Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk

YTAAP-13545; No of Pages 11 Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Toxicology and Applied Pharmacology journal homepage: www.elsevier.com/locate/ytaap

Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk Pedro M. Costa, Bengt Fadeel ⁎ Nanosafety & Nanomedicine Laboratory, Division of Molecular Toxicology, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden

a r t i c l e

i n f o

Article history: Received 3 November 2015 Revised 11 December 2015 Accepted 21 December 2015 Available online xxxx Keywords: Engineered nanomaterials Proteomics Metabolomics Systems toxicology Toxicogenomics

a b s t r a c t Engineered nanomaterials are being developed for a variety of technological applications. However, the increasing use of nanomaterials in society has led to concerns about their potential adverse effects on human health and the environment. During the first decade of nanotoxicological research, the realization has emerged that effective risk assessment of the multitudes of new nanomaterials would benefit from a comprehensive understanding of their toxicological mechanisms, which is difficult to achieve with traditional, low-throughput, single end-point oriented approaches. Therefore, systems biology approaches are being progressively applied within the nano(eco)toxicological sciences. This novel paradigm implies that the study of biological systems should be integrative resulting in quantitative and predictive models of nanomaterial behaviour in a biological system. To this end, global ‘omics’ approaches with which to assess changes in genes, proteins, metabolites, etc. are deployed allowing for computational modelling of the biological effects of nanomaterials. Here, we highlight omics and systems biology studies in nanotoxicology, aiming towards the implementation of a systems nanotoxicology and mechanism-based risk assessment of nanomaterials. © 2015 Elsevier Inc. All rights reserved.

1. Introduction In the past decade, nanotoxicology has emerged as a specific domain within the toxicological sciences (Donaldson et al., 2004; Fadeel et al., 2013). In fact, we have witnessed an exponential rise in the number of papers on the subject, but as pointed out recently, nanotoxicology as a discipline is still struggling with the fundamental question: are there specific concerns associated with nanomaterials due to their particular or novel properties, that call for specific regulations to be applied in the case of nano-enabled products or technologies? (Krug, 2014). Furthermore, while considerable progress has been made, nanotoxicology still faces a number of challenges including the harmonization of nanoparticle dosimetry, the validation of in vitro assays for toxicity testing, and so on (Hussain et al., 2015). Indeed, while there are surely problems associated with many of the early papers in the field (Krug, 2014), one would be amiss to assume that all the papers published to date are of

Abbreviations: AOP, adverse outcome pathway; CRP, C-reactive protein; 2D-DIGE, two-dimensional differential in-gel electrophoresis; ENM, engineered nanomaterial; GO, gene ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; MIAME, Minimum Information About a Microarray Experiment; MIAPE, Minimum Information About a Proteomics Experiment; MS, mass spectrometry; NGS, next-generation sequencing; MWCNT, multi-walled carbon nanotube; NMR, nuclear magnetic resonance; NOTEL, no observed transcriptomic adverse effect level; OECD, Organisation for Economic Cooperation and Development; RNA-Seq, RNA sequencing. ⁎ Corresponding author at: Division of Molecular Toxicology, Institute of Environmental Medicine, Karolinska Institutet, Nobels väg 13, 171 77 Stockholm, Sweden. E-mail address: [email protected] (B. Fadeel).

poor quality or that no lessons have been learned. Researchers are now fully cognisant of the importance of a thorough physicochemical characterization of the nanomaterials (Fadeel et al., 2015), and the role of the so-called biological “identity” of nanoparticles is also recognized (Nel et al., 2009; Monopoli et al., 2012). Nanotoxicologists have also understood that “not all nanomaterials are created equal” and that even slight differences in material properties could elicit a different biological response (Hussain et al., 2015). This, in turn, further emphasizes the need for a careful characterization of nanomaterials as well as standardized and validated procedures for toxicity testing, both in vitro and in vivo, to enable the comparison of results across different studies. However, to keep up with the rapid pace of development of new classes of nanomaterials of ever increasing sophistication, it is also clear that new approaches are needed in nanotoxicology; indeed, it may be argued that this is true for (regulatory) toxicology in general (Hartung, 2009). Understanding the potential health and environmental risks associated with exposure to chemicals and nanomaterials requires accurate and predictive risk assessment approaches. As pointed out in an excellent, recent perspective, developing such approaches requires a detailed mechanistic understanding of the ways in which substances perturb biological systems and lead to adverse outcomes (Sturla et al., 2014). Systems biology approaches to human disease are grounded in the idea that diseases may perturb the normal network of a biological system through genetic perturbations and/or by pathological environmental cues (Hood et al., 2004). Systems biology has more recently been integrated with toxicology to give birth to systems toxicology, which

http://dx.doi.org/10.1016/j.taap.2015.12.014 0041-008X/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

2

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

essentially aims at a holistic understanding of the mechanisms of interaction between substances and living systems at various levels of biological organization, in order to attempt computational modelling of complex toxicological pathways to ultimately support risk assessment (Sturla et al., 2014; Fadeel, 2015). To achieve such ambitious goals, systems toxicology must rely on accurate quantitative methods that enable the screening of a wide range of responses to a toxic insult. The ability to screen for multiple changes permits a much better understanding of toxicological pathways, in comparison to the traditional single endpoint approach. To this end, so-called omics methods are being deployed; omics approaches are sometimes viewed as “high-throughput”, but it can be argued that even though vast amounts of data are generated, this is not necessarily done in a high-throughput manner, as the data analysis can be demanding. In this context, toxicogenomics is a generic term commonly referring to molecular approaches to screen for alterations in gene expression and products of protein function in living systems subjected to toxicological challenge (Chen et al., 2012). The term comprises transcriptomics, proteomics, and other more recent approaches such as metabolomics and epigenomics, which are, in essence, related to different steps along the complex chain of events of gene expression and its consequences. Needless to say, computational tools and the ability to interpret complex pathways are of paramount importance. Indeed, as pointed out recently, systems biology should not be seen merely as the generation of lists of genes, proteins, or metabolites using omics platforms; the objective is to exploit these data and to develop quantitative, predictive models that describe the biological system and its response to individual perturbations (Fadeel, 2015). Notwithstanding, the application of omics in nanotoxicology is rapidly attaining maturity. Here, an overview is provided of omics techniques and their application in nanotoxicology, focusing mainly on work published in the last five years, and how this may contribute to a systems toxicology approach to support risk assessment. Published papers were selected that best served to illustrate the use of omics approaches to guide mechanism-based toxicological studies (see Table 1). 2. Omics and bioinformatics approaches The suffix -ome as used in molecular biology refers to a totality of some sort; omics are thus used to assess globally all the genes, proteins, metabolites, etc. that are affected by a specific substance, or condition. Besides the ability to screen for multiple end-points in a single analytical run, omics techniques share the fact that they focus on changes at the molecular level. Technically, the methods applied differ according to their target, i.e., genes, transcripts, proteins, metabolites, and so on. We provide a brief description of omics and bioinformatics methods below, and how these different methods are being applied in nanotoxicology, and the reader is referred to more specialized reviews for further details. Additionally, it must be noted that systems toxicology is, by definition, a multi-level screening, which implies that the most informative research is likely that which integrates omics with more conventional end-points, to provide some measure of phenotypic anchoring of the data. Indeed, the key in systems biology (and, hence, in systems toxicology) is that phenotypic features of the system must be tied directly to the behaviour of the protein and gene regulatory networks (Ideker et al., 2001a). Moreover, systems biology, in essence, should capture global sets of biological information from as many hierarchical levels as possible (gene and protein regulatory networks, organs, individuals, populations, ecosystems) and integrate them (Ideker et al., 2001a). 2.1. Bioinformatics The ultimate goal of systems biology is to produce predictive and preferably quantitative models of biological pathways, and computational tools therefore play a pivotal role. Bioinformatics provides crucial and ever-evolving tools for the analysis and interpretation of omics

data. Specifically, bioinformatics can be deployed for the following three tasks in the toxicological sciences: (i) determination of which end-points (i.e., transcripts, proteins and others) are effectively deregulated relatively to a control or calibrator plus the quantification of such changes; (ii) association of de-regulated end-points to specific biological pathways; and (iii) development of predictive models that can be used to support risk assessment (of chemicals, or nanomaterials) based on the understanding of complex networks of molecular interactions that are affected by exposure. The first task is rooted in the need to sort de-regulated end-points through adequate statistical processes that typically involve data normalization along with analysis of variance with false-positive discovery rates and other multiple test corrections of p-values. The second task serves to assist the complex data-mining process that follows the short-listing of end-points. In recent years, several bioinformatics tools have been developed for assessment of omics data, typically linked to public access databases with emphasis on genes, and proteins. As such, the first step of this task is to obtain annotations for datasets through database searches, followed by combining cluster analysis with functional annotation, based on, for instance, Gene Ontology (GO) and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analyses. Clustering (of genes) aims for the identification of regulated biological processes through the evaluation of coregulated genes. There are several software tools developed for the purpose for bioinformatics analysis of omics data, from R-based packages such as limma, minet and wgcna (with a range of algorithms for data normalization, statistics, clustering, etc), to more user-friendly web applications like BLAST (for sequence homology searches) and DAVID (for gene enrichment and other analyses) (McGinnis and Madden, 2004; Huang et al., 2009). Finally, building predictive models of toxicological pathways remains the major challenge. To this end, analytical methods such as the Ingenuity Pathway Analysis (IPA) and Gene Set Enrichment Analysis (GSEA) software offer a starting point to unravel dynamic biological pathways and networks (Calvano et al., 2005; Subramanian et al., 2005). In a vast majority of studies, a simple univariate strategy of testing the features (genes, transcripts, proteins, etc) one by one is used. However, this results merely in a list of differentially expressed or abundant molecules and does not necessarily provide information on their potential interactions. Also, in the context of identifying molecular biomarkers (see section below), this strategy has a number of limitations and more advanced feature selection methods using multivariate analysis are preferred, based on the assumption that subsets of interacting molecules should be identified at once (see Fortino et al., 2014 for some examples of such methods, implemented in R). It must be emphasized that there is a concern regarding the accuracy and reproducibility of omics data, which has led to proposals for a Minimum Information About a Microarray Experiment (MIAME) (Brazma et al., 2001) and a Minimum Information About a Proteomics Experiment (MIAPE) (Taylor et al., 2007) by bioinformatics experts as checklists of mandatory information to warrant validation. There are also efforts to amend these standards for specific applications, such as the MIAME/Toxicogenomics, or MIAME-Tox (Burgoon, 2007). In nanotoxicology, it is especially important that the omics experiments are performed using nanomaterials that are carefully characterized and that the models used are reliable and relevant; otherwise, one may run the risk of generating enormous amounts of useless data (Fadeel, 2015). 2.2. Transcriptomics Transcriptomics aims essentially at quantifying changes in gene expression through detection of the number of mRNA copies. Unlike conventional qRT-PCR, transcriptomics technologies allow for the measurement of mRNA levels for thousands of genes simultaneously. Transcriptomics is likely the most common approach to survey both effects and mechanisms within the toxicological sciences and can be said to comprise of two distinct methods: oligonucleotide microarrays and next-generation sequencing (NGS), specifically, whole-transcriptome

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

sequencing (RNA-Seq). Microarrays remain popular in toxicological sciences, despite the advances in NGS, the latter being more expeditious (permitting the analysis of a much larger number of individual transcripts), but less cost-effective and often yielding large datasets whose interpretation may be non-trivial. In fact, most studies dealing with transcriptomic responses to nanomaterials still resort to traditional microarray technology which can be applied to a broad range of in vivo and in vitro models (from mice and zebrafish to mammalian cell lines), provided that a high level of genomic annotation is available, since the technology is based on the hybridization between labelled sample cRNAs and pre-defined oligonucleotide probes. On the other hand, RNA-Seq does not necessarily require a reference genome for reconstruction of the transcriptome, even though it is helpful. As such, RNA-Seq studies with non-model organisms, possessing high environmental relevance, may be endeavoured, aimed at supporting environmental risk assessment of nanomaterials. Indeed, RNA-Seq has been successfully applied in nano-ecotoxicology, as illustrated by the work of van Aerle et al. (2013) and Simon et al. (2013). The continuous developments in NGS will probably render RNA-Seq the successor of microarray technology in the future (Wang et al., 2009). The growing interest in transcriptomics in nanotoxicology prompted the development of the NanoMiner database in the frame of the EUfunded project, FP7-NANOMMUNE, comprising a comprehensive set of transcriptomics data based on in vitro studies of nanomaterials (Kong et al., 2013). All the samples in NanoMiner were annotated, and normalized using standard methods that ensure the quality of the data analyses and enable users to utilize the database systematically across the different experimental setups and platforms. Thus, NanoMiner serves as a repository for transcriptomics data and facilitates bioinformatics analysis of the data (Fig. 1). The resource is freely available for academic users.

2.3. Proteomics The proteome can be thought of as the direct mediator between toxicants and subsequent cellular responses to insult. Therefore, proteomics has rapidly attained popularity among toxicologists. Additionally, proteomics may provide important clues on proteinprotein interactions and post-translational modifications (van Summeren et al., 2012). Due to its high sensitivity and accuracy, mass spectrometry (MS) based methods are the keystone of proteomics, although the protocols for extraction, separation and quantification of proteins vary. Proteomics typically involve some process of first-stage protein separation, followed by quantification or quantitation of deregulated proteins relative to a control or calibrator group. The peptides of interest are then subjected to digestion and subsequent MS fingerprinting and contrasting to available databases, such as SwissProt, or UniProt. As such, proteomics may be effective even without much prior knowledge about the proteome or transcriptome of the model system in question. Two-dimensional gel electrophoresis (2D-PAGE) or two-dimensional differential in-gel electrophoresis (2D-DIGE), based on separation by isoelectric focusing and by molecular mass, are the most common protein separation techniques. In-gel methods involve relative quantification of proteins through densitometry analyses, followed by spot excision for MS sequencing. Besides HPLC for protein fractioning, more recent gel-independent approaches (regarded as more accurate for peptide quantitation and identification) such as isobaric tagging for relative and absolute quantitation (iTRAQ) are already being applied in nanotoxicology (Verano-Braga et al., 2014). Furthermore, although beyond the scope of this review, proteomics is a valuable tool to study the protein corona on nanoparticles (see Docter et al., 2014; 2015, for a technical and conceptual overview; and Vogt et al., 2015, for a recent example of bioinformatics-driven classification of bio-coronas).

3

2.4. Metabolomics Metabolomics is the comprehensive analysis of all the metabolites of an organism or specified biological sample (Robertson et al., 2011). There is some confusion in the literature with regards to the terms metabolomics and metabonomics; however, there appears to be a growing consensus that metabolomics is primarily concerned with metabolic profiling of the endogenous metabolism, whereas metabonomics encompasses perturbations of metabolism caused by environmental factors, disease processes, and coexisting organisms, such as the gut microbiome (Robertson et al., 2011). Historically, the metabonomics approach, pioneered by Nicholson, was one of the first methods to apply a systems biology approach to studies of metabolism (Nicholson et al., 2002). Nevertheless, in practice, there is still a large degree of overlap in the way both terms are used, and they are often applied interchangeably. From a toxicologist's point-of-view, metabolomics (the term that we will use hereafter) aims essentially at determining the dynamic shifts in the production of metabolites following a toxicological challenge. Unlike other omics, metabolomics does not imply a specific set of techniques; instead, the methods are diverse and dependent upon the target substances. Nevertheless, nuclear magnetic resonance (NMR) spectroscopy and MS are the most commonly used for metabolic profiling. Metabolomics can be performed on every type of biological specimen, from peripheral fluids to solid tissues, which is a clear analytical advantage. Recent advances point towards the analysis of the metabolome of single cells, although such techniques are still far from being incorporated in the toxicological arsenal (Zenobi, 2013). The integration of metabolomics with other omics, especially transcriptomics and proteomics, has been suggested as a means to obtain a comprehensive overview of the mechanisms and consequences of toxicological challenge (Sturla et al., 2014). However, the literature dealing with the application of metabolomics in nanotoxicology is relatively sparse (Lv et al., 2015). The EU-funded project, FP7-MARINA (http://www.marina-fp7.eu), is applying a multi-omics approach based on a combination of transcriptomics, proteomics and metabolomics to study nanotoxicity. 2.5. Genomics and epigenomics Genomics is a field that includes efforts to determine the entire DNA sequence of organisms. In addition, although not a widespread approach in the toxicological sciences, whole-genome sequencing has been used in biomedicine to screen for mutations in DNA, usually in relation to neoplastic disease. In spite of the importance of determining genotoxicity and mutagenicity of chemicals and nanoparticles, only scant literature can be found on the applications of whole-genome methods in toxicology. However, recent work on model mutagens such as UV light and benzo[a]pyrene indicates the relevance of such approaches in toxicology (Nik-Zainal et al., 2015). Moreover, exome sequencing (i.e., sequencing of all the coding regions) to screen for DNA alterations following exposure to mutagens has been reported (Severson et al., 2014). There is growing evidence that epigenetic modifications such as DNA methylation and histone modifications may be caused by environmental factors (Jirtle and Skinner, 2007). Such alterations, which by definition do not imply changes in the DNA sequence, are now believed to influence complex disease-related pathways, including tumorigenesis, even though the related mechanisms are not yet fully understood. In spite of the growing concern regarding epigenetic alterations in response to xenobiotics, high-throughput epigenomics approaches are still in their infancy in the field of nanotoxicology (Shyamasundar et al., 2015). Different microarray and next-generation sequencing protocols are available for whole-genome epigenomic screening aiming at either DNA and chromatin modifications, or profiling of microRNAs. In a recent study, Eom et al. (2014) combined mRNA and microRNA microarrays and identified epigenetic mechanisms in response to silver (Ag)

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

4

“Omics” approach

Assay NM(s)

Transcriptomics In vivo

In vitro

MWCNT

Model

Organ/tissue

Exposure

Dose range

Method(s)

Affected pathways

References

Rat

Lung

6 h (aerosol) +

11 mg/m3

cDNA microarray

Immune-related, cell differentiation and proliferation, ion transport

Ellinger-Ziegelbauer and Pauluhn (2009)

40 μg/mL fullerenes and 170 ng/mL TiO2 Up to 50 μg/L

cDNA microarray

Circadian rhythm, immune response, basal metabolism

Jovanović et al. (2011)

cDNA microarray

DNA repair, gene transcription, development Stress response, regeneration

Griffitt et al. (2013)

90-day post-exposure period Whole-organism Microinjection (48 h incubation)

TiO2 (anatase) and hydroxylated fullerenes

Danio rerio (embryo)

Ag

Danio rerio (adult)

Gills

SiO2

Hydra vulgaris

Whole-organism Water (24 h)

25 nM

RNAseq (Illumina)

TiO2 (nano and bulk)

Caenorhabditis elegans

Whole-organism Water (24 h)

Up to 10 μg/mL

cDNA microarray

TiO2 (free and sanding dust-bound)

Mouse

Lung

18–162

cDNA microarray

Basal metabolic pathways, including oxidative-related, and development-related Immune/inflammation-related genes

SiO2, TiO2, ZnO and Fe2O3 Ag, TiO2, ZnO and Cd/Te quantum dots TiO2 and ZnO

RKO and CaCo-2 (human) Chlamydomonas reinhardtii

Colon 4h Whole-organism 2 h

cDNA microarray RNAseq (SOLiD)

Protein folding Photosynthesis, cell structure

Moos et al. (2011) Simon et al. (2013)

Monocyte-derived macrophages and dendritic cells; Jurkat (human) EA.hy926 (human)

Immune system

cDNA microarray

Tuomela et al. (2013)

Endothelium

No changes recorded for TiO2. ZnO disrupted multiple pathways, from cell death to immune-related Inflammation, apoptosis, cell cycle and basal metabolism

Fröhlich et al. (2014)

HepG2 (human)

Liver

TGFβ1-mediated (GO) or NF-kB-mediated responses (rGO)

Chatterjee et al. (2014)

CaCo-2 (human)

Colon

Cell adhesion, oxidative stress

Böhmert et al. (2015)

Inflammation, cell signalling

Snyder-Talkington et al. (2015)

No overt cytotoxicity; cell cycle arrest; and senescence gene signatures

Feliu et al. (2015)

Polystyrene (PS), plain and carboxylated; SWNCTs, MWCNTs Graphene oxide (GO) and reduced graphene oxide (rGO) Ag

MWCNT

PAMAMs

Water (24–48 h)

Single instillation (intratracheal)

6 and 24 h

μg/instillation 5–50 μg/cm2 1 μg/mL NPs, 0.125 μg/mL QDs 1 μg/ml and 10 μg/m

Small airway epithelial cells and Lung, microvascular endothelial cells endothelium (human) Primary broncho-epithelial Lung cells (human)

24 h

cDNA microarray Up to 200 μg/mL (PS) and 50 (CNP) μg/mL 20 mg/L (GO) and cDNA microarray 8 mg/L (rGO)

24 h

2.5 and 25 μg/mL

6 or 24 h

1.2 μg/mL

cDNA microarray and custom RT-PCR array cDNA microarray

48 h

0.1 μM

RNAseq (Illumina)

Ambrosone et al. (2014) Rocheleau et al. (2015) Halappanavar et al. (2015)

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

Table 1 Representative transcriptomics, proteomics, and metabolomics studies in nanotoxicology research.

In vivo

In vitro

Metabolomics

In vivo

In vitro

Multi-omics

In vivo In vitro

SWCNT, asbestos, ultrafine carbon

Mouse

Lung

Aerosol (3 × week

40 μg/mouse

HPLC-FTICR-MS

× 3 weeks)

Mostly related to immune response; proteins deregulated by nanotubes similar to those affected by asbestos Inflammation, gene expression, fatty acid metabolism and proteasome function Immune/inflammation

Teeguarden et al. (2011) Gao et al. (2011)

TiO2 (82% anatase, 16% rutile) Mouse

Lymph nodes

Single intradermal injection

25 mg/kg

2DLC-MS/MS

MWCNT and Al2O3-coated MWCNT

Mouse

Bronchoalveolar fluid

4 mg/kg

TiO2 (80% anatase + 20% rutile) Ag

Rat

Blood

30 μg/animal

In-solution digestion LC–MS/MS LC–MS/MS

LoVo (human)

Colon

Pharyngeal aspiration (single exposure) Aerosol (single exposure, 4–6 h) 24 h

10 μg/mL

iTRAQ

Au, functionalised with anti-sense cDNAs

HCT-116 (human)

Colon

48 h

Equivalent to 30

2DE and MALDI-TOF MS

No significantly perturbed pathways

Amino polystyrene nanospheres (with and without Pd conjugation) TiO2

HEK-293T (human) and L929 (mouse)

Kidney; connective tissue Lung

48 h

nM of cDNAs 86 μg/mL

Glycolysis, cytoskeleton motility

Pietrovito et al. (2015)

2 months

1–50 μg/mL

2DE and MALDI-TOF/TOF MS 2DE and NanoLC–MS/MS 2D-DIGE and LC–MS/MS

DNA damage response

Armand et al. (2015) Palomäki et al. (2015)

A549 (human)

MWCNT and asbestos

Monocyte-derived macrophages (human)

Immune system

6h

100 μg/mL

MWCNT

U937 (human)

Immune system

96 h

1 μg/mL

TiO2 (nano and bulk)

Whole-organism Water (24 h)

Fullerene

Caenorhabditis elegans (L4 larvae) Eisenia fetida

TiO2

L929 (mouse)

TiO2 and TiO2 functionalised with IL-1β SWCNT (carboxylated and graphene oxide) Ag (citrate-stabilized) MWCNT (pristine and hydroxylated)

Gingival fibroblasts (human)

2DE and MALDI-TOF/MS 7.7 and 38.5 μg/ml GC–MS H1 NMR

Acute phase response signalling, LXR/RXR, and FXR/RXR activation Protein kinase signalling cascade

Focus on secreted and not intracellular proteins; inflammation and apoptosis Basal metabolism and cellular stress Amino acid metabolism, TCA cycle

Hilton et al. (2015)

Maurer et al. (2016) Verano-Braga et al. (2014) Conde et al. (2014)

Haniu et al. (2010) Ratnasekhar et al. (2015) Lakandurai et al. (2015) Bo et al. (2014)

Whole-organism Soil and direct skin contact (2–7 days) Connective 24 h tissue Connective 30 min + 24 h tissue (with IL-1β) Whole-organism 96 h

Up to 3000 mg/kg soil 25–300 μg/mL

GC-TOF/MS

Altered peptide synthesis and energy metabolism Amino acid metabolism

0.2 to 3.2 mM

CapE-ESI-MS

Amino acid and peptide metabolism

0.01–10 mg/L

GC–MS

Oxidative stress

Skin 48 h Whole-organism Water (up to 24 h)

10 and 40 μg/mL 1 mg/L

Immune system

48 h

5 μg/mL

Magnetic, SiO2-coated

HEK293 (human)

Kidney

12 h

Up to 1.0 μg/μL

24 h

2

NMR cDNA microarray and 2DE and LC–MS/MS cDNA microarray and 2DE-ESI-TOF-MS cDNA microarray and GC–MS cDNA microarray and LC–MS/MS

Energy production, oxidative stress Carrola et al. (2015) Uptake processes and oxidative stress Eom et al. (2015)

Au

HaCaT (human) Caenorhabditis elegans (wild-type and mutant (impaired phagocytosis) K562 (human)

Chlorella vulgaris

SiO2

A549 (human)

Lung

Cerium

Chlamydomonas reinhardtii (strain CCAP 11/32c)

Whole-organism 72 h

0.1 to 6 μg/cm

0.029–10.000 μg/L FTICR MS and cDNA microarray

Garcia-Contreras et al. (2015) Hu et al. (2015)

Multiple pathways indicative of intracellular stress

Tsai et al. (2011)

Mitochondrial function/energy production pathways Focus on secreted and not intracellular proteins; oxidative stress-related pathways, inflammation, xenobiotic metabolism Photosynthesis-related pathways

Shim et al. (2012)

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

Proteomics

Pisani et al. (2015)

Taylor et al. (2015)

Abbreviations: 2D-PAGE, 2-dimensional polyacrylamide gel electrophoresis; 2D-DIGE, 2-dimensional differential in-gel electrophoresis; 2DLC, 2-dimensional liquid chromatography; CapE-ESI-MS, capillary electrophoresis electrospray ionization mass spectrometry; ESI-MS, electrospray ionization mass spectrometry; FTICR, Fourier transform ion cyclotron resonance; GC–MS, gas chromatography mass spectrometry; iTRAQ, isobaric tagging for relative and absolute quantitation; LC–MS/MS; MALDITOF-MS, matrix-assisted desorption/ionization time-of-flight mass spectrometry; NanoLC–MS/MS, nanoscale liquid chromatography coupled with tandem mass spectrometry; NM, nanomaterial; NMR, nuclear magnetic resonance; NP, nanoparticle; RTPCR, real-time polymerase chain reaction.

5

6

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Fig. 1. NanoMiner database: a repository for transcriptomics data. For details, refer to Kong et al. (2013). Reprinted from Kong et al. (2013) under a Creative Commons Attribution (CC-CY) licence.

nanoparticles; the authors could also show that Ag nanoparticles and Ag ions triggered a distinct set of epigenetic changes, using the Jurkat cell line as a model. However, potential trans-generational effects of nanoparticle-induced epigenetic alterations remain to be explored. 3. Omics applications in nanotoxicology 3.1. Human and environmental safety Nanosafety encompasses both human health risk assessment and environmental or “ecological” risk assessment of engineered nanomaterials. However, there is often a (real or perceived) barrier between the two scientific communities, possibly related to customs and methodological biases in each respective field of research. Nevertheless, the human toxicology and eco-toxicology research communities need to communicate, to reduce duplication of efforts and optimize resources (Malysheva et al., 2015). This is true not least in the emerging field of systems toxicology, which makes use of costly and laborious experimental platforms. It is also important to note that in systems biology, information on several hierarchical levels including molecular, cellular, organ, organism, and population levels is incorporated to model the interactions of a toxicant and a biological system (Sturla et al., 2014); eco-toxicologists are more attuned to the concept of the ecological impact of toxicants and human toxicologists who frequently deal with effects in a single individual (a single cell type, a single strain of inbred mice) may benefit from such a “population-based” view. In order to understand the effects of chemicals including nanomaterials in humans, animal testing is still mandatory. The high level of genomic annotation renders the laboratory mouse a preferred model in nanotoxicological studies involving high-throughput omics techniques (see, for instance, Hilton et al., 2015; Halappanavar et al., 2015). However, eco-toxicology may offer alternative, non-mammalian model systems for screening of nanomaterial effects. Hence, laboratory strains of non-mammalian species, e.g., the zebrafish, Danio rerio, or the nematode, Caenorhabditis elegans, which may be of less environmental relevance, hold great advantage in terms of genomic annotation, available transgenic/mutant lines, and reduced intraspecific variability. Such models are now being widely applied in toxicological research, with emerging examples of omics-based analysis of nanomaterial effects (see, for instance, Griffitt et al., 2013; Rocheleau et al., 2015) (and see Table 1). However, it is pertinent to note that contrasting species from different taxonomical groups poses an important challenge, due to the

evolution of the biological pathways in question. For instance, the immune system of vertebrates is more complex and more specialized than that of invertebrates, and certain metabolic pathways may not be universally relevant for different species. In other words, whereas pathways that are conserved among taxa may be readily compared, it is important to consider the issue of translatability of information when embarking on multi-species comparisons of data. Additionally, less orthodox biological models are also deployed for mechanism-oriented research. This is the case, for instance, in a recent RNA-Seq study of silica nanoparticles using the freshwater invertebrate, Hydra vulgaris (Ambrosone et al., 2014). Similarly, Simon et al. (2013) applied RNASeq to investigate the molecular effects of metal nanoparticles and quantum dots in the unicellular green alga, Chlamydomonas reinhardtii. Taylor et al. (2015) used a combination of microarray-based transcriptomics and MS-based metabolomics to study effects of cerium oxide nanoparticles in C. reinhardtii. Growth inhibition was monitored using standard OECD test guidelines, to better anchor the omics data. Notably, molecular perturbations were detected only at unrealistically high (supra-environmental) concentrations. KEGG pathway enrichment analysis showed that multiple molecular pathways were perturbed at high concentrations, the majority of which were represented by both transcripts and metabolites, indicating a consistency between gene expression and subsequent metabolic change (Taylor et al., 2015). On the other hand, in vitro assays have many advantages comparatively to animal testing. They are compliant with the 3-R (Refinement, Reduction and Replacement) ethical guidelines for laboratory animal science, and are more expeditious than in vivo models, thus allowing for (automated) high-throughput toxicity testing of nanomaterials (Feliu and Fadeel, 2010; Arora et al., 2012). Additionally, recent advances in co-cultures of different cell types in an attempt to mimic the in vivo physiology, combined with microfluidic approaches to better control the tissue microenvironment, suggest new models for toxicity testing of drugs and chemicals, and could help to reduce animal testing (Bhatia and Ingber, 2014; Braakhuis et al., 2015). Few studies have been published to date in which omics approaches to assess nanomaterials have been applied in co-culture model systems. However, in a recent instalment, Snyder-Talkington et al. (2015) compared gene expression profiles in human lung epithelial and microvascular endothelial cells exposed to multi-walled CNTs (MWCNTs) in monoculture and co-culture with gene expression from mouse lungs exposed to MWCNTs. When analysis of microarray data was confined to only those genes involved in inflammation and fibrosis – known outcomes of in vivo exposure to

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

MWCNTs – there were more concordant genes expressed in co-cultures than in monocultures. The authors concluded that co-cultures of relevant cell types can provide an improved system for high-throughput in vitro testing that better mimics the in vivo situation, and may reduce the need for animal testing of nanomaterials. For a further discussion on alternative test strategies for nanomaterials, see Nel (2013). 3.2. Mechanism-oriented research One of the problems in contemporary nanotoxicological research is that we too often are looking for the keys under the lamppost (Feliu and Fadeel, 2010). In contrast, with global omics approaches, novel and/or unanticipated end-points may be identified, which could also yield novel biomarkers (see below); whether or not there are nanospecific effects remains a contentious subject and the reader is referred to other recent reviews (Donaldson and Poland, 2013; Gallud and Fadeel, 2015). There is, on the other hand, a common assumption that the application of omics technologies is paramount to hypothesis-free research. However, while omics methods provide an opportunity for unbiased assessment of the perturbations of genes, proteins, metabolites, etc. in response to a certain stimulus, this does not mean that the study itself should be devoid of a hypothesis. Moreover, in many studies in which high-throughput omics methods have been applied, researchers tend to view their data in light of the prevailing paradigms (such as, the oxidative stress paradigm; see Shvedova et al., 2012), and it may be argued, therefore, that it is the scientist, and not the method, that is biased. Nonetheless, there are also recent examples of omics studies that, when coupled with appropriate bioinformatics evaluation, have indicated novel and/or low-dose effects, not captured by conventional cellular assays, as well as indirect effects which are manifested only in response to challenge of normal cellular function. Using genome-wide microarrays, Kodali et al. (2013) determined that preincubation of primary murine bone marrow-derived macrophages with iron oxide nanoparticles that did not elicit acute cytotoxicity caused extensive transcriptional “reprogramming” in response to bacterial LPS. Quantitative RT-PCR analysis confirmed these microarray results. Furthermore, macrophages exposed to nanoparticles displayed an altered phenotype suggesting an impaired ability to undergo activation, and diminished phagocytic activity towards the lung pathogen, Streptococcus pneumonia. The authors concluded that the biological effects of engineered nanomaterials may be indirectly manifested after challenging normal cell function (eg., pathogen recognition) (Kodali et al., 2013). In a recent study using primary human bronchial epithelial cells as a model, we found that cationic poly(amidoamine) dendrimers (PAMAMNH2) triggered significant changes in gene expression, using doses at which these nanoparticles did not trigger cell death according to conventional cell viability assays (Feliu et al., 2015). Importantly, global gene expression profiling using RNA-Seq coupled with detailed bioinformatics assessment revealed that all of the most significantly differentially expressed gene categories were related to cell cycle and cell division. In other words, we were able to identify the predominant, transcriptional effects of the nanoparticles and subsequent biological experiments validated these findings. In addition, using pathway analysis software, we identified NF-KB as a potential upstream regulator of gene expression, and this in silico-based prediction was confirmed in functional assays (Feliu et al., 2015). Furthermore, we performed Connectivity Map analysis (Lamb et al., 2006) to identify putative similarities between PAMAMNH2 and other compounds with known modes of action. The analysis showed that the gene expression profile of PAMAM-NH2 matched the profiles of several other compounds known to cause S-phase arrest, thus further corroborating our model (Feliu et al., 2015). In a related study, Lucafò et al. (2013) evaluated the effects of fullerenes in the human MCF-7 breast carcinoma cell line using RNA-Seq followed by Connectivity Map analysis. The authors could show that the gene expression signature of fullerene-treated cells was strikingly similar to those

7

induced by selective inhibitors of mammalian target of rapamycin (mTOR) signalling, thus suggesting a molecular mechanism. As noted above, a systems biology approach implies the assessment of multiple hierarchical levels of biology, meaning that multi-omics analyses are required, along with appropriate bioinformatics tools to manage the data. Wilmes et al. (2013) applied integrated transcriptomics, proteomics and metabolomics to investigate the molecular processes of the immunosuppressive drug, cyclosporine A (CsA), which is also known to be nephrotoxic, in cultured human renal epithelial cells. They could demonstrate activation of the nuclear factor erythroid 2-related factor 2 (Nrf2) pathway related to oxidative stress and the unfolded protein response (UPR) pathway at high concentrations, but not at lower concentrations which were closer to the clinical peak plasma concentration of CsA (Wilmes et al., 2013). Thus, using this integrated, multi-omics approach, the authors could achieve a global view of compound-induced cellular stress and could distinguish pharmacological from toxicological effects. The EU-funded project FP7-NANOSOLUTIONS (http://nanosolutionsfp7. com) aims to provide a means to develop a safety classification for engineered nanomaterials based on an understanding of their interactions with living organisms at the molecular, cellular and organism level, using conventional toxicological end-points and omics data (i.e., transcriptomics, proteomics, and epigenomics data). In Table 1, a few examples of multi-omics studies are provided; typically, a combination of proteomics and transcriptomics was applied. 3.3. Biomarkers and risk assessment One of the key paradigms in biomonitoring and risk assessment of hazardous substances and nanomaterials is the biomarker concept. In toxicology, biomarkers are usually defined as any sub-individual endpoint whose change can indicate an interaction between a biological system and a given toxicological stressor (Bergamaschi and Magrini, 2012). Even though the classification of biomarkers is highly reliant on the way in which they are interpreted, they are typically divided into three functional categories: biomarkers of exposure, effect and susceptibility. Biomarkers of exposure are defined as exogenous or endogenous substances whose presence can indicate an interaction between a xenobiotic and a living system. In turn, biomarkers of effect are quantifiable measures of end-points that can be associated to a potential adverse effect resulting from exposure to a xenobiotic. Finally, biomarkers of susceptibility relate to the ability (innate or acquired) of an organism to respond to a given stressor. High-throughput (omics) methods, if applied in a rigorous manner, hold great promise for the development of (novel) biomarkers and biomarker signatures (Jennings, 2015). Azuaje et al. (2009) provided a description of a biomarker discovery framework in which omics data is incorporated. As noted above, multivariate analysis, able to take into consideration the interactions existing among the potential biomarkers, is important (Robotti et al., 2014). Omics have mainly been used in effects-oriented nanosafety research for the purpose of hazard identification, but have not been widely applied in biomonitoring, or identification of biomarkers. In a recent transcriptomics study performed using primary human immune cells, the induction of metallothioneins was suggested as a potential biomarker of exposure to ZnO nanoparticles (Tuomela et al., 2013). However, in vivo validation of these findings is needed. Higashisaka et al. (2011) employed a SDS-PAGE-based proteomics approach to identify candidate biomarkers of exposure and toxicity of silica nanoparticles following intravenous administration in mice, and proposed that acute phase proteins may act as biomarkers for assessing the risk of exposure to silica nanoparticles. In a subsequent study, the authors performed a similar screen for biomarkers using two-dimensional differential in-gel electrophoresis (2D-DIGE), a gel-based approach like SDS-PAGE (i.e., involving separation on the basis of molecular weight and isoelectric point) but applying differential tagging of peptides, and identified hemopexin (another acute phase protein) as a potential biomarker for predicting the effects of silica nanoparticles (Higashisaka et al., 2012).

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

8

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Acute phase proteins, such as C-reactive protein (CRP), are commonly used as biomarkers of (inflammatory) disease, and associations have also been shown between inflammatory biomarkers including CRP and exposure to particulate matter (PM) air pollution (see Bergamaschi and Magrini, 2012, for a review). In a recent study, Pisani et al. (2015) used a microarray-based approach combined with secretomics (proteomics of the secreted proteome of exposed cells) to assess cellular responses to fumed silica nanoparticles in the human lung carcinoma cell line, A549. The authors derived a so-called “no observed transcriptomic adverse effect level” (NOTEL) that was lower (i.e., more sensitive) than values obtained using conventional cell viability tests, and argued that transcriptomics could be applied to benchmark potentially any type of toxicant, alone or in a mixture, in a predictive risk assessment framework. Related to this work, Palomäki et al. (2015) recently performed proteomics analyses of primary human macrophages exposed to tangled or rigid, long MWCNTs, or crocidolite asbestos, using 2D-DIGE and LC–MS/MS based protocols, and could conclude that not all types of CNTs are as hazardous as asbestos fibres; specifically, rigid MWCNTs caused severe toxicity (in this model system), while this was not the case for tangled MWCNTs. These findings have implications for the risk assessment of this class of nanomaterials and illustrate the utility of unbiased omics approaches. Finally, while we have discussed hazard assessment of nanomaterials, exposure data are also needed. Wild (2005) proposed the term exposome one decade ago, to draw attention to the need for better and more complete environmental exposure data in epidemiological research; the exposome, which is dynamic and variable, thus consists of all of the internal and external exposures an individual incurs over a lifetime. There are considerable challenges in developing the exposome concept into a workable approach for epidemiological research (Vrijheid, 2014). Nevertheless, the exposome is a paradigm, or perhaps a vision, that puts a spotlight on the totality of environmental (i.e., non-genetic) exposures and serves to complement the genome (Wild, 2005). For the purpose of the present discussion, it is worth noting that omics technologies could provide tools to measure biological responses to exposures, as reflected by profiles or signatures of transcripts, proteins, or metabolites. However, a

comprehensive discussion of the exposome is beyond the scope of the present review. 4. Future outlook: towards systems nanotoxicology What may one expect from systems toxicology? Sturla et al. (2014) concluded, in their perspective, that “an overarching expectation is that systems toxicology approaches will provide more sound information on which to judge how chemicals cause biological perturbations, moving knowledge beyond knowing only what phenotypes are altered”. In other words, our knowledge of the underlying mechanisms and the interactions between substances and biomolecules – or networks and pathways of biomolecules – will inform risk assessment. In this manner, systems toxicology “will enable the gradual shift from toxicological assessment using solely apical end-points towards understanding the biological pathways perturbed by active substances” (Sturla et al., 2014). This is strongly linked to, and aligned with, the concept of adverse outcome pathways (AOPs) (Ankley et al., 2010), a concept first espoused in ecotoxicology and defined as “existing knowledge concerning the linkage between a direct molecular initiating event and an adverse outcome at a biological level of organization relevant to risk assessment” (Ankley et al., 2010). For a depiction of a systems toxicology framework enabling mechanism-based risk assessment of nanomaterials, see Fig. 2. Has a systems biology view of nanomaterial toxicity been achieved? The short answer is: not yet. First, it must be reiterated that applying omics methods in an in vitro or in vivo setting does not automatically render it into a systems biology study. Omics methods are tools which, when coupled with bioinformatics (computational) approaches, can be used to identify pathways that can be quantitatively modelled. Moreover, and this is something that we have not discussed in great detail in the present review, a systems biology strategy consists of several steps of which the generation of data using omics techniques is only the first step. In their landmark paper, Ideker et al. (2001b) described four distinct steps: (i) define all of the genes, proteins, and other small molecules constituting the pathway of interest, and, if possible, define an initial model of the molecular interactions governing pathway

Fig. 2. Systems toxicology framework: role of omics. The ultimate outputs of omics-enabled modelling of nanomaterial-induced perturbations are the so-called adverse outcome pathways or AOPs.

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

function; (ii) perturb each pathway component through a series of genetic of environmental manipulations, and detect the corresponding changes in global cellular responses; (iii) integrate the observed gene and protein responses with the pathway-specific model; and (iv) formulate new hypotheses to explain observations not predicted by the model, and design new experiments to test these. In other words, systems biology calls for an iterative approach of perturbations and global measurements; in this manner, systems biology can be seen as an interplay between discovery- and hypothesis-driven science (Ideker et al., 2001a). Notably, high-throughput approaches can be utilized not only for molecular profiling, but also for molecular perturbation (Yao et al., 2015). Functional studies are needed to validate the models. Moreover, as we have previously pointed out (Fadeel, 2015), care should be taken when performing high-throughput omics analyses of nanomaterials; if the experiment is poorly designed or controlled, and/or if the nanomaterials that are under study are poorly characterized in terms of their physicochemical properties, then the resulting omics data will, inevitably, be of poor quality too. Indeed, a detailed understanding of the physicochemical properties of nanomaterials, and how these properties change when nanomaterials come into contact with a biological system, through biotransformation, or through the acquisition, or shedding, of a bio-corona, is required for an adequate interpretation of the data (Fadeel et al., 2015). In sum, we have witnessed a surge of omics studies in nanotoxicology in recent years, but the field is still adolescent and examples of true systems biology approaches are, admittedly, scarce. However, there is a tremendous potential and we expect that the field of systems toxicology will continue to evolve, thus enabling improved risk assessment of nanomaterials. We also believe cross-fertilization between environmental and human toxicology is important, and suggest that the two communities may benefit greatly from joint efforts in this field. Conflict of interest statement The authors declare that there are no conflicts of interest. Transparency document The Transparency document associated with this article can be found, in online version. Acknowledgements This work is supported by the European Commission (FP7-MARINA, Grant No. 263215; FP7-SUN, Grant No. 604305; FP7-NANOSOLUTIONS, Grant No. 309329), and the Swedish Foundation for Strategic Environmental Research (MISTRA Environmental Nanosafety). Members of the working group on Systems Biology within the EU Nanosafety Cluster are acknowledged for helpful discussions; however, the authors are solely responsible for the conclusions and opinions in this report. References Ambrosone, A., di Vettimo, M.R.S., Malvindi, M.A., Roopin, M., Levy, O., Marchesano, V., Pompa, P.P., Tortiglione, C., Tino, A., 2014. Impact of amorphous SiO2 nanoparticles on a living organism: morphological, behavioral, and molecular biology implications. Front. Bioeng. Biotechnol. 2, 37. Ankley, G.T., Bennett, R.S., Erickson, R.J., Hoff, D.J., Hornung, M.W., Johnson, R.D., Mount, D.R., Nichols, J.W., Russom, C.L., Schmieder, P.K., Serrrano, J.A., Tietge, J.E., Villeneuve, D.L., 2010. Adverse outcome pathways: a conceptual framework to support ecotoxicology research and risk assessment. Environ. Toxicol. Chem. 29, 730–741. Armand, L., Biola-Clier, M., Bobyk, L., Collin-Faure, V., Diemer, H., Strub, J.-M., Cianferani, S., van Dorsselaer, A., Herlin-Boime, N., Rabilloud, T., Carriere, M., 2015. Molecular responses of alveolar epithelial A549 cells to chronic exposure to titanium dioxide nanoparticles: a proteomic view. J. Proteome http://dx.doi.org/10.1016/j.jprot.2015. 08.006 (Aug. 11, Epub ahead of print). Arora, S., Rajwade, M., Paknikar, K.M., 2012. Nanotoxicology and in vitro studies: the need of the hour. Toxicol. Appl. Pharmacol. 258, 151–165.

9

Azuaje, F., Devaux, Y., Wagner, D., 2009. Computational biology for cardiovascular biomarker discovery. Brief. Bioinform. 10, 367–377. Bergamaschi, E., Magrini, A., 2012. Biomonitoring. In: Fadeel, B., Pietroiusti, A., Shvedova, A. (Eds.), Adverse Effects of Engineered Nanomaterials. Exposure, Toxicology, and Impact on Human Health. Academic Press, Elsevier, Amsterdam, pp. 45–62. Bhatia, S.N., Ingber, D.E., 2014. Microfluidic organs-on-chips. Nat. Biotechnol. 32, 760–772. Bo, Y., Jin, C., Liu, Y., Yu, W., Kang, H., 2014. Metabolomic analysis on the toxicological effects of TiO₂ nanoparticles in mouse fibroblast cells: from the perspective of perturbations in amino acid metabolism. Toxicol. Mech. Methods 24, 461–469. Böhmert, L., Niemann, B., Lichtenstein, D., Juling, S., Lampen, A., 2015. Molecular mechanism of silver nanoparticles in human intestinal cells. Nanotoxicology 9, 852–860. Braakhuis, H.M., Kloet, S.K., Kezic, S., Kuper, F., Park, M.V.D.Z., Bellmann, S., van der Zande, M., Le Gac, S., Krystek, P., Peters, R.J.B., Rietjens, M.C.M., Bouwmeester, H., 2015. Progress and future of in vitro models to study translocation of nanoparticles. Arch. Toxicol. 89, 1469–1495. Brazma, A., Hingamp, P., Quackenbush, J., Sherlock, G., Spellman, P., Stoeckert, C., Aach, J., Ansorge, W., Ball, C.A., Causton, H.C., Gaasterland, T., Glenisson, P., Holstege, F.C., Kim, I.F., Markowitz, V., Matese, J.C., Parkinson, H., Robinson, A., Sarkans, U., SchulzeKremer, S., Stewart, J., Taylor, R., Vilo, J., Vingron, M., 2001. Minimum information about a microarray experiment (MIAME)-toward standards for microarray data. Nat. Genet. 29, 365–371. Burgoon, L.D., 2007. Clearing the standards landscape: the semantics of terminology and their impact on toxicogenomics. Toxicol. Sci. 99, 403–412. Calvano, S.E., Xiao, W., Richards, D.R., Felciano, R.M., Baker, H.V., Cho, R.J., Chen, R.O., Brownstein, B.H., Cobb, J.P., Tschoeke, S.K., Miller-Graziano, C., Moldawer, L.L., Mindrinos, M.N., Davis, R.W., Tompkins, R.G., Lowry, S.F., 2005. A network-based analysis of systemic inflammation in humans. Nature 437, 1032–1037. Carrola, J., Bastos, V., Ferreira de Oliveira, J.M., Oliveira, H., Santos, C., Gil, A.M., Duarte, I.F., 2015. Insights into the impact of silver nanoparticles on human keratinocytes metabolism through NMR metabolomics. Arch. Biochem. Biophys. http://dx.doi.org/10. 1016/j.abb.2015.08.022 (Sep. 3, Epub ahead of print). Chatterjee, N., Eom, H.-J., Choi, J., 2014. A systems toxicology approach to the surface functionality control of graphene-cell interactions. Biomaterials 35, 1109–1127. Chen, M., Zhang, M., Borlak, J., Tong, W., 2012. A decade of toxicogenomic research and its contribution to toxicological science. Toxicol. Sci. 130, 217–228. Conde, J., Larguinho, M., Cordeiro, A., Raposo, L.R., Costa, P.M., Santos, S., Diniz, M.S., Fernandes, A.R., Baptista, P.V., 2014. Gold-nanobeacons for gene therapy: evaluation of genotoxicity, cell toxicity and proteome profiling analysis. Nanotoxicology 8, 521–532. Docter, D., Distler, U., Storck, W., Kuharev, J., Wünsch, D., Hahlbrock, A., Knauer, S.K., Tenzer, S., Stauber, R.H., 2014. Quantitative profiling of the protein coronas that form around nanoparticles. Nat. Protoc. 9, 2030–2044. Docter, D., Westmeier, D., Markiewicz, M., Stolte, S., Knauer, S.K., Stauber, R.H., 2015. The nanoparticle biomolecule corona: lessons learned — challenge accepted? Chem. Soc. Rev. 44, 6094–6121. Donaldson, K., Poland, C.A., 2013. Nanotoxicity: challenging the myth of nano-specific toxicity. Curr. Opin. Biotechnol. 24, 724–734. Donaldson, K., Stone, V., Tran, C.L., Kreyling, W., Borm, P.J., 2004. Nanotoxicology. Occup. Environ. Med. 61, 727–728. Ellinger-Ziegelbauer, H., Pauluhn, J., 2009. Pulmonary toxicity of multi-walled carbon nanotubes (Baytubes®) relative to alpha-quartz following a single 6 h inhalation exposure of rats and a 3 months post-exposure period. Toxicology 266, 16–29. Eom, H.J., Chatterjee, N., Lee, J., Choi, J., 2014. Integrated mRNA and micro RNA profiling reveals epigenetic mechanism of differential sensitivity of Jurkat T cells to AgNPs and Ag ions. Toxicol. Lett. 229, 311–318. Eom, H.-J., Roca, C.P., Roh, J.-Y., Chatterjee, N., Jeong, J.-S., Shim, I., Kim, H.-M., Kim, P.-J., Choi, K., Giralt, F., Choi, J., 2015. A systems toxicology approach on the mechanism of uptake and toxicity of MWCNT in Caenorhabditis elegans. Chem. Biol. Interact. 239, 153–163. Fadeel, B., 2015. Systems biology in nanosafety research. Nanomedicine (London) 10, 1039–1041. Fadeel, B., Alenius, H., Savolainen, K., 2013. Nanotoxicology. Toxicology 313, 1–2. Fadeel, B., Fornara, A., Toprak, M.S., Bhattacharya, K., 2015. Keeping it real: the importance of material characterization in nanotoxicology. Biochem. Biophys. Res. Commun. http://dx.doi.org/10.1016/j.bbrc.2015.06.178 (Jul 15, Epub ahead of print). Feliu, N., Fadeel, B., 2010. Nanotoxicology: no small matter. Nanoscale 2, 2514–2520. Feliu, N., Kohonen, P., Ji, J., Zhang, Y., Karlssson, H.L., Palmberg, L., Nyström, A., Fadeel, B., 2015. Next-generation sequencing reveals low-dose effects of cationic dendrimers in primary human bronchial epithelial cells. ACS Nano 9, 146–163. Fortino, V., Kinaret, P., Fyhrquist, N., Alenius, H., Greco, D., 2014. A robust and accurate method for feature selection and prioritization from multi-class OMICs data. PLoS One 9, e107801. Fröhlich, E., Meindl, C., Wagner, G., Roblegg, E., 2014. Use of whole genome expression analysis in the toxicity screening of nanoparticles. Toxicol. Appl. Pharmacol. 280, 272–284. Gallud, A., Fadeel, B., 2015. Keeping it small: towards a molecular definition of nanotoxicology. Eur. J. Nanomed. 7, 143–151. Gao, Y., Gopee, N.V., Howard, P.C., Yu, L.-R., 2011. Proteomic analysis of early response lymph node proteins in mice treated with titanium dioxide nanoparticles. J. Proteome 74, 2745–2759. Garcia-Contreras, R., Sugimoto, M., Umemura, N., Kaneko, M., Hatakeyama, Y., Soga, T., Tomita, M., Scougall-Vilchis, R.J., Contreras-Bulnes, R., Nakajima, H., Sakagami, H., 2015. Alteration of metabolomic profiles by titanium dioxide nanoparticles in human gingivitis model. Biomaterials 57, 33–40. Griffitt, R.J., Lavelle, C.M., Kane, A.S., Denslow, N.D., Barber, D.S., 2013. Chronic nanoparticulate silver exposure results in tissue accumulation and transcriptomic changes in zebrafish. Aquat. Toxicol. 130-131, 192–200.

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

10

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Halappanavar, S., Saber, A.T., Decan, N., Jensen, K.A., Wu, D., Jacobsen, N.R., Guo, C., Rogowski, J., Koponen, I.K., Levin, M., Madsen, A.M., Atluri, R., Snitka, V., Birkedal, R.K., Rickerby, D., Williams, A., Wallin, H., Yauk, C.L., Vogel, U., 2015. Transcriptional profiling identifies physicochemical properties of nanomaterials that are determinants of the in vivo pulmonary response. Environ. Mol. Mutagen. 56, 245–264. Haniu, H., Matsuda, Y., Takeuchi, K., Kim, Y.A., Hayashi, T., Endo, M., 2010. Proteomicsbased safety evaluation of multi-walled carbon nanotubes. Toxicol. Appl. Pharmacol. 242, 256–262. Hartung, T., 2009. Toxicology for the twenty-first century. Nature 460, 208–212. Higashisaka, K., Yoshioka, Y., Yamashita, K., Morishita, Y., Fujimura, M., Nabeshi, H., Nagano, K., Abe, Y., Kamada, H., Tsunoda, S.-I., Yoshikawa, T., Itoh, N., Tsutsumi, Y., 2011. Acute phase proteins as biomarkers for predicting the exposure and toxicity of nanomaterials. Biomaterials 32, 3–9. Higashisaka, K., Yoshioka, Y., Yamashita, K., Morishita, Y., Pan, H., Ogura, T., Nagano, T., Kunieda, A., Nagano, K., Abe, Y., Kamada, H., Tsunoda, S., Nabeshi, H., Yoshikawa, T., Tsutsumi, Y., 2012. Hemopexin as biomarkers for analyzing the biological responses associated with exposure to silica nanoparticles. Nanoscale Res. Lett. 7, 555. Hilton, G.M., Taylor, A.J., McClure, C.D., Parsons, G.N., Bonner, J.C., Bereman, M.S., 2015. Toxicoproteomic analysis of pulmonary carbon nanotube exposure using LC–MS/ MS. Toxicology 329, 80–87. Hood, L., Heath, J.R., Phelps, M.E., Lin, B., 2004. Systems biology and new technologies enable predictive and preventative medicine. Science 306, 640–643. Hu, X., Ouyang, S., Mu, L., An, J., Zhou, Q., 2015. Effects of graphene oxide and oxidized carbon nanotubes on the cellular division, microstructure, uptake, oxidative stress, and metabolic profiles. Environ. Sci. Technol. 49, 10825–10833. Huang, D.W., Sherman, B.T., Lempicki, R.A., 2009. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57. Hussain, S.M., Warheit, D.B., Ng, S.P., Comfort, K.K., Grabinski, C.M., Braydich-Stolle, L.K., 2015. At the crossroads of nanotoxicology in vitro: past achievements and current challenges. Toxicol. Sci. 147, 5–16. Ideker, T., Galitski, T., Hood, L., 2001a. A new approach to decoding life: systems biology. Annu. Rev. Genomics Hum. Genet. 2, 343–372. Ideker, T., Thorsson, V., Ranish, J.A., Christmas, R., Buhler, J., Eng, J.K., Bumgarner, R., Goodlett, D.R., Aebersold, R., Hood, L., 2001b. Integrated genomic and proteomic analyses of a systematically perturbed metabolic network. Science 292, 929–934. Jennings, P., 2015. The future of in vitro toxicology. Toxicol. in Vitro 29, 1217–1221. Jirtle, R.L., Skinner, M.K., 2007. Environmental epigenomics and disease susceptibility. Nat. Rev. Genet. 8, 253–262. Jovanović, B., Ji, T., Palić, D., 2011. Gene expression of zebrafish embryos exposed to titanium dioxide nanoparticles and hydroxylated fullerenes. Ecotoxicol. Environ. Saf. 74, 1518–1525. Kodali, V., Littke, M.H., Tilton, S.C., Teeguarden, J.G., Shi, L., Frevert, C.W., Wang, W., Pounds, J.G., Thrall, B.D., 2013. Dysregulation of macrophage activation profiles by engineered nanoparticles. ACS Nano 7, 6997–7010. Kong, L., Tuomela, S., Hahne, L., Ahlfors, H., Yli-Harja, O., Fadeel, B., Lahesmaa, R., Autio, R., 2013. NanoMiner — integrative human transcriptomics data resource for nanoparticle research. PLoS One 8, e68414. Krug, H.F., 2014. Nanosafety research—are we on the right track? Angew. Chem. Int. Ed. Engl. 53, 12304–12319. Lakandurai, B.P., Nagato, E.G., Simpson, A.R., Simpson, M.J., 2015. Analysis of Eisenia fetida earthworm responses to sub-lethal C60 nanoparticle exposure using 1 H-NMR based metabolomics. Ecotoxicol. Environ. Saf. 120, 48–58. Lamb, J., Crawford, E.D., Peck, D., Modell, J.W., Blat, I.C., Wrobel, M.J., Lerner, J., Brunet, J.P., Subramanian, A., Ross, K.N., Reich, M., Hieronymus, H., Wei, G., Armstrong, S.A., Haggarty, S.J., Clemons, P.A., Wei, R., Carr, S.A., Lander, E.S., Golub, T.R., 2006. The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease. Science 313, 1929–1935. Lucafò, M., Gerdol, M., Pallavicini, A., Pacor, S., Zorzet, S., Da Ros, T., Prato, M., Sava, G., 2013. Profiling the molecular mechanism of fullerene cytotoxicity on tumor cells by RNA-seq. Toxicology 314, 183–192. Lv, M., Huang, W., Chen, Z., Jiang, H., Chen, J., Tian, Y., Zhang, Z., Xu, F., 2015. Metabolomics techniques for nanotoxicity investigations. Bioanalysis 7, 1527–1544. Malysheva, A., Lombi, E., Voelcker, N.H., 2015. Bridging the divide between human and environmental nanotoxicology. Nat. Nanotechnol. 10, 835–844. Maurer, M.M., Donohoe, G.C., Maleki, H., Yi, J., McBride, C., Nurkiewicz, T.R., Valentine, S.J., 2016. Comparative plasma proteomic studies of pulmonary TiO2 nanoparticle exposure in rats using liquid chromatography tandem mass spectrometry. J. Proteome 130, 85–93. McGinnis, S., Madden, T.L., 2004. BLAST: at the core of a powerful and diverse set of sequence analysis tools. Nucleic Acids Res. 32 (S2), W20–W25. Monopoli, M.P., Åberg, C., Salvati, A., Dawson, K.A., 2012. Biomolecular coronas provide the biological identity of nanosized materials. Nat. Nanotechnol. 7, 779–786. Moos, P.J., Olszewski, K., Honeggar, M., Cassidy, P., Leachman, S., Woessner, D., Cutler, N.S., Verantha, J.M., 2011. Responses of human cells to ZnO nanoparticles: a gene transcription study. Metallomics 3, 1199–1211. Nel, A.E., 2013. Implementation of alternative test strategies for the safety assessment of engineered nanomaterials. J. Intern. Med. 274, 561–577. Nel, A.E., Mädler, L., Velegol, D., Xia, T., Hoek, E.M., Somasundaran, P., Klaessig, F., Castranova, V., Thompson, M., 2009. Understanding biophysicochemical interactions at the nano-bio interface. Nat. Mater. 8, 543–557. Nicholson, J.K., Connelly, J., Lindon, J.C., Holmes, E., 2002. Metabonomics: a platform for studying drug toxicity and gene function. Nat. Rev. Drug Discov. 1, 153–161. Nik-Zainal, S., Kucak, J.E., Morganella, S., Glodzik, D., Alexandrov, L.N., Artl, V.M., Weninger, A., Hollstein, M., Stratton, M.E., Phillips, D.H., 2015. The genome as a record of environmental exposure. Mutagenesis 30, 763–770.

Palomäki, J., Sund, J., Vippola, M., Kinaret, P., Greco, D., Savolainen, K., Puustinen, A., Alenius, H., 2015. A secretomics analysis reveals major differences in the macrophage responses towards different types of carbon nanotubes. Nanotoxicology 9, 719–728. Pietrovito, L., Cano-Cortés, V., Gamberi, T., Magherini, F., Bianchi, L., Bini, L., SánchezMartín, Fasano, M., Modesti, A., 2015. Cellular response to empty and palladiumconjugated amino-polystyrene nanospheres uptake: a proteomic study. Proteomics 15, 34–42. Pisani, C., Gaillard, J.C., Nouvel, V., Odorico, M., Armengaud, J., Prat, O., 2015. Highthroughput, quantitative assessment of the effects of low-dose silica nanoparticles on lung cells: grasping complex toxicity with a great depth of field. BMC Genomics 16, 315. Ratnasekhar, C., Sonane, M., Satish, A., Mudiam, M.K.R., 2015. Metabolomics reveals the perturbations in the metabolome of Caenorhabditis elegans exposed to titanium dioxide nanoparticles. Nanotoxicology 9, 994–1004. Robertson, D.G., Watkins, P.B., Reily, M.D., 2011. Metabolomics in toxicology: preclinical and clinical applications. Toxicol. Sci. 120 (S1), S146–S170. Robotti, E., Manfredi, M., Marengo, E., 2014. Biomarkers discovery through multivariate statistical methods: a review of recently developed methods and applications in proteomics. J. Proteomics Bioinform. S3, 003. Rocheleau, S., Arbour, M., Elias, M., Sunahara, G., Masson, L., 2015. Toxicogenomic effects of nano- and bulk-TiO2 particles in the soil nematode Caenorhabditis elegans. Nanotoxicology 8, 502–512. Severson, P.L., Vrba, L., Stampfer, M.R., Futscher, B.W., 2014. Exome-wide mutation profile in benzo[a]pyrene-derived post-stasis and immortal human mammary epithelial cells. Mutat. Res. 775-776, 48–54. Shim, W., Paik, M.J., Nguyen, D.T., Lee, J.K., Lee, Y., Kim, J.H., Shin, E.H., Kang, J.S., Jung, H.S., Choi, S., Park, S., Shim, J.S., Lee, G., 2012. Analysis of changes in gene expression and metabolic profiles induced by silica-coated magnetic nanoparticles. ACS Nano 6, 7665–7680. Shvedova, A.A., Pietroiusti, A., Fadeel, B., Kagan, V., 2012. Mechanisms of carbon nanotubeinduced toxicity: focus on oxidative stress. Toxicol. Appl. Pharmacol. 261, 121–133. Shyamasundar, S., Ng, C.T., Yung, L.Y., Dheen, S.T., Bay, B.H., 2015. Epigenetic mechanisms in nanomaterial-induced toxicity. Epigenomics 7, 395–411. Simon, D.F., Domingos, R.F., Hauser, C., Hutchins, C.M., Zerges, W., Wilkinson, K.J., 2013. Transcriptome sequencing (RNA-seq) analysis of the effects of metal nanoparticle exposure on the transcriptome of Chlamydomonas reinhardtii. Appl. Environ. Microbiol. 79, 4774–4785. Snyder-Talkington, B.N., Dong, C., Zhao, X., Dymacek, J., Porter, D.W., Wolfarth, M.G., Castranova, V., Qian, Y., Guo, N.L., 2015. Multi-walled carbon nanotube-induced gene expression in vitro: concordance with in vivo studies. Toxicology 328, 66–74. Sturla, S.J., Boobis, A.R., Fitzgerald, R.E., Hoeng, J., Kavlock, R.J., Schirmer, K., Whelan, M., Wilks, M.F., Peitsch, M.C., 2014. Systems toxicology: from basic research to risk assessment. Chem. Res. Toxicol. 27, 314–329. Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gillette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., Mesirov, J.P., 2005. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U. S. A. 102, 15545–15550. Taylor, C.F., Paton, N.W., Lilley, K.S., Binz, P.A., Julian, R.K., Jones, A.R., Zhu, W., Apweiler, R., Aebersold, R., Deutsch, E.W., Dunn, M.J., Heck, A.J., Leitner, A., Macht, M., Mann, M., Martens, L., Neubert, T.A., Patterson, S.D., Ping, P., Seymour, S.L., Souda, P., Tsugita, A., Vandekerckhove, J., Vondriska, T.M., Whitelegge, J.P., Wilkins, M.R., Xenarios, I., Yates, J.R., Hermjakob, H., 2007. The minimum information about a proteomics experiment (MIAPE). Nat. Biotechnol. 25, 887–893. Taylor, N.A., Merrifield, R., Williams, T.D., Chipman, J.K., Lead, J.R., Viant, M.R., 2015. Molecular toxicity of cerium oxide nanoparticles to the freshwater alga Chlamydomonas reinhardtii is associated with supra-environmental exposure concentrations. Nanotoxicology http://dx.doi.org/10.3109/17435390.2014.1002868 (Sep. 4, Epub ahead of print). Teeguarden, J.G., Webb-Robertson, B.-J., Waters, K., Murray, A.R., Kisin, E.R., Varnum, S.M., Jacobs, J.M., Pounds, J.G., Zanger, R.C., Shvedova, A.A., 2011. Comparative proteomics and pulmonary toxicity of instilled single-walled carbon nanotubes, crocidolite asbestos, and ultrafine carbon black in mice. Toxicol. Sci. 120, 123–135. Tsai, Y.Y., Huang, Y.H., Chao, Y.L., Hu, K.Y., Chin, L.T., Chou, S.H., Hour, A.L., Yao, Y.D., Tu, C.S., Liang, Y.J., Tsai, C.Y., Wu, H.Y., Tan, S.W., Chen, H.M., 2011. Identification of the nanogold particle-induced endoplasmic reticulum stress by omic techniques and systems biology analysis. ACS Nano 5, 9354–9369. Tuomela, S., Aution, R., Buerki-Thurherr, Arslan, O., Kunzmann, A., Andersson-Willman, B., Wick, P., Mathur, S., Scheynius, A., Krug, H., Fadeel, B., Lahesmaa, 2013. Gene expression profiling of immune-competent human cells exposed to engineered zinc oxide or titanium dioxide nanoparticles. PLoS One 8, e68415. van Aerle, R., Lange, A., Moorhouse, A., Paszkiewicz, K., Ball, K., Johnston, B.D., de Bastos, E., Booth, T., Tyler, C.R., Santos, E.M., 2013. Molecular mechanisms of toxicity of silver nanoparticles in zebrafish embryos. Environ. Sci. Technol. 47, 8005–8014. van Summeren, A., Renes, J., van Delft, J.H., Kleinjans, J.C., Mariman, E.C., 2012. Proteomics in the search for mechanisms and biomarkers of drug-induced hepatotoxicity. Toxicol. in Vitro 26, 373–385. Verano-Braga, T., Methling-Graff, R., Wojdyla, K., Rogowska-Wresinska, A., Brewer, J.R., Erdmann, H., Kjeldsen, F., 2014. Insights into the cellular response triggered by silver nanoparticles using quantitative proteomics. ACS Nano 8, 2161–2175. Vogt, C., Pernemalm, M., Kohonen, P., Laurent, S., Hultenby, K., Vahter, M., Lehtiö, J., Toprak, M.S., Fadeel, B., 2015. Proteomics analysis reveals distinct corona composition on magnetic nanoparticles with different surface coatings: implications for interactions with primary human macrophages. PLoS One 10, e0129008. Vrijheid, M., 2014. The exposome: a new paradigm to study the impact of environment on health. Thorax 69, 876–878.

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014

P.M. Costa, B. Fadeel / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx Wang, Z., Gerstein, M., Snyder, M., 2009. RNA-Seq: a revolutionary tool for transcriptomics. Nat. Rev. Genet. 10, 57–63. Wild, C.P., 2005. Complementing the genome with an “exposome”: the outstanding challenge of environmental exposure measurement in molecular epidemiology. Cancer Epidemiol. Biomark. Prev. 14, 1847–1850. Wilmes, A., Limonciel, A., Aschauer, L., Moenks, K., Bielow, C., Leonard, M.O., Hamon, J., Carpi, D., Ruzek, S., Handler, A., Schmal, O., Herrgen, K., Bellwon, P., Burek, C., Truisi, G.L., Hewitt, P., Di Consiglio, E., Testai, E., Blaauboer, B.J., Guillou, C., Huber, C.G.,

11

Lukas, A., Pfaller, W., Mueller, S.O., Bois, F.Y., Dekant, W., Jennings, P., 2013. Application of integrated transcriptomic, proteomic and metabolomic profiling for the delineation of mechanisms of drug induced cell stress. J. Proteome 21 (79), 180–194. Yao, Z., Petschnigg, J., Ketteler, R., Stagljar, I., 2015. Application guide for omics approaches to cell signaling. Nat. Chem. Biol. 11, 387–397. Zenobi, R., 2013. Single-cell metabolomics: analytical and biological perspectives. Science 342, 1243259.

Please cite this article as: Costa, P.M., Fadeel, B., Emerging systems biology approaches in nanotoxicology: Towards a mechanism-based understanding of nanomaterial hazard and risk, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.12.014