Published: April 16, 2016
Nanotechnology encompasses a wide variety of nanosized materials, or nanomaterials (NMs), which are defined by the International Organization for Standardization (ISO) as materials with any external dimension in the nanoscale or having internal structure or surface structure in the nanoscale (ISO, 2011). There is a large number of NMs in the global marketplace with the estimated nanotechnology market valued at almost $20 billion in 2015 (Rapaka et al., 2014). Unlike their bulk counterparts, the diversity of NMs is enhanced by differences in physico-chemical properties that then impact both biological effects and fate and transport in the environment. NMs can vary extensively from each other in their physico-chemical properties. For example, NMs can be metallic, metal oxides, carbon-based, or polymer-based, and not only is each basic type of material available in different shapes, sizes, and surface areas, but also basic materials are often further enhanced via surface modification or combined with other types of materials to generate composites.The differences in NM properties may be intentional, to fit the application requirement, or unintentional, due to unanticipated size-related effects. Furthermore, NMs can undergo transformations – such as aggregation/ agglomeration, oxidation, changes in surface properties, and ionization – that are specific to the immediate environment that they are exposed to and the precise nature of the potential transformations over the life cycle (which includes conditions of use and disposal) for the NM product.
The sheer number and diversity of NMs has been a driving force encouraging researchers to develop in vitro and in silico (“non-animal”) methods to assess NM safety. Non-animal tests have the potential to be more high-throughput and less expensive than testing on animals. In addition to providing an animal welfare advantage, non-animal methods allow for the study of NM effects in a controlled environment that enables linking of NM properties to the observed human-relevant biological effect. Thus, economic, logistic, ethical and scientific concerns necessitate the development of methods for characterization and safety testing of NMs that are high-throughput, relevant to realistic exposures and use patterns, and have high predictability for receptor-based mechanistic endpoints/outcomes. Discussed here are several topics of consideration amongst the nanocommunity to streamline the safety testing of NMs. These topics include: 1) NM definition and regulatory approaches; 2) NM characterization; 3) grouping and read-across; and 4) development of high-quality in vitro data.
It’s all in the definition
Generally speaking, NMs are materials with at least one dimension measuring between 1-100 nm. As simple as this statement sounds, there has been considerable debate among various jurisdictions on how NMs should be precisely defined within legislature for the purposes of regulation. As a result, various versions of nanotechnology-related definitions exist and have been included in Table 1.
The European Commission (EC) published a recommendation in 2011 on the definition of a NM (2011/696/EU) (EU, 2011). This definition applies to Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) and CLP (Classification, Labelling and Packaging). REACH and CLP are European Union (EU) chemical safety and labeling regulations which contain no explicit considerations of NMs. However, the newer Biocidal Products Regulation (BPR) has specific provisions for NMs that meet the EU definition and calls for separate dossiers for nanoforms and bulk forms of active ingredients (http://echa.europa.eu/regulations/nanomaterials-under-bpr). In addition, discussions continue on how to revise the Novel Food Regulation (EC 258/97) to best address the definition of NMs. Amenta et al. provide a review of the regulation of nanotechnology in the European agriculture and food & feed sectors (Amenta et al., 2015).
Health Canada’s (HC’s) definition of a NM is general, allowing collection of information from NM characterization and measurement (Table 1, (HC, 2011)). The definition forms the basis of the consultation document proposed by HC and Environment Canada in early 2015 on a “Proposed Approach to Address Nanoscale Forms of Substances on the Domestic Substances List” (EC, 2015). The proposal is a stepwise approach to acquire and evaluate information to address NMs considered in commerce in Canada (on Canada’s public inventory) (HC, 2011). In follow-up to the consultation, a stakeholder workshop is planned for 2016 to discuss the future activities based on the comments received.
“A natural, incidental or manufactured material containing particles, in an unbound state or as an aggregate or as an agglomerate and where, for 50 % or more of the particles in the number size distribution, one or more external dimensions is in the size range 1 nm – 100 nm. In specific cases and where warranted by concerns for the environment, health, safety or competitiveness the number size distribution threshold of 50% may be replaced by a threshold between 1 and 50 %. Fullerenes, graphene flakes and single wall carbon nanotubes with one or more external dimensions below 1 nm should also be considered as nanomaterials.”
European Commission (EU, 2011)
Any manufactured substance or product and any component material, ingredient, device, or structure is a NM if: “it is at or within the nanoscale in at least one external dimension, or has internal or surface structure at the nanoscale, or; it is smaller or larger than the nanoscale in all dimensions and exhibits one or more nanoscale properties/phenomena.”
Health Canada (HC, 2011)
“Many nanoscale materials are regarded as chemical substances under the Toxic Substances Control Act (TSCA). Specifically, chemical substances that have structures with dimensions at the nanoscale – approximately 1-100 nanometers (nm) – are commonly referred to as nanoscale materials or nanoscale substances.”
U.S. EPA (EPA, 2015)
“Nanotechnology is the understanding and control of matter at dimensions between approximately 1 and 100 nanometers, where unique phenomena enable novel applications. Encompassing nanoscale science, engineering, and technology, nanotechnology involves imaging, measuring, modeling, and manipulating matter at this length scale.”
U.S. National Nanotechnology Initiative (NNI, 2011)
NMs are “both materials that have at least one dimension in the size range of approximately 1 nanometers (nm) to 100 nm and certain materials that otherwise exhibit related dimension dependent properties or phenomena.” FDA alsoconcludes that it “finds it reasonable to consider evaluation of materials or end products engineered to exhibit properties or phenomena attributable to dimensions up to 1,000 nm, as a means to screen materials for further examination and to determine whether these materials exhibit properties or phenomena attributable to their dimension(s) and associated with the application of nanotechnology.”
U.S. Food and Drug Administration (FDA, 2014).
“Industrial materials intentionally produced, manufactured or engineered to have unique properties or specific composition at the nanoscale, that is a size range typically between 1 nm and 100 nm, and is either a nano-object (i.e. that is confined in one, two, or three dimensions at the nanoscale) or is nanostructured (i.e. having an internal or surface structure at the nanoscale).”
Australia National Industrial Chemicals Notification and Assessment Scheme (NICNAS, 2010)
“A material with any external dimensions in the nanoscale or having internal structure or surface structure in the nanoscale.”
The International Organization for Standardization (ISO, 2011)
In the United States (U.S.), the more than 20 federal agencies that comprise the National Nanotechnology Initiative (NNI, established in 2000) generally follow the NNI’s definition. However, given the wide variability in agency mandates and responsibility, different federal agencies have released differing descriptions of what it considers a NM while determining the future regulatory path for a specific product. These descriptions can differ within federal and some state agencies. For example, both the Food and Drug Administration (FDA) Office of Pharmaceutical Science and the California Department of Toxic Substance Control have stated previously that NMs as large as 1,000 nm in one dimension might be considered under its specific regulations (Janus et al., 2010). In addition to the definition put forth by the NNI, the FDA has issued guidance for industry regarding the types of properties (size included) that would classify a material as a NM (FDA, 2014).
It has been argued that establishing a strict definition of a NM may impede regulation, and one such proponent has argued for establishing “trigger points” (empirically-derived metrics or decision points that signal when the risk is too high) in lieu of categorical definitions (Maynard, 2011). Definitions based solely on mass or volume do not seem to sufficiently address NMs (such as the tonnage restrictions in REACH) and a one-size-fits-all approach has been called scientifically indefensible (Maynard, 2011; Boverhof et al., 2015). Terms such as “engineered” and “novel properties” are subjective and can be associated with large, potentially futile, data calls for the regulator to understand how a material behaves in a number of different conditions and what impact that might have on human health and the environment (Dana, 2013). Given that many government entities consider materials up to 1 micrometer (1,000 nanometers) in diameter (e.g., U.S. FDA, U.S. Department of Agriculture National Organic Standards Board, UK House of Lords, and California Safer Consumer Products Act), a hybrid approach and an approach that defines NMs solely by whether an observed effect/ behavior is associated with dimension or size, regardless of actual size, may be optimal (Maynard, 2011; Dana, 2013).
Boverhof et al. provide a detailed comparison of the differing NM definitions, and conclude that frameworks for risk evaluation are the greater need (within which situationally and context-specific definitions can be included in order to achieve a case-by-case approach to NMs) (Boverhof et al., 2015). All definitions of NMs require measurement, but NM morphology is dependent on its interaction with the external environment and subsequent life cycle or use pattern changes; therefore, the definition might apply to one life cycle stage of NM but not another (e.g., to the “as manufactured” form, but not the form present in the external matrix). Furthermore, many of the analytical techniques that are available have limitations in characterizing NMs in complex matrices.
The need to characterize NMs was realized as soon as the capability of synthesizing custom NMs was acquired; however, the early nanotoxicology literature shows a variety of characterization approaches in addition to differing levels of completeness. In the mid-2000s, as the literature on NMs and nanoparticles (NPs; a nano-object with all three external dimensions in the nanoscale (ISO, 2011)) was starting to balloon, an interdisciplinary workshop was held to discuss what data would be needed to better understand the human health effects of NMs and how to best obtain these data (Balbus et al., 2007).
A long-term, hypothesis-based research program aimed at elucidating mechanistic information and identifying characteristics that determine nanotoxicity was deemed an essential component. In order to understand the mechanisms of nanotoxicity, NM characterization is of paramount importance and includes particle size, particle distribution, shape, surface area (including bioavailable surface area), redox potential and properties, purity and contaminant profile, catalytic activity, and potential for generation of reactive oxygen species. Multiple measurements of these properties, taken using standardized protocols, were encouraged in various environmental matrices with consideration given to potential for transformation, route of entry, and target-organ specific biologically effective doses. Balbus et al. also pointed out several critical research needs, including adequate methods for NP characterization in biological matrices, standardized non-animal methods to develop mechanistic toxicology data, development of toxicokinetic and toxicodynamic data (“ADME” studies), and “sufficient experience and data to be able to predict the toxicity of nanoparticles on the basis of their physiciochemical characteristics and mechanistic assays.” (Balbus et al., 2007). Finally, it was suggested that standardized reporting requirements for physico-chemical and nanotoxicological data be established at scientific journals and public research agencies. This final suggestion for standardized reporting has been echoed by others to avoid the publication of seemingly discordant results on the same material from different laboratories and to provide a standardized platform from which the “sufficient experience and data” can be generated (Oberdorster et al., 2005; Powers et al., 2006; Boverhof & David, 2010).
Participants at an additional interdisciplinary meeting in 2008, referred to as the Minimum Information for Nanomaterial Characterization (MinChar) Initiative, developed the list of recommended minimum physical and chemical parameters (or measurands) for the characterization of NMs (Table 2) for nanotoxicology studies (Boverhof & David, 2010). Table 2 reflects the question-based approach taken by the MinChar initiative (MinChar, 2008) which is a minimal list of measurands and is not intended to replace more detailed and robust guidelines, as offered by the Organisation for Economic Co-operation and Development (OECD) or groups such as the International Organization for Standardization (ISO). ISO launched a technical committee (TC229) in 2005 whose stated responsibilities are “developing standards for: terminology and nomenclature; metrology and instrumentation, including specifications for reference materials; test methodologies; modeling and simulations; and science-based health, safety, and environmental practices.”
Table 2. Recommended approach to minimum characterization developed at MinChar Initiative in 2008 – the first list of “recommended minimum physical and chemical parameters” (or measurands) for characterization of NMs for nanotoxicology studies [adapted from Boverhof & David, 2010]
Physical and chemical properties
What does the material look like?
What is the material made of?
What factors affect the material’s interaction with surroundings?
What are the other overarching considerations?
In addition to characterizing NMs in their manufactured form, it is important to evaluate transformations in various matrices as well as to evaluate transformations that can occur as a result of sample preparation during experiments and measurement. Clift et al. pointed out that researchers must “fully understand not only the cellular system but also the NP suspensions that they are using” for in vitro studies to understand the potential for transformation and to facilitate the collection of the appropriate characterization data (Clift et al., 2011).
A number of techniques are available that can precisely characterize NMs in their pristine form, but challenges exist in the application of such techniques for characterization in complex matrices (Goenaga-Infante & Larsen, 2014). To address this issue, the British Standards Institute has developed guidelines (PAS 139, 2012) on the detection and characterization of manufactured nano-objects in complex matrices (BSI, 2012). Furthermore, challenges associated with NM characterization in relevant matrices can be overcome by using multiple techniques to evaluate similar properties (Methner et al., 2010). For example, the Nanoparticle Emission Assessment Technique (NEAT) developed by the U.S. National Institute for Occupational Safety and Health (NIOSH) employs a combination of handheld direct-read instruments (particle counters to detect the release of airborne NMs) coupled with air sampling and subsequent chemical and microscopic analyses for NM characterization and elemental analysis (Methner et al., 2010). Baer and colleagues provide a list of tools and techniques that can be used to generate NP characterization data, as found in the ISO 14187 report on NM surface chemical analysis (Baer et al., 2013).
Grouping and Read-across
To address the wide variety of materials awaiting testing, there is much interest in grouping NMs into categories—based on parameters such as physico-chemical properties or observed effects—to facilitate the use of read-across approaches. Grouping and read-across help maximize the use of existing data for predicting the effects of unknown substances. OECD defines a chemical category as “a group of chemicals whose physico-chemical and human health and/or ecotoxicological properties and/or environmental fate properties are likely to be similar or follow a regular pattern, usually as a result of structural similarity” (OECD, 2014). Chemical categories can be used in a read-across approach, in which an endpoint (i.e., physico-chemical properties, toxicity, environmental fate, or ecotoxicity) for a source chemical is used to predict the same endpoint for a similar chemical (ECHA, 2013; OECD, 2014).
NMs can thus be grouped based on physico-chemical properties, biological activity, predicted exposures, or anticipated use in products. Often, groups are assembled based on structural similarity, as defined via [Quantitative] Structure-Activity Relationships, or QSARs, which are in silico models used to predict outcome based on structure (ECHA, 2008). For example, using 10 putatively active and 10 putatively inactive carbon nanotubes decorated with ligands, Fourches et al. demonstrated that QSAR approaches can be used to design NMs with desired profiles and biological effects (Fourches et al., 2015). When combined with existing experimental data, the use of QSARs allow for the grouping of NMs and can eliminate the need to test every novel material (Cohen et al., 2013; Liu et al., 2013; Lubinski et al., 2013; Winkler et al., 2013; Fourches et al., 2015; Godwin et al., 2015).
Various grouping approaches have been proposed by experts, but choosing the best approach that can aid in making regulatory decisions has been a matter of debate. NIOSH recommends grouping NMs based on the following physical states to improve safe handling and reduce worker exposure: (1) bound or fixed nanostructures (polymer matrix); (2) liquid suspension and dispersion; (3) dry dispersible NMs and agglomerates; and (4) nanoaerosols and gas phase synthesis (on substrate) (NIOSH, 2012). To ensure safe disposal of NMs, Hallock et al. suggested the following categories based on the product matrix: (1) pure NMs; (2) items contaminated with NMs; (3) liquid suspensions; and (4) solid matrices (Hallock et al., 2009). Another approach by Foss Hansen et al. suggests grouping NMs based on their location in the final product, for example, as part of a bulk substance (e.g., nanoelectronics), on the surface (e.g., films), or as particles (e.g., liquid suspensions) to determine the exposure potential of NMs (Foss Hansen et al., 2007). The Regulatory Cooperation Council (RCC), comprising members from U.S. and Canada, proposed a classification scheme based on chemical composition of NMs including carbon nanotubes, inorganic carbon, metal oxides and metalloid oxides, metals, metal salts, and metalloids, semi-conductor, quantum dots, and organics. The aforementioned approaches in addition to others have been reviewed in a recent report by the Dutch National Institute for Public Health and the Environment, RIVM (RIVM, 2015).
While these approaches based on chemical composition, physical state, and location of NMs in the products can provide information regarding the potential exposure, they fail to link inherent properties to observed bioactivity. Recently, the European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC) Task Force on Grouping and Integrated Approaches for Testing and Assessments of Nanomaterials developed a decision-making framework for grouping and testing NMs (DF4nanoGrouping) based on material properties, use and exposure considerations, biophysical interactions, uptake, biokinetics, and possible early and apical toxicological effects. The four main groups included in this approach are (1) soluble NMs, (2) biopersistent high aspect ratio NMs, (3) passive NMs, and (4) active NMs (Arts et al., 2015a). Using existing information for 24 materials covering carbonaceous NMs, metal oxide and metal sulphate NMs, amorphous silica and organic pigments, DF4nanoGrouping approach was able to group the materials in their respective sub-groups and could also identify NMs that needed further evaluation (Arts et al., 2015b). Such an approach that considers multiple aspects related to NMs and their biological effects can aid in regulatory decision-making related to risks associated with NMs. But for these grouping approaches to be useful in the risk assessment of NMs in the regulatory context, there is a need to generate high-quality in vitro data generated using human-based cells that can link NM properties to the observed biological effects.
Considerations for generating high-quality data
The need to generate high-quality data is critical to develop a greater understanding of the biological, environmental, and ecological impacts of NMs. There are a number of factors that define data quality from in vitro approaches, including (1) the choice of relevant cell types in a physiologically relevant configuration; (2) thorough characterization of the test-material throughout the assay, including life cycle transformations; (3) the choice of realistic test-material concentration and form relevant to real exposures; (4) the use of context-specific dispersants; (5) the use of assays with minimal or no NM interference to capture relevant mechanisms of NM toxicity, and (6) the use of appropriate exposure route and duration (Farcal et al. 2015; Sharma et al., 2016). Consideration of the aforementioned parameters during testing can generate information regarding the NM behavior and potential context-specific effects that can be useful in the risk and hazard assessment of NMs. It is critical to point out that the successful application of NM “predictive systems” (for example, see Meng et al., 2009; Nel et al., 2013; Gernand & Casman, 2014) is reliant upon high-quality data. Discussed below are a few points that are critical for the generation and use of quality data for regulatory decision-making.
Use of tiered approaches
Realistic exposure scenarios involve complex systems, and therefore, a tiered approach to toxicity testing is often the most comprehensive and effective method for NM assessment. A tiered testing approach uses a range of systems increasing in complexity, beginning with in silico or in chemico analysis, followed by in vitro tests including individual cell types and co-cultures and tissue or organotypic systems. Such tiered strategies can be fine-tuned to be fit-for-purpose (i.e., regulatory needs) while fulfilling minimum characterization needs and safety profiling (i.e., public needs). Similar to the EPA Endocrine Disruptor Screening Program (EDSP), triggers for higher tiers of data should be based on results of lower tiers and be based on transparent evidence-based physico-chemical and toxicological thresholds. At the problem formulation stage, intelligent (e.g., ITS-NANO (Stone et al., 2014)) and integrated approaches to testing (IATAs) should be defined per regulatory and public information needs. The Managing Risks of Nanoparticles (MARINA) risk assessment strategy calls for a two-phase process including “problem framing” (Phase I) and “risk assessment” (Phase II), during which minimal necessary data requirements for hazard, fate/ kinetics, and exposure are generated to satisfy the needs of the relevant exposure scenarios determined in problem formulation (Bos et al., 2015).
Use of standardized methods and materials
Standardization of methods is important for their use in hazard identification and risk assessment purposes. There are a number of in vitro methods available for testing chemicals and those methods can be applied to NMs, given the properties and possible interferences of NMs with the assays are considered. Previous publications have addressed the existence of well-characterized alternative methods and their relevance to NP studies (Hartung & Sabbioni, 2011) as well as human cell culture models useful for nanotoxicological investigation (Clift et al., 2011). The Scientific Committee on Consumer Safety’s (SCCS) Nano-Guidance (SCCS/1484/12) lists the in vitro methods that are used for traditional chemicals, along with recommendations on parameters to consider while testing NMs (SCCS, 2012). The SCCS report states that in the absence of scientific evidence proving otherwise (e.g., as has been demonstrated for the in vitro Comet assay), existing in vitro assays already validated for chemicals should provide valuable information regarding the toxicity of NMs (Petersen et al., 2014). In instances requiring the development of new methods, the ISO Technical Committee on nanotechnology and the OECD Working Party for Manufactured Nanomaterials (WPMN) have developed and are continuing to develop standards and guidelines for NM characterization and testing.
In addition to the standardization of methods, reference materials (RMs) should be used in comparison to the test material for method optimization and a better understanding of NM life cycle transformations. An RM is a “material, sufficiently homogeneous and stable with respect to one or more specified properties, which has been established to be fit for its intended use in a measurement process” (Roebben et al., 2013). The EC-JRC has established a repository that hosts different types of representative test materials. The U.S. National Institute of Standards and Technology (NIST) also provides standard RMs.
Development and use of a data-storage, -access, and -management system
Before generating new data, it is important to consider existing data. A centralized database allows for a space to collate and organize information. Consulting such a database before conducting a study will reduce duplicative testing and allow for the design of tests that more effectively address current research gaps (Johnston et al., 2013). Examples of NM databases include the cancer Nanotechnology Laboratory (caNanoLab) portal (https://cananolab.nci.nih.gov/caNanoLab/#/), NanoMaterial Registry (https://www.nanomaterialregistry.org/), and NanoHUB (http://nanohub.org/). A concerted effort to add information to these databases will further centralize NM data. The U.S. National Cancer Informatics Program has introduced a Nanomaterial Data Curation Initiative that seeks to define and implement some degree of “functional interoperability” across them (Hendren et al., 2015). Part of this effort will be a series of publications on NM data completeness and quality that are currently in preparation. Along with the need to house large volumes of data comes the need to rapidly analyze and query large data sets, as well as frameworks for data management and, often, novel hardware and software solutions to a particular data effort (“nanoinformatics systems”).
Organization of existing information
Adverse outcome pathways (AOPs) provide a framework to organize data that can be used to better understand the mechanism of action of chemicals including NMs (Ankley et al., 2010; Kleinstreuer et al., 2016). As part of a collaborative effort between the EC-JRC, the U.S. EPA, and the OECD, an AOP-Wiki has been created to provide an interactive and virtual platform for AOP development and to help scientists worldwide develop AOPs. A number of AOPs are currently under development and the OECD has also published a guidance document on ‘Developing and assessing adverse outcome pathways’ (https://aopkb.org/aopwiki/index.php/Main_Page) (OECD 2013). One challenge for using AOPs for NMs is the limited availability of information on how internal, external, and composition-based properties mechanistically influence pathways of toxicity (Lynch et al., 2014; Murphy et al., 2015). Towards this end, Hendren and colleagues have suggested a series of “intermediary, semi-empirical measures” (termed functional assays) that can “bridge the gap between NM properties and potential outcomes in complex settings” (Hendren et al., 2015). Well-designed in vitro assays can be used to study the different life cycle forms of NM and to correlate the NM properties to the activation of complex signaling pathways.
A consistent definition for NMs across jurisdictions is important for the assessment of NMs throughout the life cycle of NMs and the products that they are used in. The ‘what’ and ‘how’ of characterization is central to every discussion related to NMs as it defines the course of NM assessment. Useful non-animal testing approaches include (1) organizing existing information to develop conceptual frameworks (such as AOPs) to guide the assessment process, (2) use of in silico modeling to predict potential effects, and (3) use of in vitro systems (using human-relevant acellular, mono- and multi-cellular systems) to assess biological and ecological effects. Similar to other chemicals, the value of non-animal methods for the assessment of NMs has been widely realized as they enable a comprehensive (qualitative and quantitative) evaluation of NMs using a variety of approaches. However, there are specific considerations to contemplate, for example, it is not expected that in vitro methods will provide a 1-for-1 replacement for animal tests; rather, integrated approaches to testing and assessment that combine in vitro and in silico assays will be important, as will training scientists on how to use these approaches. Another hurdle is gaining global regulatory acceptance of non-animal methods, and achieving this goal will require collaboration between government, industry, academia, and non-governmental organizations worldwide. Standardization of non-animal methods is important to achieve inter-laboratory reproducibility of results. Establishment of data-sharing platforms can reduce the duplication of data and thereby improve the efficiency of toxicological research. Use of such non-animal approaches can expedite the process of testing thousands of materials awaiting testing while generating human-relevant data.