was upregulated in allclones but only half as high relative to the control in the S clone. Theenzyme encoded by this gene has been shown to play a pivotal role inthe ability of the organism to detoxify Cu (Hossain et al., 2012), andexpression of gst is often used as a biomarker for many toxins that areknown to cause oxidative stress (Cunha et al., 2007). Reduced upre-gulation of gst in the S clone possibly contributes to its Cu sensitivity ifthe encoded protein cannot help eliminate Cu from cells. Our resultssuggest the need to further investigate the role of gst expression in Cutoxicity in Daphnia, as well as understand how it relates to Cu sensitivityin different Daphnia clones. Interestingly, the K clone, which has ex-perienced historical Cu contamination, shared the highest proportion ofDE genes with the two other clones, whereas the expression patterns ofD and S were the least similar to one another. Instead of inducingstronger transcriptional responses to Cu, the previous exposure to Cu inthe K clone might initiate a more efficient or quicker response of thesame general Cu sensitive genes seen in the other clones that was notcaptured in our 24 h experiment.

The genes that were only differentially expressed in one clone areputatively recently emerged genes, and were either Daphnia-specificwith no known orthologs or Daphnia-specific duplications. Such genescan provide insight into how clones respond differently to acute Custress. The S clone shows downregulation of some uniquely DE genesencoding digestive enzymes and serine peptidases, and upregulation ofgenes encoding digestive enzymes and endopeptidase. The other sixunique DE genes in the S clone showed downregulation of structuralconstituents of the cuticle under Cu exposure, which is due to differentbasal gene expression between clones in the controls (Fig. 5). All clonesshowed elevated average expression of mt1 (both mt1a and mt1b) afterCu exposure, although significant upregulation was only found in the Dclone (in mt1b) and when all clones are considered together (Fig. 5).Only a hemoglobin complex gene was significantly downregulated inthe D clone, suggesting that oxygen transport was significantly im-paired in the D clone and not in the others. An ATPase activity gene wasalso significantly downregulated only in the D clone; ATPase activity isknown to be a mechanism of transporting Cu ions out of cells and or-ganelles in mammals (Dameron and Harrison, 1998). This could suggestthat the D clone is not able to pump Cu ions out of the cell as effectivelyas the other clones.

De Coninck et al. (2013) have cautioned against using single clonallineages to extrapolate findings to an entire species. The clone-specificDE genes we identified support this and provide candidate targets todetermine how each clone responds differently to Cu stress. Differentlineage sensitivities to Cu stress (and other metals) can have importantpopulation-level implications (Venâncio et al., 2016). Although therehas been research on phenotypic and fitness differences between clonallineages in response to metal stress, to our knowledge there have beenno large-scale gene expression studies in D. pulex that investigate dif-ferences between clonal lineages. However, a recent transcriptomeanalysis of environmental perturbations in D. magna revealed thatabout one third of the genes involved in response to numerous

environmental stressors are shared among genotypes, and that genesthat differ in expression based on genotype and condition are enrichedin Daphnia-specific genes (Orsini et al., 2018). This is consistent withthe idea that newly evolved lineage-specific genes play a role inadaptation to novel environments (Colbourne et al., 2011; Kondrashov,2012; Chain et al., 2014; Santos et al., 2017). Our study also shows thatDaphnia-specific genes, and particularly recently duplicated genes inDaphnia are significant contributors to expression responses to Cu ex-posure, and represent clone-specific differences. There is a need formore studies of this kind to understand the nuances in gene expressionpatterns between clonal lineages in Daphnia.

4.2. Co-expression analysis

The network analysis revealed five co-expressed gene modules, twoof which contain enriched gene ontology categories that highlight themajor pathways involved in response to Cu. The M1 module was en-riched with the gene functions chitin binding, chitin metabolic pro-cesses, digestive enzymes and peptidase activity (serine-type en-dopeptidase activity, metallocarboxypeptidase activity), suggestingthat genes involved in exoskeleton structure and digestion co-respondto Cu exposure. A potential site of activity of these genes is in theperitrophic membrane (PM), which is made of chitin in arthropod gutsand is an important structural and immunological defense against in-gested pathogens (Dinglasan et al., 2009). The thickness of the PM canbe regulated by chitinases to create a structural barrier for parasites orother molecules (Benedito Filho et al., 2002). Since we know there aregut-specific chitinases in other arthropods (Kramer et al., 2011; Shenand Jacobs-Lorena, 1997; Arakane et al., 2005; Dinglasan et al., 2009),it would be interesting to know if D. pulex is expressing a specificchitinase in response to Cu exposure. Beckerman et al. (2013) suggestedit may be developmentally advantageous to develop one physiologicalpathway in response to many types of stressors. It is also interesting tonote that other proteins with chitin-binding domains are not able tobind to chitin and thus cannot remodel the PM (Shi and Paskewitz,2004; Badariotti et al., 2007; McTaggart et al., 2009). Their function isyet to be determined, but it is thought they are predominantly active inthe gut and are involved in an immune-related response (Siva-Jothyet al., 2005).

Co-expression of downregulated genes in the M4 module are en-riched with the gene ontology categories structural constituent of cu-ticle, carbonate dehydratase activity, and collagen trimer, supportingthe hypothesis suggested by Engel and Brouwer (1987) that Me-tallothioneins, ecdysteroids and metallo-enzymes (including Carbonatedehydrogenase, also known as Carbonic anhydrase) all interact witheach other in the molt cycle of blue crabs (Callinectes sapidus). Thishypothesis is supported by studies in Daphnia (Bodar et al., 1990) andother crustaceans (Abidi et al., 2016). For example, metals such as Cuand Cd are known to disrupt Carbonate dehydrogenase enzymatic ac-tivity in crustaceans as well as in vitro erythrocyte bioassays (Vitaleet al., 1999; Lionetto et al., 2005). Studies have also reported that Cddelays molting in fiddler crabs (Weis, 1976) and increases time betweenmolting in Daphnia (Bodar et al., 1990). Genes encoding the structuralconstituent of the cuticle are downregulated in this co-expressionmodule with a P- value between 0.1 and 0.05, so they were not includedin our interpretation of differentially expressed genes (P < 0.05) inwhich exoskeletal genes were upregulated. This incongruence requiresfuture work to fully understand how dosage and duration factor into thecomplex relationship between the expression of exoskeletal genes andCu exposure.

Other downregulated genes that were enriched in the M4 moduleare associated with GO annotations for the one-carbon metabolic pro-cess, which is involved in epigenetic regulation of converting methylgroups from dietary sources to methyl donors that cause DNA methy-lation (Anderson et al., 2012). Downregulation of genes involved inmethylation activity suggests epigenetic modifications were beginning


��

to occur during Cu exposure. Changes in methylation patterns havebeen seen in D. magna when exposed to Zn, however Cd did not result inany changes (Vandegehuchte et al., 2009a,b). Although epigeneticregulation of gene expression in response to stressors and toxins hasbeen studied widely in mammals, there has been little work done tounderstand these processes in Daphnia (Asselman et al., 2015). How-ever, the downregulation of genes involved in the one-carbon metabolicprocess observed in this study suggest that epigenetic modificationsshould be taken into account in future ecotoxicogenomics studies(Vandegehuchte and Janssen, 2011).

Although our DE and network analyses reveal strong links to func-tions expected under copper stress, both false positives and negativesare possible with low sample sizes. In particular, the DE analyses ofindividual clonal lineages relies on two control samples, which couldexplain the low number of DE genes detected. The co-expression net-work analysis with WGCNA was performed with 20 samples, which is atthe low end of acceptable sample sizes. Furthermore, the use of adultDaphnia could introduce noise in the expression data and its analysisdue to genetic, epigenetic, or developmental differences among samples(Bruning et al., 2016). Our joint analysis on the effects of Cu using threedifferent genetic lineages acts to limit such confounding effects, but theDE genes identified need further experimental testing to support theirrole in Cu response.

4.3. Immune suppression

In this study, downregulated genes include six carbohydrate bindinggenes with C-type lectin domains, and one gene that is involved in adefense response to viruses suggesting suppression of the immunesystem in response to Cu exposure. Conversely, certain immune re-cognition genes were upregulated in this study including genes in-volved in scavenger receptor activity, response to a viral capsid andlysozyme activity. An increased level of such a function could plausiblydeplete energy reserves over a long period of time. Depletion of energyreserves with increasing nickel (Ni) concentration has been observed inD. magna (Vandenbrouck et al., 2009). If an organism cannot sustain aheightened state of immune response for a long time, increasing certainimmune functions may be a last resort to compensate for the action ofthe acute stressor.

Suppression of the immune system has been a proposed toxic effectspecific to Cu exposure (compared to other metals) based on reductionin blood cell count and activity in molluscs (Parry and Pipe, 2004) andincreased disease susceptibility (Yeh et al., 2004) in other organisms.There is still much to understand about the molecular mechanisms ofthe Daphnia immune system (Rozenberg et al., 2015), but a recent insilico identification of immune-related genes inferred by comparinghomologs in other arthropods has advanced our understanding(McTaggart et al., 2009; reviewed by Auld, 2014). In our study, genesthat encode C-type lectins responsible for recognition of pathogenswere downregulated. However, genes encoding other types of re-cognition proteins such as scavenger receptors and peptidoglycan cat-abolism were upregulated. In addition, chitinase activity, which is im-portant for molting but also has a role in the innate immune response,was elevated in all three clones after Cu exposure. Further functionalinformation about different types of recognition proteins and the effectsof Cu exposure will help us understand which specific part of the im-mune system is affected by excess Cu ions.

4.4. Oxidative stress

As seen in other studies of Daphnia (Poynton et al., 2007, 2008a,b;Shaw et al., 2007), we found that nine genes related to oxidation re-duction processes, response to oxidative stress, and peroxidase activitywere upregulated including gst. It has been shown that gst expression isregulated by oxidative stress (Casalino et al., 2004), which is a commoneffect of some heavy metals and toxins (Melegari et al., 2013). Cu is

known to cause reactive oxygen species when present in excess amounts(Kawakami et al., 2008) and produces mutagenic and carcinogenic by-products through redox reactions (Valko et al., 2005). Other knownbiomarkers of oxidative stress, including genes encoding superoxidedismutase and catalase activity (Ruas et al., 2008) were not differen-tially expressed in our study. One gene (DAPPUDRAFT_234836) relatedto oxygen binding and transport was downregulated. The upregulationof genes that are expected to respond as part of the antioxidant pathwayindicate that oxidative stress is a mode of action of Cu toxicity.

4.5. Digestion

In this study we observed upregulation of 27 genes involved inpeptidase activity such as serine type endopeptidase activity, whichaffects lipid metabolism, carbohydrate metabolism, and hydrolase ac-tivity (von Elert et al., 2004). We also saw upregulation of 28 genesinvolved in hydrolase activity. This suggests that some digestive pro-cesses have increased due to Cu exposure, although we also sawdownregulation of genes involved in lipid transport. These resultscontradict a proposed toxic effect of Cu involving disruption of diges-tive processes in invertebrates (Chen et al., 2002). Supporting evidenceincludes downregulation of genes encoding digestive enzymes (Poyntonet al., 2007, 2008a,b), reduction in feeding behaviour (De Coen andJanssen, 1998), and physiological evidence of necrosis of the hepato-pancreas in other crustaceans (Li et al., 2007).

Dynamic energy budget models (Nisbet et al., 2000) and laboratoryexperiments on D. magna (Nogueira et al., 2004) suggest that increasedenergy consumption to mitigate stressors draws energy away fromgrowth and reproduction. These increased metabolic demands to de-fend against Cu damage will lead to a reduction in energy reserves,which will decline further over time, as shown by a gradually wideninggap in brood production and size in D. magna compared to non-exposedindividuals (De Schamphelaere et al., 2007). An initial increase inmetabolic and digestive processes after exposure to Cu may serve toprovide energy to mitigate the immediate toxic effects of Cu if thestressor is severe enough. If the stressor is sublethal, these processesmay be downregulated to prioritize other pathways. The resulting in-crease or decrease in growth and brood production may be dictated bythe amount of energy that is ultimately used to deal with the stressor,and depends on factors such as concentration and duration of exposure(Bossuyt et al., 2005). Therefore we suggest that for the duration,concentration, and mode of exposure used in this study, disruption ofdigestion does not appear to be a toxic effect of Cu. In this case, up-regulation of genes involved in digestion could actually be protectingthe organism by providing energy for use in other pathways that defendagainst Cu toxicity.

4.6. Exoskeletal proteins

In our study, all 11 significantly DE genes related to chitin binding,chitinase activity and metabolic processes were upregulated. One ex-planation for this upregulation is that we used individuals activelypreparing to shed their exoskeleton (primiparous). Genes related toecdysis (molting) activity were upregulated in individuals exposed toCu compared to the controls, perhaps as a stress response to release thebrood as soon as possible; either in an effort to reduce metal exposure tothe brood, or as a self-preservation technique as molting can be used toeliminate metals in Daphnia (Muyssen and Janssen, 2002; Riddell et al.,2005). When D. pulex was exposed for 48 h to Cd, a more toxic metal,chitin-related genes were upregulated (Shaw et al., 2007). This was alsoseen in D. magna exposed to Cd (Connon et al., 2008). Another reasonfor upregulation of exoskeleton genes could be the concentration of thestressor, and therefore the severity. However, using lower concentra-tions of Cu exposure in D. magna (between 1 and 30 μg/L), Poyntonet al. (2008a,b) found that exoskeleton genes were downregulated,perhaps to prioritize more important functions. Copper exposure in an


��

amphipod revealed that chitinase is upregulated at low dissolved Cuexposure but downregulated at high dissolved Cu exposure and bothlow and high dietary Cu exposure (Hook et al., 2014), suggesting bothan effect of route of exposure and dose. We suggest that for the con-centration of Cu and primiparous state of the animals in this study,upregulation of exoskeletal genes may be advantageous for Daphnia andappears to be a means to eliminate excess Cu. But the genetic responsesand mechanism of toxicity appears to be idiosyncratic, depending onfactors other than genetic background such as the exposure pathwaysand concentrations.

4.7. Hemoglobin production

The effects of Cu on Hb production in Daphnia are not well under-stood. Dave (1984) observed a complex dose-response relationshipbetween Cu concentration and Hb production in D. magna wherein Hbwas decreased at low and high Cu concentrations but increased at in-termediate concentrations. In this study, expression of a gene involvedin oxygen binding, transport, hemoglobin complex, and heme-bindingdecreased under Cu exposure. This significant downregulation wasobserved when grouping all clones together, and in the clone with thelowest LC50 (D), but not in the other clones separately. We suggest thatthe downregulation of Hb genes and potential impairment of oxygenbinding and transportation is an interesting future target of study in thetoxicity of Cu in Daphnia, especially in light of the fact that Cu is es-sential for Hb synthesis. Conversely, other heme-binding protein en-coding genes associated with GO terms such as response to oxidativestress, were upregulated. Pleiotropy and epistasis might help explainthese complex relationships, and future studies will help understand therole of Hb in the context of Cu toxicity.

4.8. Metallothioneins

Metallothioneins are a class of proteins that bind specifically tometal ions to regulate inter-cellular concentrations of essential heavymetals, including Cu, and to mitigate the effects of free radicals.Incorporation into metalloproteins is a reversible process that is assistedby Glutathione (Maret, 1994; Jiang et al., 1998). Mt genes are found inall eukaryotes and some prokaryotes, often in multiple copies (Palmiter,1998), which may allow mt specialization for a particular heavy metal.D. pulex has five mt genes, as mt1 is duplicated. Mt2 and mt4 are knownto respond to Cu exposure (Asselman et al., 2013) while mt1 and mt3are known to respond to Cd exposure (Shaw et al., 2007).

The patterns of mt expression we observed do not align with pre-vious results; mt2 and mt4 were either expressed at the same level in thetreatment and control or slightly downregulated after Cu exposure. Incontrast, mt1a and mt1b were the only mt genes to be significantlyupregulated (mt1a only when grouping all clones together, mt1b only inthe D clone individually). Since Cd is a more toxic stressor than Cu, it ispossible that Cu becomes as potent a stressor as Cd at a certain con-centration, thus inducing expression of mt1. Asselman et al. (2013)found that chronic EC50 Cd exposure caused upregulation of mt1 (day4) in D. pulex followed by repression of that gene while the other Cd-responsive gene, mt4 was upregulated later in the exposure period (day8). This suggests different mt homologs may respond not only to specificmetals, but the response may also depend on exposure duration as wellas concentration. Furthermore, the promoter region of mt1a/b inDaphnia contains ecdysone-responsive elements (Asselman et al., 2013),consistent with Cu-specific mts playing a role in molting metabolismand changes in metal localization (Engel and Brouwer, 1987, 1991).Life stage of the organism can influence metal detoxification and mtexpression patterns, suggesting that the primiparous life stage at whichthe Daphnia in this study were exposed might not capture the full extentof genotype differences.

There are several hypotheses to explain the dynamic expressionpatterns of the D. pulex mt genes. These include a time-response

relationship between exposure and gene expression (as shown inAsselman et al., 2013), a dose-response relationship (Onosaka andCherian, 1981, reviewed in Amiard et al., 2006), or an acclimation-response, which can occur over a few hours to generations (Muyssenand Janssen, 2004). The regulation of the mt homologs likely involves acomplex interaction between promoter regions, which can differ innumber, location and sequence of metal response elements (MREs)(Asselman et al., 2013). According to the inhibition-induction me-chanism, the homolog most able to combat the stressor will be ex-pressed while the others will be repressed (Asselman et al., 2013). Thisis seen in other types of stress responses such as heat shock proteins(Franzellitti and Fabbri, 2005), the response elements of which are alsofound in the promoter region of the D. pulex mts. Chelation of extra-neous metal ions can occur after Metallothionein proteins are synthe-sized, thus continual transcription of the mt genes becomes unnecessaryover time (Amiard et al., 2006). As Asselman et al. (2013) suggest,cyclical-, time- and dose-dependent expression patterns betweenhomologs highlight the need for temporal comparisons of inductionbetween all the mt genes. Understanding the mechanism of activation ofthe mt genes in D. pulex can help understand changes in mt expression inresponse to different metal stressors.

5. Conclusions

We have identified over 200 genes whose expression responds toacute Cu stress in D. pulex. We have suggested several candidates forfollow-up assays that would enhance our knowledge of their involve-ment in toxic Cu response in this species. Although proposed majorpathways involved in toxicity were shared by all clonal lineages, ourcomparison of gene expression between clones shows that geneticbackground influences the expression patterns of genes responsive toacute Cu stress. Regulation of key genes, such as gst, differed betweenclones, cautioning the use of one lineage to draw conclusions for anentire population or species. This illustrates the need for future studiesto incorporate more genetic backgrounds to further understand thebreadth of variation in gene expression response to toxicity.

The results of our study have important implications for ecotox-icogenomics studies on environmental stress, specifically the role ofDaphnia as an ecotoxicogenomic model. We showed that previouslyproposed toxic effects of Cu exposure, including suppression of theimmune system and oxidative stress are important in some D. pulexlineages. Other proposed toxic effects such as disruption of molting anddigestion were not supported by our study. These differences raise in-teresting questions about the use of different doses, and the durationand mode of exposure between studies. Each study provides a ‘snapshot’ of the gene expression for those given conditions. Future studiesshould incorporate gene expression levels monitored from acute tochronic timeframes over varying dosages.

Contributions

SF conducted toxicity tests, gene expression experiments, and mo-lecular work. FC and SF analysed the RNA-sequencing data. FC and SFdrafted the manuscript. TC and MC contributed to the experimentaldesign and data analysis. All authors edited the manuscript.

Acknowledgements

We thank all graduate and undergraduate students who helped withlab work, in particular Piumi Abeynayaka for assistance with toxicitytests and Katie Millette for help with sampling. We would also like tothank Norman Yan and Martha Celis-Salgado for valuable advice on theexperimental design and Daniel Schoen and Rowan Barrett for pro-viding feedback on the project. This project was supported by an NSERCCREATE grant (397997-11) on Aquatic Ecosystem Health to MC andTC, and an NSERC CGS-M scholarship to SF.


��

Appendix A. Supplementary data

Supplementary material related to this article can be found, in theonline version, at doi:https://doi.org/10.1016/j.aquatox.2019.02.016.

References

Aardema, M.J., MacGregor, J.T., 2002. Toxicology and genetic toxicology in the new eraof “toxicogenomics”: impact of “-omics” technologies. Mutat. Res. Mol. Mech.Mutagen. 499, 13–25.

Abidi, S., Abbaci, K.T., Geffard, O., Boumaiza, M., Dumet, A., Garric, J., Mondy, N., 2016.Impact of cadmium on the ecdysteroids production in Gammarus fossarum.Ecotoxicology 25, 1–8.

Alexa, A., Rahnenfuhrer, J., 2006. topGO: Enrichment Analysis for Gene Ontology. RPackage Version 2.12.0.

Altshuler, I., Demiri, B., Xu, S., Constantin, A., Yan, N.D., Cristescu, M.E., 2011. An in-tegrated multi-disciplinary approach for studying multiple stressors in freshwaterecosystems: daphnia as a model organism. Integr. Comp. Biol. 51, 623–633.

Amiard, J.-C., Amiard-Triquet, C., Barka, S., Pellerin, J., Rainbow, P.S., 2006.Metallothioneins in aquatic invertebrates: their role in metal detoxification and theiruse as biomarkers. Aquat. Toxicol. 76, 160–202.

Anders, S., Huber, W., 2010. Differential expression analysis for sequence count data.Genome Biol. 11, R106.

Anderson, O.S., Sant, K.E., Dolinoy, D.C., 2012. Nutrition and epigenetics: an interplay ofdietary methyl donors, one-carbon metabolism and DNA methylation. J. Nutr.Biochem. 23, 853–859.

Andrews, S., 2010. FastQC: A Quality Control Tool for High Throughput Sequence Data.Available online at. http://www.bioinformatics.babraham.ac.uk/projects/fastqc.

Arakane, Y., Muthukrishnan, S., Kramer, K.J., Specht, C.A., Tomoyasu, Y., Lorenzen,M.D., Kanost, M., Beeman, R.W., 2005. The Tribolium chitin synthase genes TcCHS1and TcCHS2 are specialized for synthesis of epidermal cuticle and midgut peritrophicmatrix. Insect Mol. Biol. 14, 453–463.

Asselman, J., Glaholt, S.P., Smith, Z., Smagghe, G., Janssen, C.R., Colbourne, J.K., Shaw,J.R., De Schamphelaere, K.A.C., 2012. Functional characterization of four me-tallothionein genes in Daphnia pulex exposed to environmental stressors. Aquat.Toxicol. 110, 54–65.

Asselman, J., Shaw, J.R., Glaholt, S.P., Colbourne, J.K., De Schamphelaere, K.A., 2013.Transcription patterns of genes encoding four metallothionein homologs in Daphniapulex exposed to copper and cadmium are time-and homolog-dependent. Aquat.Toxicol. 142, 422–430.

Asselman, J., De Coninck, D.I., Vandegehuchte, M.B., Jansen, M., Decaestecker, E., DeMeester, L., Bussche, J.V., Vanhaecke, L., Janssen, C.R., De Schamphelaere, K.A.,2015. Global cytosine methylation in Daphnia magna depends on genotype, en-vironment, and their interaction. Environ. Toxicol. Chem. 34, 1056–1061.

Auld, S.K.J.R., 2014. In: Smirnov, N.N. (Ed.), Physiology of the Cladocera. AcademicPress, Elsevier, USA, pp. 219–233.

Badariotti, F., Thuau, R., Lelong, C., Dubos, M.P., Favrel, P., 2007. Characterization of anatypical family 18 chitinase from the oyster Crassostrea gigas: evidence for a role inearly development and immunity. Dev. Comp. Immunol. 31, 559–570.

Baird, D.J., Barata, C., 1998. Genetic variation in the response of Daphnia to toxic sub-stances: implications for risk assessment. Genetics and Ecotoxicology. Taylor andFrancis, Ann Arbor, pp. 207–221.

Barata, C., Baird, D.J., 2000. Determining the ecotoxicological mode of action of che-micals from measurements made on individuals: results from instar-based tests withdaphnia magna straus. Aquat. Toxicol. 48, 195–209.

Barata, C., Markich, S.J., Baird, D.J., Taylor, G., Soares, A.M.V.M., 2002. Genetic varia-bility in sublethal tolerance to mixtures of cadmium and zinc in clones of Daphniamagna Straus. Aquat. Toxicol. 60, 85–99.

Beckerman, A.P., de Roij, J., Dennis, S.R., Little, T.J., 2013. A shared mechanism of de-fense against predators and parasites: chitin regulation and its implications for life-history theory. Ecol. Evol. 3, 5119–5126.

Benedito Filho, P.D., Lemos, F.J., Secundino, N.F., Páscoa, V., Pereira, S.T., Pimenta, P.F.,2002. Presence of chitinase and beta-N-acetylglucosaminidase in the Aedes aegypti: achitinolytic system involving peritrophic matrix formation and degradation. InsectBiochem. Mol. Biol. 32, 1723–1729.

Bodar, C.W.M., Voogt, P.A., Zandee, D.I., 1990. Ecdysteroids in Daphnia magna: their rolein moulting and reproduction and their levels upon exposure to cadmium. Aquat.Toxicol. 17, 339–350.

Bolger, A.M., Lohse, M., Usadel, B., 2014. Trimmomatic: a flexible trimmer for illuminasequence data. Bioinformatics 30, 2114–2120.

Bossuyt, B.T., Escobar, Y.R., Janssen, C.R., 2005. Multigeneration acclimation of Daphniamagna Straus to different bioavailable copper concentrations. Ecotoxicol. Environ.Saf. 61, 327–336.

Bruning, O., Rodenburg, W., Wackers, P.F.K., van Oostrom, C., Jonker, M.J., Dekker, R.J.,Rauwerda, H., Ensink, W.A., de Vries, A., Breit, T.M., 2016. Confounding factors inthe transcriptome analysis of an in-vivo exposure experiment. PLoS One 11,e0145252.

Casalino, E., Sblano, C., Landriscina, V., Calzaretti, G., Landriscina, C., 2004. Rat liverGlutathione S-transferase activity stimulation following acute cadmium or manga-nese intoxication. Toxicology 200, 29–38.

Celis-Salgado, M.P., Cairns, A., Kim, N., Yan, N.D., 2008. The FLAMES medium: a new,soft-water culture and bioassay medium for Cladocera. Internationale Vereinigungfur Theoretische und Angewandte Limnologie Verhandlungen 30, 265–271.

Chain, F.J.J., Feulner, P.G.D., Panchal, M., Eizaguirre, C., Samonte, I.E., Kalbe, M., Lenz,T.L., Stoll, M., Bornberg-Bauer, E., Milinski, M., Reusch, T.B.H., 2014. Extensivecopy-number variation of young genes across stickleback populations. PLoS Genet.2014 (10), e1004830.

Chen, Z., Mayer, L.M., Weston, D.P., Bock, M.J., Jumars, P.A., 2002. Inhibition of

digestive enzyme activities by copper in the guts of various marine benthic in-vertebrates. Environ. Toxicol. Chem. 21, 1243–1248.

City of Greater Sudbury, 2001. City of Greater Sudbury’s 2001 Lake Water QualityInitiative. Living With Lakes Initiative.

Clemens, S., 2001. Molecular mechanisms of plant metal tolerance and homeostasis.Planta 212, 475–486.

Colbourne, J.K., Pfrender, M.E., Gilbert, D., et al., 2011. The ecoresponsive genome ofDaphnia pulex. Science 331, 555–561.

Connon, R., Hooper, H.L., Sibly, R.M., Lim, F.L., Heckmann, L.H., Moore, D.J., HajimeWatanabe, H., Soetaert, A., Cook, K., Maund, S.J., Hutchinson, T.H., Moggs, J., Coen,W.D., Iguchi, T., Callaghan, A., 2008. Linking molecular and population stress re-sponses in Daphnia magna exposed to cadmium. Environ. Sci. Technol. 42,2181–2188.

Cunha, I., Mangas-Ramirez, E., Guilhermino, L., 2007. Effects of copper and cadmium oncholinesterase and glutathione S-transferase activities of two marine gastropods(Monodonta lineata and Nucella lapillus). Comp. Biochem. Physiol. Part C Toxicol.Pharmacol. 145, 648–657.

Dameron, C.T., Harrison, M.D., 1998. Mechanisms for protection against copper toxicity.Am. J. Clin. Nutr. 67, 1091S–1097S.

Dave, G., 1984. Effects of copper on growth, reproduction, survival and haemoglobin inDaphnia magna. Comp. Biochem. Physiol. Part C Comp. Pharmacol. 78, 439–443.

De Coen, W.M., Janssen, C.R., 1998. The use of biomarkers in Daphnia magna toxicitytesting. Hydrobiologia 367, 199–209.

De Coninck, D.I., De Schamphelaere, K.A., Jansen, M., De Meester, L., Janssen, C.R.,2013. Interactive effects of a bacterial parasite and the insecticide carbaryl to life-history and physiology of two Daphnia magna clones differing in carbaryl sensitivity.Aquat. Toxicol. 130, 149–159.

De Schamphelaere, K.A.C., Forrez, I., Dierckens, K., Sorgeloos, P., Janssen, C.R., 2007.Chronic toxicity of dietary copper to Daphnia magna. Aquat. Toxicol. 81, 409–418.

DeLuca, D.S., Levin, J.Z., Sivachenko, A., Fennell, T., Nazaire, M.D., Williams, C., Reich,M., Winckler, W., Getz, G., 2012. RNA-SeQC: RNA-seq metrics for quality control andprocess optimization. Bioinformatics 28, 1530–1532.

Dinglasan, R.R., Devenport, M., Florens, L., Johnson, J.R., McHugh, C.A., Donnelly-Doman, M., Caruccic, D.J., Yates, J.R., Jacobs-Lorena, M., 2009. The Anophelesgambiae adult midgut peritrophic matrix proteome. Insect Biochem. Mol. Biol. 39,125–134.

Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P.,Chaisson, M., Gingeras, T.R., 2013. STAR: ultrafast universal RNA-seq aligner.Bioinformatics 29, 15–21.

Engel, D.W., Brouwer, M., 1987. Metal regulation and molting in the blue crab, Callinectessapidus: metallothionein function in metal metabolism. Biol. Bull. 173, 239–251.

Engel, D.W., Brouwer, M., 1991. Short-term metallothionein and copper changes in bluecrabs at ecdysis. Biol. Bull. 180, 447–452.

Fedorenkova, A., Vonk, J.A., Lenders, H.R., Ouborg, N.J., Breure, A.M., Hendriks, A.J.,2010. Ecotoxicogenomics: bridging the gap between genes and populations. Environ.Sci. Technol. 44, 4328–4333.

Franzellitti, S., Fabbri, E., 2005. Differential HSP70 gene expression in the Mediterraneanmussel exposed to various stressors. Biochem. Biophys. Res. Commun. 336,1157–1163.

Haap, T., Köhler, H.R., 2009. Cadmium tolerance in seven Daphnia magna clones is as-sociated with reduced hsp70 baseline levels and induction. Aquat. Toxicol. 94,131–137.

Haap, T., Schwarz, S., Köhler, H.R., 2016. Metallothionein and Hsp70 trade-off againstone another in Daphnia magna cross-tolerance to cadmium and heat stress. Aquat.Toxicol. 170, 112–119.

Heugens, E.H., Hendriks, A.J., Dekker, T., Straalen, N.M.V., Admiraal, W., 2001. A reviewof the effects of multiple stressors on aquatic organisms and analysis of uncertaintyfactors for use in risk assessment. Crit. Rev. Toxicol. 31, 247–284.

Hook, S.E., Osborn, H.L., Golding, L.A., Spadaro, D.A., Simpson, S.L., 2014. Dissolved andparticulate copper exposure induces differing gene expression profiles and mechan-isms of toxicity in the deposit feeding amphipod Melita plumulosa. Environ. Sci.Technol. 48, 3504–3512.

Hossain, M.A., Piyatida, P., da Silva, J.A.T., Fujita, M., 2012. Molecular mechanism ofheavy metal toxicity and tolerance in plants: central role of glutathione in detox-ification of reactive oxygen species and methylglyoxal and in heavy metal chelation.J. Bot. 2012, 1–37.

Jamers, A., Blust, R., De Coen, W., Griffin, J.L., Jones, O.A., 2013. An omics based as-sessment of cadmium toxicity in the green alga Chlamydomonas reinhardtii. Aquat.Toxicol. 126, 355–364.

Janssens, T.S., Roelofs, D., vanstraalen, N., 2009. Molecular mechanisms of heavy metaltolerance and evolution in invertebrates. Insect Sci. 16, 3.

Järup, L., 2003. Hazards of heavy metal contamination. Br. Med. Bull. 68, 167–182.Jiang, L.J., Maret, W., Vallee, B.L., 1998. The glutathione redox couple modulates zinc

transfer from metallothionein to zinc-depleted sorbitol dehydrogenase. Proc. Natl.Acad. Sci. 95, 3483–3488.

Finney, D.J., 1952. Probit Analysis. Cambridge University Press, Cambridge, England.Johnson, M.G., Owen, G.E., 1966. Report on the Biological Survey of Streams and Lakes

in the Sudbury Area, 1965. Ontario Water Resources Commission.Kawakami, M., Inagawa, R., Hosokawa, T., Saito, T., Kurasaki, M., 2008. Mechanism of

apoptosis induced by copper in PC12 cells. Food Chem. Toxicol. 46, 2157–2164.Keller, W., Heneberry, J., Gunn, J.M., Snucins, E., Morgan, G., Leduc, J., 2004. Recovery

of Acid and Metal-damaged Lakes Near Sudbury Ontario: Trends and Status.Cooperative Freshwater Ecology Unit, Sudbury.

Kim, B.M., Choi, B.S., Lee, K.W., Ki, J.S., Kim, I.C., Choi, I.Y., Rhee, J.-S., Lee, J.S., 2013.Expression profile analysis of antioxidative stress and developmental pathway genesin the manganese-exposed intertidal copepod Tigriopus japonicus with 6K oligochip.Chemosphere 92, 1214–1223.

Koedrith, P., Kim, H., Weon, J.I., Seo, Y.R., 2013. Toxicogenomic approaches for un-derstanding molecular mechanisms of heavy metal mutagenicity and carcinogenicity.Int. J. Hyg. Environ. Health 216, 587–598.


��

Kondrashov, 2012. Gene duplication as a mechanism of genomic adaptation to a changingenvironment. Proc. R. Soc. B. 279, 5048–5057.

Kramer, V.J., Etterson, M.A., Hecker, M., Murphy, C.A., Roesijadi, G., Spade, D.J.,Spromberg, J.A., Wang, M., Ankley, G.T., 2011. Adverse outcome pathways andecological risk assessment: bridging to population‐level effects. Environ. Toxicol.Chem. 30, 64–76.

Kvitek, D.J., Will, J.L., Gasch, A.P., 2008. Variations in stress sensitivity and genomicexpression in diverse S. cerevisiae isolates. PLoS Genet. 4, e1000223.

Langfelder, P., Horvath, S., 2008. WGCNA: an R package for weighted correlation net-work analysis. BMC Bioinformatics 9, 559.

Le, Q.-A.V., Sekhon, S.S., Lee, L., Ko, J.H., Min, J., 2016. Daphnia in water quality bio-monitoring - "omic" approaches. Toxicol. Environ. Health Sci. 8, 1–6.

Lee, G.R., Cartwright, G.E., Wintrobe, M.M., 1968. Heme biosynthesis in copper deficientswine. Exp. Biol. Med. 127, 977–981.

Lei, R., Ye, K., Gu, Z., Sun, X., 2015. Diminishing returns in next-generation sequencing(NGS) transcriptome data. Gene 557, 82–87.

Li, N., Zhao, Y., Yang, J., 2007. Impact of waterborne copper on the structure of gills andhepatopancreas and its impact on the content of metallothionein in juvenile giantfreshwater prawn Macrobrachium rosenbergii (Crustacea: decapoda). Arch. Environ.Contam. Toxicol. 52, 73–79.

Lionetto, M.G., Caricato, R., Erroi, E., Giordano, M.E., Schettino, T., 2005. Carbonic an-hydrase-based environmental bioassay. Int. J. Environ. Anal. Chem. 85, 895–903.

Maret, W., 1994. Oxidative metal release from metallothionein via zinc-thiol/disulfideinterchange. Proc. Natl. Acad. Sci. 91, 237–241.

McKenzie, J.A., Yen, J.L., 1995. Genotype, environment and the asymmetry phenotype.Dieldrin-resistance in Lucilia cuprina (the Australian sheep blowfly). Heredity 75,181–187.

McTaggart, S.J., Conlon, C., Colbourne, J.K., Blaxter, M.L., Little, T.J., 2009. The com-ponents of the Daphnia pulex immune system as revealed by complete genome se-quencing. BMC Genomics 10, 1.

Melegari, S.P., Perreault, F., Costa, R.H.R., Popovic, R., Matias, W.G., 2013. Evaluation oftoxicity and oxidative stress induced by copper oxide nanoparticles in the green algaChlamydomonas reinhardtii. Aquat. Toxicol. 142, 431–440.

Mussali-Galante, P., Tovar-Sánchez, E., Valverde, M., Rojas, C.E., 2013. Biomarkers ofexposure for assessing environmental metal pollution: from molecules to ecosystems.Revista Internacional de Contaminación Ambient 29, 117–140.

Muyssen, B.T.A., Janssen, C.R., 2002. Accumulation and regulation of zinc in Daphniamagna: links with homeostasis and toxicity. Arch. Environ. Contam. Toxicol. 43,0492–0496.

Muyssen, B.T., Janssen, C.R., 2004. Multi-generation cadmium acclimation and tolerancein Daphnia magna Straus. Environ. Pollut. 130, 309–316.

Muyssen, B.T.A., Janssen, C.R., 2007. Age and exposure duration as a factor influencingCu and Zn toxicity toward Daphnia magna. Ecotoxicol. Environ. Saf. 68, 436–442.

Muyssen, B.T., Messiaen, M., Janssen, C.R., 2010. Combined cadmium and temperatureacclimation in Daphnia magna: physiological and sub-cellular effects. Ecotoxicol.Environ. Saf. 73, 735–742.

Nisbet, R.M., Muller, E.B., Lika, K., Kooijman, S.A.L.M., 2000. From molecules to eco-systems through dynamic energy budget models. J. Anim. Ecol. 69, 913–926.

Nogueira, A.J.A., Baird, D.J., Soares, A.M.V.M., 2004. Testing physiologically-based re-source allocation rules in laboratory experiments with Daphnia magna Straus. AnnalesDe Limnologie-Int. J. Limnol. 40, 257–267.

Onosaka, S., Cherian, M.G., 1981. The induced synthesis of metallothionein in varioustissues of rat in response to metals. I. Effect of repeated injection of cadmium salts.Toxicology 22, 91–101.

Orsini, L., Brown, J.B., Shams Solari, O., Li, D., He, S., Podicheti, R., Stoiber, M.H.,Spanier, K.I., Gilbert, D., Jansen, M., Rusch, D.B., Pfrender, M.E., Colbourne, J.K.,Frilander, M.J., Kvist, J., Decaestecker, E., De Schamphelaere, K.A.C., De Meester, L.,2018. Early transcriptional response pathways in Daphnia magna are coordinated innetworks of crustacean‐specific genes. Mol. Ecol. 27, 886–897.

Palmiter, R.D., 1998. The elusive function of metallothioneins. Proc. Natl. Acad. Sci. 95,8428–8430.

Parry, H.E., Pipe, R.K., 2004. Interactive effects of temperature and copper on im-munocompetence and disease susceptibility in mussels (Mytilus edulis). Aquat.Toxicol. 69 (4), 311–325.

Poynton, H.C., Varshavsky, J.R., Chang, B., Cavigiolio, G., Chan, S., Holman, P.S.,Loguinov, A.V., Bauer, D.J., Komachi, K., Theil, E.C., Perkins, E.J., Hughes, O.,Perkins, E.J., 2007. Daphnia magna ecotoxicogenomics provides mechanistic insightsinto metal toxicity. Environ. Sci. Technol. 41, 1044–1050.

Poynton, H.C., Loguinov, A.V., Varshavsky, J.R., Chan, S., Perkins, E.J., Vulpe, C.D.,2008a. Gene expression profiling in Daphnia magna part I: concentration-dependentprofiles provide support for the no observed transcriptional effect level. Environ. Sci.Technol. 42, 6250–6256.

Poynton, H.C., Zuzow, R., Loguinov, A.V., Perkins, E.J., Vulpe, C.D., 2008b. Gene ex-pression profiling in Daphnia magna, part II: validation of a copper specific geneexpression signature with effluent from two copper mines in California. Environ. Sci.Technol. 42, 6257–6263.

Riddell, D.J., Culp, J.M., Baird, D.J., 2005. Behavioral responses to sublethal cadmiumexposure within an experimental aquatic food web. Environ. Toxicol. Chem. 24,431–441.

Robbens, J., van der Ven, K., Maras, M., Blust, R., De Coen, W., 2007. Ecotoxicologicalrisk assessment using DNA chips and cellular reporters. Trends Biotechnol. 25,460–466.

Robinson, M.D., McCarthy, D.J., Smyth, G.K., 2010. edgeR: a Bioconductor package for

differential expression analysis of digital gene expression data. Bioinformatics 26, 1.Rozenberg, A., Parida, M., Leese, F., Weiss, L.C., Tollrian, R., Manak, J.R., 2015.

Transcriptional profiling of predator-induced phenotypic plasticity in Daphnia pulex.Front. Zool. 12, 1.

Ruas, C.B.G., dos Santos Carvalho, C., de Araújo, H.S.S., Espíndola, E.L.G., Fernandes,M.N., 2008. Oxidative stress biomarkers of exposure in the blood of cichlid speciesfrom a metal-contaminated river. Ecotoxicol. Environ. Saf. 71, 86–93.

Santos, M., Le Bouqin, A., Crumiere, A., Khila, A., 2017. Taxon-restricted genes at theorigin of a novel trait allowing access to a new environment. Science 358, 386–390.

Schindler, D.W., 1987. Detecting ecosystem responses to anthropogenic stress. Can. J.Fish. Aquat. Sci. 44, s6–s25.

Seda, J., Petrusek, A., 2011. Daphnia as a model organism in limnology and aquaticbiology: introductory remarks. J. Limnol. 70, 337–344.

Shaw, J.R., Colbourne, J.K., Davey, J.C., Glaholt, S.P., Hampton, T.H., Chen, C.Y., Folt,C.L., Hamilton, J.W., 2007. Gene response profiles for Daphnia pulex exposed to theenvironmental stressor cadmium reveals novel crustacean metallothioneins. BMCGenomics 8, 1.

Shen, Z., Jacobs-Lorena, M., 1997. Characterization of a novel gut-specific chitinase genefrom the human malaria vector Anopheles gambiae. J. Biol. Chem. 272, 28895–28900.

Shi, L., Paskewitz, S.M., 2004. Identification and molecular characterization of two im-mune‐responsive chitinase‐like proteins from Anopheles gambiae. Insect Mol. Biol. 13,387–398.

Siva-Jothy, M.T., Moret, Y., Rolff, J., 2005. Insect immunity: an evolutionary ecologyperspective. Adv. In Insect Phys. 32, 1–48.

Soetaert, A., Vandenbrouck, T., van der Ven, K., Maras, M., van Remortel, P., Blust, R., DeCoen, W.M., 2007. Molecular responses during cadmium-induced stress in Daphniamagna: integration of differential gene expression with higher-level effects. Aquat.Toxicol. 83 (3), 212–222.

Stein, J.R., 1979. Handbook of Phycological Methods: Culture Methods and GrowthMeasurements, vol. 1 CUP Archive.

Trapnell, C., Williams, B.A., Pertea, G., Mortazavi, A., Kwan, G., Van Baren, M.J.,Salzberg, S.L., Wold, B.J., Pachter, L., 2010. Transcript assembly and quantificationby RNA-Seq reveals unannotated transcripts and isoform switching during cell dif-ferentiation. Nat. Biotechnol. 28, 511–515.

Valko, M.M.H.C.M., Morris, H., Cronin, M.T.D., 2005. Metals, toxicity and oxidativestress. Curr. Med. Chem. 12, 1161–1208.

Van Straalen, N.M., Janssens, T.K., Roelofs, D., 2011. Micro-evolution of toxicant toler-ance: from single genes to the genome’s tangled bank. Ecotoxicology 20, 574–579.

Vandegehuchte, M.B., Janssen, C.R., 2011. Epigenetics and its implications for ecotox-icology. Ecotoxicology 20, 607–624.

Vandegehuchte, M.B., Lemière, F., Janssen, C.R., 2009a. Quantitative DNA-methylationin Daphnia magna and effects of multigeneration Zn exposure. Comp. Biochem.Physiol. Part C Toxicol. Pharmacol. 150, 343–348.

Vandegehuchte, M.B., Kyndt, T., Vanholme, B., Haegeman, A., Gheysen, G., Janssen, C.R.,2009b. Occurrence of DNA methylation in Daphnia magna and influence of multi-generation Cd exposure. Environ. Int. 35, 700–706.

Vandenbrouck, T., Soetaert, A., van der Ven, K., Blust, R., De Coen, W., 2009. Nickel andbinary metal mixture responses in Daphnia magna: molecular fingerprints and (sub)organismal effects. Aquat. Toxicol. 92, 18–29.

Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics With S, 4th ed. Springer,New York.

Venâncio, C., Ribeiro, R., Soares, A., Lopes, I., 2016. Multiple stressor differential toler-ances: possible implications at the population level. PLoS One 11, e0151847.

Vilella, A.J., Severin, J., Ureta-Vidal, A., Heng, L., Durbin, R., Birney, E., 2009.EnsemblCompara GeneTrees: complete, duplication-aware phylogenetic trees invertebrates. Gen Res. 19, 327–335.

Vitale, A.M., Monserrat, J.M., Castilho, P., Rodriguez, E.M., 1999. Inhibitory effects ofcadmium on carbonic anhydrase activity and ionic regulation of the estuarine crabChasmagnathus granulata (Decapoda, Grapsidae). Comp. Biochem. Physiol. C,Pharmacol. Toxicol. Endocrinol. 122, 121–129.

von Elert, E., Agrawal, M.K., Gebauer, C., Jaensch, H., Bauer, U., Zitt, A., 2004. Proteaseactivity in gut of Daphnia magna: evidence for trypsin and chymotrypsin enzymes.Comp. Biochem. Physiol. B, Biochem. Mol. Biol. 137, 287–296.

Waters, M.D., Fostel, J.M., 2004. Toxicogenomics and systems toxicology: aims andprospects. Nat. Rev. Genet. 5, 936–948.

Weis, J.S., 1976. Effects of environmental factors on regeneration and molting in fiddlercrabs. Biol. Bull. 150, 152–162.

Wilson, V.S., Keshava, N., Hester, S., Segal, D., Chiu, W., Thompson, C.M., Euling, S.Y.,2013. Utilizing toxicogenomic data to understand chemical mechanism of action inrisk assessment. Toxicol. Appl. Pharmacol. 271, 299–308.

Yadav, S.K., 2010. Heavy metals toxicity in plants: an overview on the role of glutathioneand phytochelatins in heavy metal stress tolerance of plants. South Afr. J. Bot. 76,167–179.

Yeh, S.T., Liu, C.H., Chen, J.C., 2004. Effect of copper sulfate on the immune response andsusceptibility to Vibrio alginolyticus in the white shrimp Litopenaeus vannamei. FishShellfish Immunol. 17, 437–446.

Zhou, Q., Zhang, J., Fu, J., Shi, J., Jiang, G., 2008. Biomonitoring: an appealing tool forassessment of metal pollution in the aquatic ecosystem. Anal. Chim. Acta 606,135–150.

Zhu, B., Liu, L., Li, D.L., Ling, F., Wang, G.X., 2014. Developmental toxicity in rareminnow (Gobiocypris rarus) embryos exposed to Cu, Zn and Cd. Ecotoxicol. Environ.Saf. 104, 269–277.


��

aquatic toxicology - mcgill universitybiology.mcgill.ca/faculty/cristescu/articles/chain et al....

Documents