Health impact and toxicological effects of nanomaterials in the lung

Authors

Errata

This article is corrected by:

  1. Errata: CORRIGENDUM Volume 17, Issue 6, 1027, Article first published online: 29 July 2012

  • The Authors: Dr Kendall is an expert environmental scientist, with 12 years of international faculty-level experience who specializes in exposure measurement, nano-characterization and health-impact assessment of fine particulate matter and nanoparticles. She was awarded the prestigious Rosenblith Prize by the Health Effects Institute in 2004. Professor Stephen T. Holgate is an MRC Clinical Professor in Immunopharmacology with research interests in the mechanisms of asthma and its treatment. He has published over 980 peer reviewed papers and is co-editor of key textbooks. He is past Chair and now member of the UK Department of Health's Committee on Medical Effects of Air Pollutants.

  • SERIES EDITORS: IAN YANG AND STEPHEN HOLGATE

Michaela Kendall, European Centre for Environment and Human Health, Peninsula College of Medicine and Dentistry, The Knowledge Spa, Royal Cornwall Hospital, Truro TR1 3HD, UK. Email: michaela_kendall@yahoo.co.uk

ABSTRACT

The manufacture, use and disposal of nanomaterials will result in increased human exposures to engineered nanoparticles (ENPs), potentially via the lung. ENPs differ physically and chemically from natural- or combustion-derived nanoparticles (NP) in important respects. While there are parallels with ultrafine aerosol particles in the atmosphere and colloids in water, there remain some unique issues and impacts of engineered materials on lung health that require consideration and urgent study.

The study of toxicity of nanomaterials in biological systems—nanotoxicology—emerged from the observed effects of inhaled particulate matter (PM) and NP. Some engineered nanomaterials deserve special toxicological examination because of their unique properties in biological systems; novel toxicological approaches may be required for their assessment. Translocation in biological systems—a key feature of ENPs—is dependent on ENP size and surface interactions with macromolecules at the portal of entry, upstream of cellular interaction. Of particular significance is the agglomeration processes associated with macromolecule adsorption at ENP surfaces, which determine clearance rates and cellular response. ENP toxicity is therefore dominated by three linked physico-chemical factors: size-shape, surface and ‘corona’ (formed by adhering macromolecules from the suscipient host). Measuring and predicting ENP translocation and effects following lung entry have proven to be particularly challenging, but understanding ENP behaviour in vivo is fundamental for safe design for effective and targeted drug delivery. Human exposures via medical and dental applications appear important in terms of dose and toxicity, and may need to be assessed for risk on a case-by-case basis.

INTRODUCTION

Engineered nanoparticles (ENPs; typically defined as <100 nm in any one dimension1) are increasingly being utilized in a range of commercial and medical products.2 Most ENPs are currently used without toxicological consideration of the physical and chemical properties exploited in the products they make up, with some important exceptions (carbon nanotubes3). They do not need special toxicological review under law despite their novel properties, even though humans may be directly exposed to essentially new materials and/or receive elevated doses in rare cases.4,5 This raises concerns about the potential public health consequences of nanomaterial products released to the general market or in applications that result in direct exposures, such as nanomedicine. From toxicological studies, it is clear that reducing particle size to the nanoscale significantly influences toxicity, and so this strategy remains under discussion.6 ENPs provide fresh scientific challenges, related to their detection and detection of their effects in complex media (e.g. in the environment or biological systems such as the human lung), and some argue their toxicity must be considered.7

New materials, old paradigms?

Nanoparticles (NP) are not new, but many ENP are because of our improved ability to control nanomaterials on an industrial scale. While many common materials occur in the submicron range (nanoscale), processes have been developed to generate large quantities of uniform dispersions of nanoscale and ENP-containing materials, some of which do not occur in nature.8 These approximately monodisperse materials have unique properties used in products to improve everyday products.9 Some meticulously controlled properties of ENPs are new and surprising, and potentially provide both beneficial and problematic effects. Material size is often reduced to lower material usage and associated costs, exploit material surface properties more efficiently and develop unique applications, providing societal benefits.10 As nanotechnology becomes more widespread, second-generation ENPs are now being developed with complex formulations and structures.11,12 While the toxicity of ENPs measured today may have limited relevance for materials released in the future, in the next section, we discuss some common principles of toxicity that have emerged. Balancing risks and benefits is key.

New applications, new insights

The availability of new nano-analytical techniques increased our ability to detect and assess the nanoscale and contemporary exposures to NP. The volume of recent publications in nanotoxicology may be partially explained this way, and it is important to contextualize the new data: the lung has evolved under a narrow range of exposures to nanomaterials, and we can now examine its role for the first time using new tools.

For example, ENPs represent a new opportunity for studying the phenomena of material uptake and transport through the human lung, to study fundamental processes and mechanistic pathways previously not clearly defined. For example, by using tailored ENPs, the specific characteristics of airborne particulate matter (PM) determine that exposure-related health outcomes are being isolated. Via ENP studies innate and adaptive immune responses to ENPs are being isolated.13 Engaging commercial scientific knowledge gathered in the process of developing nanoproducts may support proper risk assessment on new ENPs under development.

However, the release of NP into the environment and the predictable human exposures must be managed responsibly because the public health and wider environmental impacts are—to date—largely unknown. In order to conduct an appropriate risk-benefit assessment, it will be important to match the speed of nanotechnology development with a clearly established process for risk assessment and regulation, prior to market release. Forensic examination of exposures and linked effects can prevent undue damage in humans. The ‘safe by design’ approach based on sound scientific evidence is crucial to avoid premature market introduction. Regulatory frameworks are currently under development to protect both public health and the environment.14 The financial, ethical and social dimensions of ENPs innovated for commercial uses also require serious consideration.15

This review presents the key evidence and understanding of risk following the exposure of the human respiratory system to NP. Assessing this risk is crucial not just to protect human health but also for the nanotechnology industry. Evidence for the toxicity of ENPs in humans is multidisciplinary was drawn from research in air pollution science, toxicology, engineering, medicine, nanoscience, and environmental and public health. A homogenized vocabulary accessible to all disciplines was used to avoid over-complexity and confusion. While other epithelial surfaces are also breached by nanomaterials including the gastrointestinal tract, conjunctiva and skin, these exposures were only considered in relation to the lung here. We examined the key types and sources of human lung exposures together with the proposed mechanistic explanations, and the broad scientific understanding this has generated.

KEY CHARACTERISTICS IN ENP TOXICITY

To examine the toxicity of ENP, it is critical to forensically examine the relationships between their physical and chemical character and biological effects. Many publications that develop hierarchies of the vast range of physico-chemical properties of ENP toxicity have been generated.6,8,11 All factors affecting NP effects may be investigated before interpreting ENP—cellular interactions, but human toxicity of nanomaterials can broadly be described as dependent on four interlinked physical factors:

  • • Size (at least one or more dimension of the material is nano)
  • • Shape
  • • Surface
  • • Corona

We will now examine the importance of these characteristics in the human respiratory system, after first examining the lessons we learnt from air pollution particle exposures.

Ultrafines: lessons from the lung

Because the nano-industry is embryonic, there are few epidemiological studies indicating human health hazards from ENPs, and fewer still were well controlled.16,17 In contrast, the epidemiology and toxicology of human lung exposures to ultrafines (or NP <100 nm) and fine PM (PM2.5; PM with aerodynamic diameter ≤ 2.5 µm) were widely studied around the world.18,19Figure 1 shows the similarities in morphology between airborne and certain manufactured ENPs. In these studies, ultrafines and regulated PM metrics (e.g. PM2.5) were consistently associated with increases in human and animal mortality and morbidity.20 In the lung, NP exposures were associated with pneumonia, chronic obstructive pulmonary disease and asthma, especially in susceptible groups. These dose–response relationships have proven robust in the face of significant scientific, political, industrial and legal challenge and are now widely applied. For example, using 2008 UK mortality data, human-made PM2.5 exposures of the UK population was cautiously estimated to cause the loss of 340 000 years of life in 2008, equivalent to 29 000 deaths.21 Combustion dominates as a source of these ultrafine particles, especially in urban areas (Fig. 2 (D. Tinker et al. unpubl.data)). The primary remaining uncertainty is the precise mechanism(s) of action of particles in the lung, but some of the key PM characteristics associated with the toxicity of combustion-derived NP in the lung have emerged.

Figure 1.

Comparison of selected engineered nanoparticles and urban atmospheric nanoparticles (<100 nm): Transmission electron microscopy images of (a, b) titanium dioxide and (c, d) silica; (e) scanning electron microscopy image of atmospheric nanoparticles collected in London. These nanoparticles tend to form larger agglomerates over time, especially in environmental conditions.

Figure 2.

The hourly average nanoparticle number concentration in air at a UK government monitoring site in Central London (Marylebone Road) shows the diurnal pattern of nanoparticles near a busy roadside over 24 h. (D. Tinker et al. unpubl.data) Combustion dominates as a source of ultrafine particles (<100 nm), especially in urban areas.

The American Cancer Society prospective cohort data for 1.2 million adults in the Cancer Prevention Study II provided a unique data set to analyse the lung and cardiovascular outcomes from a PM2.5 dose–response curve that covers several orders of magnitude of dose.22 Relative risks were estimated for increments of particle exposures from urban atmospheres, environmental tobacco smoke and cigarette smoking, carefully adjusting for various individual risk factors. A non-linear relationship between urban exposures to PM2.5 and cardiovascular disease mortality emerged, demonstrating a steep increase in risk at low mass concentration exposures (urban atmospheres) and flattening out at higher environmental tobacco smoke and active smoking exposures. At low exposure levels, cardiovascular deaths dominated the burden of disease—at high PM2.5 concentrations, lung cancer dominated. Air pollution can therefore be considered a risk factor in coronary disease.

Atmospheric NP are more effective at accessing the lung alveoli than larger particles because of their small size.23 Evidence suggests that the smaller particles (including the NP fraction largely resulting from combustion) may be especially important because they are cleared less efficiently by phagocytosis in the human lung.24,25 In normal particle exposures, mobile alveolar macrophages work to clear the particles; under chronic high particle loads, airway remodelling, inflammation and alveolar destruction (emphysema) and/or fibrosis occur.26 Impairment of pulmonary macrophage function, the decrease of monocytes/macrophage phagocytic activity and mobility, and impairment of macrophage cytoskeleton were induced by disease status, NP deposition and accumulation, indicating clearance dysfunction.27,28 Phagocytes either degrade the particle, remove it to the mucociliary escalator to be swallowed or transfer particles to the lymphatic system29 or blood vessels,30,31 leading to their wider dissemination. Their small size enables them to evade normal clearance mechanisms at higher concentrations to promote localized radical-mediated inflammation and translocation.32–35 Failure to clear NPs results in the material reaching the epithelial tissue and diffusion of ultrafine particles from the alveoli into the microvascular bed.36 Depositing particles accumulate and cross the alveolar-capillary membrane to the blood and spread to other organs.19,37,38 Several studies have reported that this disturbs immune regulation in the lung.39,40

PM-mediated disturbance of lung surfactant has also been suggested.41,42 Depositing material in the respiratory system first interacts with extracellular surfactant secreted by Type II alveolar epithelial cells, before encountering extracellular molecules and mobile cells before reaching lung tissue.24 Lung surfactant stabilizes alveoli by lowering surface tension and promotes clearance of inhaled material to maintain the alveoli in a sterile and inflammation-free state.43 Lung surfactant therefore plays a critical role in innate immunity, clearing infectious and other particles from the lung without recourse to secondary inflammatory immune responses which may compromise respiratory exchange.44 Perturbation of surfactant lipid and protein may therefore contribute to the observed pulmonary inflammation observed on particle exposure and helps explain the increased45–48 susceptibility to and severity of infections.49–51 Early studies showed that urban atmospheric NP interact with components of human lung surfactant: NP agglomerated in the protein-rich fraction of human lung lining fluid, and atomic force microscopy showed increased attractive and adhesive forces of a diesel particle to a graphite surface in lung lining liquid.33 Furthermore, analysis of urban PM2.5 surfaces in London and New York immersed in lung lining liquid indicated protein and phospholipid adsorption.52

Studies of pulmonary retention, extrapulmonary translocation and redistribution have demonstrated that a proportion of the inhaled ambient particles are retained in the lung36), a proportion are removed by macrophages25 and a small fraction (<10% of the deposited dose) may translocate to the circulatory system or stimulate immune responses. In circulation, they interact with atheromatous plaques to precipitate myocardial infarction, strokes or initiate cardiac arrhythmias and strokes.36,53,54 To establish what might drive these effects, in vivo exposures to both ambient, concentrated ambient PM and ENPs have been conducted, in normal and genetically modified animal models. When transgenic mice deficient in the apolipoprotein E gene mice were exposed by inhalation to well-characterized, concentrated ambient particles, oxidative stress was identified as a predisposing factor linking PM exposure and susceptibility to neurodegeneration and cardiac dysfunction.55,56 Early in vivo animal studies of ENPs demonstrated significant differences between the retention and clearance of inhaled ultrafine ENPs and coarse particles57 and establishing the new field of nanotoxicology.19 While the exact mechanism(s) remain unclear, the coefficients relating PM to cardiovascular, respiratory or all-cause mortality and morbidity are remarkably similar around the world.58

Size: slipping through the biological net

Nanomaterials have one or more external dimension or internal structure at nanoscale, and ENPs often exhibit novel emergent characteristics as compared with the same material at a larger scale. This can deliver unprecedented mobility in biological systems, and this forms the basis for their widespread potential in medical and other applications. Size is important in the translocation of ENPs from the lung: inhalation studies with iridium demonstrated that 80 nm NP translocated less than 20 nm ENPs (approximately an order of magnitude less at 24 h59). Gold ENPs that are 1.4 and 18 nm were also significantly different in their ability to translocate and accumulate in tissue, with 8% and 0.2% translocated at 24 h, respectively.60 Migration across cellular membranes via passive diffusion and endocytosis into cells is also size dependent, with polystyrene particles less <200 nm being able to enter the cytoplasm (Fig. 3).62,63

Figure 3.

Scanning electron microscopy images of nanoparticles agglomerated by protein, in this case, fibrinogen: (a) ∼200 nm plain polystyrene beads; (b) ∼120 nm silica (200 V) nanoparticles quickly formed micron scale agglomerates in a fibrinogen suspension.61

Within the scientific community, there is broad agreement as to NP definitions and terminology, and size standardization has occurred as evidenced from the literature and for regulatory purposes.1 However, arguments against legal ENP definitions have been put forward,64,65 in part because many toxicity effects are seen for particles above >100 nm in size. Nanoscale material (with one or more dimension within the nanoscale, or less than 1 µm) may be a more effective definition to acknowledge that size regimes are important but cannot eliminate uncertainty about the effects in humans as PM10 and PM2.5 effects demonstrate.15 The difficulty with defining nano-dimension is that at the nanoscale, many materials have features with nano-dimensions, and these features may play a role in cellular interactions, including adhesion.66

The size of the material influences the deposition patterns in the lung, particle clearance efficiency, cellular uptake and translocation.38 Practically speaking, this means size partly determines which cells contact and take up ENPs, and then which cellular compartment the ENPs are directed to. Such translocation is heavily dependent on differences in agglomeration rates: ENP agglomeration rate is similarly heavily dependent on both the system conditions and the surface interactions, and extremely small interfacial differences alter rheological properties to either promote or reduce agglomeration.66 Agglomeration is size, shape and concentration dependent: for example, agglomeration rates decrease with reducing concentration.67 It is therefore possible that high concentrations of ENPs may be less toxic than low concentrations if agglomeration promotes accelerated biological clearance. While manufacturers favour ENP formulations that do not agglomerate, agglomeration processes are critical to PM/ENP and infection clearance in the lung.68 Therefore safe design of ENPs requires a deeper understanding of these lung clearance processes.

Shape: not what it appears

Another key factor determining toxicity is particle shape, especially within submicron size ranges. Few nanomaterials are perfectly spherical, and the range of size/shape combinations is almost as wide as the number of materials: micelles, worm-like filomicelles, liposomes, polymeric, didemers, porous or hollow inorganic NP, nanofibres (including nanotubes, rods and fibres) and nanoplates are classed as nano.69–73 Shape also determines agglomeration rates66,73 and therefore influences translocation.

Fibres need special consideration:3,74 animal studies have shown a consistent toxicological response to carbon nanotubes and fibres (e.g. pulmonary inflammation, fibrosis and immunosuppression50,73 independent of study design (intratracheal instillation, aspiration and inhalation). Early onset and persistence of pulmonary fibrosis were observed in carbon-nanotubes-exposed animals in short-term and sub-chronic studies,74 and migration of multi-walled carbon nanotubes from the pulmonary alveoli to the pleura occurred (i.e. the site of malignant mesothelioma development in asbestos-exposed individuals76). Injection of long multi-walled carbon nanotube into the peritoneal cavity of mice (a surrogate for the pleura) induced inflammation and granuloma formation in the pleura, suggesting that multi-walled carbon nanotubes have asbestos-like pathogenicity.77 A major caveat of this particular study was that the animals were not exposed by inhalation, although others have shown inhaled multi-walled carbon nanotube do migrate to sub-pleural tissue.78 Single walled carbon nanotubes toxicity is enhanced by the presence of metal catalysts; however,79 and this leads to the consideration of the third important factor in ENP toxicity: surface.

Surface: action at the interface

Many of the intrinsic properties of nanomaterials harnessed in nanotechnology applications are related to nano-controlled surfaces.11,80 Surface area, solubility, reactivity/stability and adsorption capacity are related to the surface of the ENP and may have little relationship with the core material.61 These surface properties may also change over the life cycle of the particle in designed or unintended ways, making the ENP surface-system interface dynamic, determined by the components, and in some cases, timeline of the system. The surface material may not indicate the bulk material of the particle providing the ‘Trojan horse’ analogy; surface dictates the nanostructure and roughness of a nanomaterial surface by changing the outer molecular layers and requires different analytical techniques from the bulk.81 Polyethylene glycol coating creates micelle that form complex surfaces and surface chemistry to control surface interactions.72 Surface coatings may also divert particles to certain cellular compartments.82 Such ‘tunable’ surfaces are typical of the next generation of nanomaterials where ‘soft’ layered, patchwork and uniformly surface functionalized ENPs are designed to promote certain properties or for specific purpose such as increased targeted drug delivery.71,83,84 For example, ENP surface design can stimulate biological responses or increase cellular binding. The surface characteristics of inhaled nanomaterial will also influence its penetration through the lung surface fluid and determine which macromolecules adsorb to the ENP surface (next section). In this section, we consider the ENP surface properties affecting the interface—we consider the system effects on the interface in the next section.

Increasing surface area acts to increase the reactivity of a material by exposing more atoms or molecules. ENP solubility or release of atoms or molecules from the surface (e.g. in the antimicrobial properties of nano-silver) is a crucial and well-established mechanism of metal ENP toxicity85 and is well described elsewhere. Oxidative stress is equally well described in the literature,11 although the physical model of how oxidative stress occurs is less clear. Titanium dioxide (TiO2) as 25 nm particles produced a far greater inflammatory response in the lungs of rats than 250 nm particles, and a strong correlation between particle surface area and effect was observed.86 In biological systems, this can result in positive or negative effects: molecule release may be more readily controlled, providing more effective therapy;87 increased complexation with proteins may deactivate or remove essential proteins from the system;88 potentially, nanomedicine therapies may even selectively remove components in circulation.89 Using a variety of different nanomaterials, further studies confirmed that, in rodents, surface area was a critical factor in driving lung and other epithelial-induced inflammatory responses, although chemical reactivity was a further crucial factor.90 Indeed, surface area may be used as a metric of dose and regulation for some nanomaterials.5

Charge at the surface of ENPs also plays an important role in their uptake and interaction with cells, partially because it dictates colloidal behaviour within the system and interactions with charged molecules on the surface of cells.11,91,92 Positively charged surfaces behaved differently to identical polystyrene particles with more negative charge.63 Together with chemistry, surface charge or reactivity is an important factor driving both uptake and toxicity, which is independent of oxidative stress. ENPs influence protein fibrillation as a function of their chemical surface properties, with those with the strongest hydrogen bonding capacity causing retarded fibrillation.93

Dispersants are often added to either ENP suspensions or media systems to stabilize colloidal suspensions.94,95 This clearly changes the nature of the ENP-system interface, interfering with biochemical interactions (e.g. agglomeration rates) and measures of toxicity.96 Biologically, agglomeration acts to reduce available surface area, increase particle size and promote clearance, and cells respond differently to agglomerated or de-agglomerated ENPs.73,97 After injection of ENPs into mice, coatings such as polyethylene glycol prolongs the circulating particle half-life and accumulation patterns by reducing agglomeration.98

Low-toxicity material (including carbon and TiO2) can induce acute inflammation in vivo in a surface area dose-dependent way for microparticles and ENPs.99 TiO2 induced responses that include persistent pulmonary inflammation and cancer in rats and mice.100 On the basis of these studies, National Institute for Occupational Safety and Health qualified ultrafine TiO2 a potential occupational carcinogen and considered surface area in setting occupational regulatory relative exposure limits (5). Insoluble low-toxicity ENPs may well have a different mechanistic pathway, which is more dependent on the receiving host (suscipient) system. Interfacial effects of the ENP are dependent on both the original ENP surface and the surrounding system. Identified a decade ago on airborne particles in the lung and in the gastrointestinal tract, we now consider the newly termed ‘corona’ formed around ENPs in biological systems and subsequent effects on biological behaviour.

Particle coronas: the changing face of ENPs

Proteins and other macromolecules have long been understood to attach to non-self surfaces in the human body.101 For example, protein attachment to particles was understood to alter particle uptake in the gastrointestinal tract for many years: surface modification could increase or decrease the uptake of variously sized polystyrene latex particles between 50 nm to 3 µm.102 Recently, scientists developed methods to identify the composition of ‘coronas’ and measure their physico-chemical behaviours.103 Corona formation leads to physical change in the ENPs, obscuring the original particle surface to form a new structural surface, affecting adhesive properties and changing agglomeration behaviour.66 What has emerged is that together with the size and original surface of the ENPs, the adherence of molecules to the ENP surface from the system is also key to toxicity effects, with significance for both toxicological and immunological behaviour. In discussing biological effects, the portal of entry largely dictates attachment. This has important implications for ENPs deposited in the respiratory system.

Adsorption of plasma proteins in serum demonstrated the role of protein adsorption in predicting clearance, toxicity and efficacy of drug delivery.104 While similar processes are expected in the lung, harvesting particles from that area is much more challenging. In serum, the protein corona composition depends on the plasma concentration and collects residues of multiple solution systems as it proceeds through them.105 As the original ENP surface is crucial to the interaction with the body, sorbed material at ENP surfaces confers biological identity via the attached molecules.47,103 The ‘corona’ offers a measurable particle history for inferring and identifying new pathways and interactions, just as the analysis of airborne particles can be used to investigate their source.

Molecules (ligands or bioconjugates) attached to the surface of ENPs change ENP biological interactions and functions. For example, surface charge and inner composition of porous NP are key factors in determining the ability of particles to cross the blood–brain barrier in vitro.106 ENP coatings in biological systems may enhance or reduce membrane crossing and cellular penetration depending on composition and the surrounding system.107 Gold NP up to ∼39 nm bound to the nuclear core complex protein to pass through nuclear membranes. Attachment of peptides to the surfaces of particles and ENPs is also used to generate antibodies against specific peptides; ENPs provide large surfaces to which peptides may attach and expose epitopes.108 Serum albumin has been shown to induce uptake and anti-inflammatory responses in macrophages, which were not present when the particles were pre-coated with surfactant to prevent albumin binding.109 Specific protein signatures, cytokines and amino acids have also been noted to accumulate on ENP and PM surfaces.42,47,94 Lactate dehydrogenase—a stable cytoplasmic enzyme present in all cells released into cell culture media upon damage of the plasma membrane—is widely used as a marker of cytotoxicity due to membrane damage. However ENP surface uptake may also interfere with this test.110

Corona formation is highly likely to be important in the lung in determining fate and the corona formed may be unique to this portal of entry. The adsorption onto a large surface area has the potential to deplete or deactivate defensive molecules.88,111 This may explain individual variation in susceptibility to the same material, where the suscipient system components differ minutely to influence health outcomes. Such a concept also contributes to the apparent disparities between particle uptake and systemic circulation in animals and in normal humans.100,111

The corona tends to equilibrate with the surrounding system, so that high-abundance proteins binding initially are replaced gradually by lower abundance, higher affinity proteins. Changes in the biomolecule environment, such as uptake or biodistribution will be reflected as changes in the corona. There are many techniques available for analyzing the size, composition and agglomeration behaviour of the corona;112 for example, heat inactivation of the proteins alters cellular uptake.113 Corona can distinguish differences that morphology measurements are unable to do. For example, fetal calf serum proteins adsorb differently to specific proteins, resulting in measurable differences in agglomeration behaviour.114

New nano-analytical surface techniques have revealed that physico-chemical changes in ENP surface can lead to activation of defence cascades and trigger attempted clearance by multiple types of phagocytic cell. For example, the corona acts to:

  • • Agglomerate PM and promote clearance
  • • Separate the ENPs from cells via the adsorbed barrier
  • • Modulate inflammatory responses, especially in the lung

MECHANISTIC MODELS OF TOXICITY

ENPs are widely reported to traverse tissue barriers and cause damage in the respiratory and gastrointestinal tracts and brain. Both in vivo and in vitro models were developed to study the mechanisms but often fail to correlate well because of the variety used and the different dosing regimes.26,100,115 Both have inherent limitations for studying ENPs: for example, in early in vitro studies, the surfaces of the ENPs were poorly controlled so that contamination by synthesis by-products such as metals or endotoxins have contributed to the reported effects.116 Single-cell-type models tend to oversimplify behaviours in vivo,117 while reduced use of animal models has limited in vivo study.64 In addition, the appropriateness of in vivo exposures using labile or non-representative radioactive labels at ENP surfaces has been questioned.118

Nanomedicine has taken a much more deliberate approach to ENP surface control, contamination and particle behaviour in vivo.96 In the diverse literature, several toxicity mechanisms have been widely reported, related to one or more of the key toxicity features of ENPs and often with parallels with air pollutant particle toxicity profiles. We now consider the specific mechanisms most significant for ENPs in the lung, where consensus has been reached.

Enhanced translocation

The ability of ENPs to translocate from one compartment to another is a key factor in their toxicity. As discussed earlier, ENPs may be dwarfed by key biomolecules involved in normal cell function; their shape may prevent normal clearance; the attachment of biomolecules (or fragments thereof) to ENP surfaces facilitates specific transport processes, for example across cell membranes. This unprecedented mobility in biological systems is key to assessing ENP hazard.

The ability of ENPs to move across tissue barriers is highly dependent upon size, shape and surface.38 Translocation across the respiratory epithelium allows the passage of particles to the circulation or lymphatics, and provides systemic access to inhaled drugs such as insulin (Exubera (Nektar Therapeutics, San Francisco, CA, USA)119). Reduced clearance of ENPs by alveolar macrophages in the lung periphery increases systemic circulation25,120 or high deposited doses may lead to activated macrophages via the release of inflammatory mediators.110 ENPs pass into lung epithelial cells in a size- and surface-dependent way via various pathways including clathrin- or caveolae-mediated endocytosis. Enhancement of specific pathways and fate of ENPs in cells may be achieved by attaching protein to their surfaces,121 utilizing similar uptake and intracellular transport routes to microorganisms and indicating that their coronas may also be similar.41 At cell membranes and inside cells, low-solubility ENPs are visible as agglomerates in lysosomes and singlets/small agglomerates in the cytosol within short time frames;122 agglomeration to larger sizes through corona growth may prevent NP from entering cells.66,67 The corona modulates particle toxicity until degraded in lysosomes, when the particle surface is once again exposed and the agglomerates may disperse.122 Later, they are visible in the mitochondria and within nuclei where they can trigger apoptosis,123 initiate metabolic dysfunction (recently observed in PM exposures124) or interfere with genetic and epigenetic targets.26,125 Positively charged particles appear to exhibit greater toxicity, but this is not clear whether this is due to instability in the media creating loosely agglomerated particles, differences in coronas or intrinsic toxicological difference.63

Enhanced translocation allows access to organs not normally reached by exogenous material. ENPs deposited in the lung accumulate in organs of the reticuloendothelial system rich in mononuclear phagocytes, including lymph nodes (in monocytes/macrophages), liver (Kupffer cells), spleen (histiocytes) and kidney.38 Intravenously injected ENPs travel around the body, deposit in tissue over a period of hours and may accumulate in the lung.126 Lymphatic uptake following intravenous injection is dependent on particle size, with ENPs >20 nm being retained in lymph nodes. Smaller injected ENPs also enter the reticuloendothelial system and are able to cross the blood–brain barrier.127 Orally administered ENPs accumulated in the liver, spleen, lymph nodes and bone marrow,128 whereas inhaled particles have the capacity to stimulate the production of monocytes in humans and rabbits129 associated with accelerated arterial atheroma plaque formation.130 Organs of the reticuloendothelial system consistently take up exogenous substances in this size range,131 and therefore techniques to reduce either the accumulation or toxicity of nanomedical products involving this uptake pathway are being investigated.132 In rodents, ENPs are also transported from the nasal mucosa to the brain via the olfactory bulb and are capable of inducing an inflammatory response,133 drawing comparisons with the polio virus which gains access to the central nervous system via this route and providing a potential route for delivering ENP-bound drugs to the brain. This ability for enhanced translocation is therefore key to ENP access to organs, tissues and cells.

Cardiovascular mechanisms

The observed cardiovascular effects of environmental inhalation of NP (including air pollution and smoking) have led to a growing understanding of how small particles have effects well beyond the first point of contact with the lung. The epidemiological findings linking atmospheric NP to cardiovascular disease can be explained by a number of potential mechanisms: (i) inflammation-enhanced atherogenesis and plaque destabilization in coronary and brain arteries; (ii) epithelial/endothelial interactions with NP alter the blood's clotting status or fibrinolytic balance favouring thrombogenesis; (iii) NP or their soluble components (e.g. metals or organics) enter the circulation to exert direct effects; and (iv) augmenting cardiac arrhythmias.26

Both in humans and in animal models, exposure to particulate pollutants induce platelet activation and intravascular accumulation to promote blood clotting.65,134,135 Although different biomarkers of coagulation were measured in many studies, vehicle pollutant particles (especially PM2.5) were consistently linked to thrombosis. Studies on particulate air pollution have led the way with a number of panel studies showing ventricular and supraventricular arrhythmias linked to ambient particle exposure especially in older people.136 Particulate air pollution is further linked to enhanced platelet activation in those at risk of coronary heart disease.137 A range of low-toxicity ENPs has also demonstrated platelet activation properties in vitro: silver,138 carbon,139 silica140 and polystyrene.66

Oxidative stress

The principle damaging mechanism associated with intracellular ENPs is through activation of oxidant pathways in the respiratory and digestive tracts, skin and eyes. Oxidative stress is one of the main mechanisms of toxicity associated with ENPs,11,141 and especially metal-containing ENPs.142 Reactive oxygen species damages proteins, lipids and DNA and participate in the development of diseases including systemic inflammation, cancer, arteriosclerosis and arthritis. Intracellular reactive oxygen species generation can trigger necrosis and apoptosis,143 mainly coordinated by specific transcription factors and the activation of specific cell signalling pathways.144 Some ENPs demonstrated unexpected properties, fullerenes for example, are highly anti-oxidant due to scavenging of free radicals.145

The formation of the surface corona indicates that the original ENP surface is not available for reactions. Researchers have therefore suggested that toxicity is either independent of oxidative stress146 or is oxidative stress mediated by interference with proteins and enzymes, potentially confusing cell signalling pathways.42,147

Immune responses

ENPs have both immune-stimulatory and immune-suppressive effects: antigenic and adjuvant properties may stimulate inflammatory responses, while interference with key molecules at surfaces and a subsequent development of tolerance may establish a quenching or inhibitory immune response.148 To date, the research focus has been on inflammatory responses to ENPs, and only a few studies have examined an immunosuppressive role.149,150 Pulmonary exposure to ENPs induced inflammation and acute phase response in the lungs of exposed mice.80,151,152 Both innate and adaptive immunity must be considered: depositing particles and ENPs induce innate immune responses by stimulating epithelial cells to increase lung resilience to microbial infection, and in particular, the surfactant protein (SP) D.153,154 The two lung surfactant proteins SP-A and -D regulate inflammation in the lung, and inflammatory lung diseases such as asthma, cystic fibrosis, lung infections and emphysema are associated with the dysfunction of surfactant.155 These patient groups may therefore be more susceptible to ENP exposures if they interfere with lung surfactant components or induce inflammation. SP-A, surfactant protein D and a recombinant fragment of surfactant protein D are involved in inhibiting infection, allergy and in the protection of the lung and other mucosal surfaces from pro-inflammatory stimuli.156,157 The recombinant fragment of surfactant protein D has anti-inflammatory activity in murine models of inflammation.155 Recent works show how surfactant protein D can be taken up by ENP surfaces, removing the protein from the system in a surface-area-dependent way and potentially promoting chronic inflammation in the lung.66,156–158 This may explain why ENPs administered to the lungs tend not to stimulate an inflammatory response in acute exposures159 but may promote inflammation over chronic exposure.

ENPs are used to serve as immunological adjuvants to enhance vaccines, and other immune responses occur via their uptake into dendritic cells. After antigen uptake, under the influence of chemokines, lung dendritic cells migrate from the epithelium through the mucosa to local lymph nodes, where they present their processed antigens via MHC Class I or II to näive T-cells. In the lung, ENPs are taken up by dendritic cells34 and macrophages and in this way are transported across the epithelium.160 Dendritic cell uptake of antigen-conjugated particles co-delivered with lipopolysaccharide (i.e. within the same phagolysosomes), induced T-cell proliferation.161 Therefore, the accumulation of ENPs in lymph nodes has provoked interest in the role of dendritic cells in ENP processing with parallels existing in the gut where there is evidence that antigen complexed with microparticles may be involved in immune disruption.161 It has been suggested that exogenous microparticles in the diet may be one factor responsible for Crohn's inflammatory bowel disease where there is strong evidence for disordered monocyte/macrophage scavenger function.163 Increased tissue barrier permeability in diseases such as eczema, asthma, celiac and inflammatory bowel diseases will tend to increase particle uptake and systemic toxicity.164 Understanding the mechanisms of immune recognition will become an important part of ENP design, especially in nanomedicine where it could affect the drug delivery efficacy and toxicity.96,132

Complement activation by ENPs in the systemic circulation system is another potential route to stimulating an inflammatory response and has been implicated in the cardiovascular effects observed in ultrafine PM exposures such as atherosclerosis.165 Dysregulation of toll-like receptors is widely understood to contribute to uncontrolled inflammation and metabolic syndromes, which contribute to the development and progression of chronic diseases, such as atherosclerosis, rheumatoid arthritis, asthma and cancer.166 ENP surfaces and curvatures with access to all cellular compartments could stimulate such interference via protein interference or coronas.

HUMAN EXPOSURES

Direct ENP exposures

Humans are constantly exposed to NP. Modern human exposures to atmospheric particles and NP were well described. (D. Tinker et al. unpubl.data) 167,168 An estimate for the number of atmospheric NP depositing daily in the human lung is 5 × 1012.169 In the UK, about 40 mg (1012) of exogenous microparticles (mainly silicates and TiO2) are ingested orally per person per day as food additives, pharmaceutical/supplements or toothpaste.163 Experiments on European drinking water have shown that 1 L contains typically 7 × 1011 natural NP, averaging 15 nm diameter, which have passed through the normal filtration and treatment processes.170

Inhalation is a key exposure route for ENPs in consumer, occupational and environmental settings, and significant efforts are active in this area of research. The increasing utilization of ENPs in nanomedicine for drug delivery, selective imaging, dentistry and wound treatment has increased ENP interactions with circulatory components. ENPs are also found in an increasing range of products, leading to direct and indirect human exposures to specific nanomaterials.171 Between 2005–2010 a steady growth in products claiming to be nano-based have been reported;2 >1300 products were listed in this inventory in 2010 compared with 54 in 2005, even though a ‘nano-claim’ was no guarantee of nanomaterial inclusion in the product, and lack of claim was no guarantee of absence.172 Such inventories provide a clear indication of market growth and establish a list of materials to conduct exposure assessments. Perhaps of particular significance to the lung of consumers are sprays (e.g. deodorants, suncreams, medicinal or for coatings) and ENPs in direct contact with humans such as silver used in clothing or food containers.

ENP toxicity in humans is dependent on exposures and their systemic bioavailability, not just toxicity or hazard. Exposure assessments attempt to define human exposures by identifying nano-enabled products using inventories (e.g. Woodrow Wilson2), through the application of mass flow models to calculate risks from current production volumes173 and predicting releases during the product life cycle. Because the toxicokinetic profiles of each type are expected to differ, we now consider selected ENPs which dominate the market in terms of their current production volumes or are expected to result in human lung exposures via medical and consumer products especially nano-silver, gold, titantium dioxide, polystyrene, silica, polystyrene and iron oxide. Direct human exposures are expected principally through medical/dentistry, food and consumer products.174

Silver ENPs are now being utilized extensively for antimicrobial purposes, and increasingly in clinical practice. Debate continues as to whether the nano-form per se is active as a microbicide or whether it is the release of silver ions within bacteria that provides the toxic effects. Nano-silver is also found in a wide range of consumer products.172 Colloidal silver (silver hydrosol) is advertised as an ‘alternative therapy’ and can be purchased from health food outlets. It is also found in clinical wound dressings (e.g. Acticoat (NUCRYST Pharmaceuticals, Fort Saskatchewan, AB, Canada)), socks and towels (to reduce bacteria and odour), toothpaste and cosmetics. Kitchen utensils and household appliances employ the microcidal effect of nano-silver (e.g. inner surfaces of refrigerators, air purifiers, vacuum cleaners, hair trimmers and food containers). It is also being considered for incorporation into food packaging to reduce food decay.15 Antimicrobial paint containing nano-silver is proposed for hospitals, schools and offices. The World Health Organization only considers silver to be a toxic at very high doses, and maximal concentrations are defined for water and air that currently cover nano-silver, despite the fact that ENPs may exhibit significantly different properties. Investigations in an in vitro diffusion cell system demonstrated the detection of nano-silver in the stratum corneum and outermost layer of the epidermis in intact and damaged human skin.175 To date, no studies have investigated the long-term effect on normal and damaged skin of nano-silver. As with most ENPs, nano-silver can be employed in drug delivery, specifically carry anti-cancer therapeutics, protein and DNA, and serum levels of silver were increased following use of silver-coated wound dressings.176 Several clinical trials of nano-silver enabled therapeutic agents are registered as on-going in the US including silver central venous catheters, nano-silver bacterial gel and silver biomaterial nanotoxicity: no current clinical trials are registered in the UK.

Nano-gold is found in tens of consumer products including cosmetics, toothpaste and a variety of antibacterial products. They are used in durable paint, water purification, faster computers, tougher shoe soles, and lighter and cheaper televisions. Gold ENPs are also employed in biodiagnostic applications, enabling the identification of a number of infectious diseases via colorimetry,177 and under investigation for drug delivery and in the detection of cancers.178 US clinical control trials using gold ENPs include treatment of atherosclerosis, anti-cancer agent vectors and investigation of the potential of gold NP as an anti-cancer agent.179

Nano-TiO2 is commonly found in sunscreens and cosmetic products due to its ability to provide barrier ultraviolet protection. In the UK, the Boots' (Nottingham, UK) product Soltan sun care range is one of the market leaders and contains TiO2 ENPs, and in 2003, this range comprised approximately 49% of the market projected to be worth approximately 290m USD. Penetration of TiO2 is negligible for healthy skin.175,180 Other commercially available TiO2 products include cleaning products, coatings for self-cleaning surfaces and car waxes and polishes (e.g. Turtlewax (Chicago, IL, USA)). Clinical applications of nano-TiO2 are limited, but in vitro and preclinical studies have suggested a therapeutic role in the treatment of a variety of tumours.84,181

Silica ENPs are produced on an industrial scale as additives to food, cosmetics (including some of the Lancome range and Leorex), car tyres, drugs, printer toners and varnish. Highly luminescent silica ENPs have been developed for the selective tagging of a wide range of biomedically important targets, such as cancer cells, bacteria (enabling identification of infectious diseases) and individual biomolecules. Nano-silica is also proposed as drug delivery vectors. Drug molecules loaded into surface-modified silica ENPs with bio-recognition entities (antibodies or proteins) allow specific cells or receptors in the body to be located.61,182 However, the toxicity of these ENPs has not been fully established. In the US, a clinical control trial is investigating silica NP in the treatment of atherosclerosis and as dentures containing silica NP to improve durability. Silica NP are already found in dental fillings (nanofillers) that are now employed commercially (Filtek; 3M, Berkshire, UK).

Polystyrene ENPs are the most widely used polymeric ENPs owing to their low cost and commercial availability. Commercially, they are found in disposable coffee cups and cutlery, food containers and CD cases. They also have a role in drug delivery systems,70 fluorescence imaging,182 in cancer diagnosis and in identifying trace amounts of infectious diseases (including anthrax, adenovirus and malaria).177

Applications of iron oxide ENPs include multi-tera bit storage devices, catalysis, sensors and a platform for high-sensitivity biomolecular magnetic resonance imaging for medical diagnosis and therapeutics. Recently, iron ENPs have been widely used in coal industry to produce clean fuels due to their catalytic activities that facilitate the chemical reactions to form and cleave carbon–carbon bonds. One key application of iron oxide ENPs is in human biomedical applications, such as labelling and magnetic separation of biological materials, imaging and diagnostic applications in human, site-directed drug delivery and anti-cancer hyperthermia therapy.183 Iron ENPs enhance the permeability of cells through the production of reactive oxygen species and the destabilization of microtubules.184

The incorporation of ENPs into medical products appears to be one of the largest non-occupational exposure routes for humans. Systemic delivery of therapeutic agents via inhalation of ENPs for organ, tissue or tumour targeting remains an attractive, non-invasive means of administration.185 Metal oxide NP are used for non-invasive vascular and tumour imaging.186 Dental materials also increasingly contain ENPs for infection control, improved material lifetimes and properties and improved biocompatibility.187

In line with other environmental contaminants, foetal exposures remain a growing concern. Intravenous and inhalation exposure studies, plus cell model exposures, have demonstrated placental transfer,188 complications in pregnancy and damage in the foetus.189 Damage mechanisms of intravenous nano-silica in mice included increased coagulation, inflammation and/or through oxidative stress.190 Finally, in mice, pregnancy enhances lung inflammatory responses to otherwise relatively innocuous inert particles such as TiO2, and exposures of non-allergic pregnant female mice to inert or toxic environmental air particles can cause increased allergic susceptibility in the offspring.190 While few of these materials tested would lead to maternal exposures or be circulating at these concentrations, the evidence suggests that some material effects are likely, and caution may be applied to particular maternal exposures.

Occupational exposures

Occupational nanomaterial release scenarios are always at the upper end of exposure ranges, and ENP exposures at work are being newly established. Measurements of occupational exposures (e.g. in the workplace191 and from product testing171) and models of exposure191 establish likely upper levels. Some lessons on how small particles behave on entering the body of humans and animals can be obtained from experience gained over the last 100 years in occupational dust exposures, but it is many years before effects may be observed observed, for example mesothelioma resulting from asbestos exposures.192,193 Currently, research scientists, medical and dental professionals often work with nanomaterials without assessment. Occupational exposures in manufacturing facilities tend to be more stringent, although elevated concentrations are associated with materials handling and even small production facilities.194 As demonstrated in accident or emergency situations, ENP exposures to combustion products from burning nanomaterials lead to first responder or fire victim exposures.17 In China, seven young female workers, exposed to NP for 5–13 months, all with shortness of breath and pleural effusions were exposed to NP consisting of polyacrylate and had lung tissue displaying non-specific pulmonary inflammation, pulmonary fibrosis and foreign-body granulomas of pleura.16 Transmission electron microscopy revealed NP observed in the cytoplasm and karyoplasms of pulmonary epithelial and mesothelial cells, but are also located in the pleural fluid. Such cases arouse concern that long-term exposure to some NP without protective measures may be related to serious damage to human lungs. Clearly, in some countries, regulation may not be so careful, and higher worker and environmental exposures may occur.

Environmental exposures via air, soil and water

Some engineered nanomaterials will be released to the environment in significant quantities.94,194 This may give rise to chronic, complex, multi-component ENP exposures that are generally not considered in toxicological exposure studies. Once in the environment, some ENPs will remain in circulation and lead to human exposures and certain materials in particular are being studied: nano-sized silver, TiO2, silica, zinc oxide, alumina, carbon black and carbon nanotubes. Modification of ENP size distribution and surface may also occur in the environment, affecting the way ENPs interact with the human body and influencing toxicity, in some cases, reducing, and in some cases, increasing effects. In terms of dose, these exposures are expected to be much lower than in the direct exposures to medical or consumer products.

REMAINING CHALLENGES IN DETERMINING ENP TOXICITY IN HUMANS

ENPs provide some scientific challenges common in environmental health sciences. Research priorities related to toxicity effects are reasonably well defined in international research programmes, such as the EU FP7, but currently, the evidence is derived from the ecotoxicological threats. In human nanotoxicology, the current gaps may be defined as:

  • Very limited human exposure information is available, partly due to the difficulty in tracing nanomaterials in products, the environment and biological systems.
  • Regardless of the route of exposure, but especially via lung exposures, ENPs have the potential to enter blood or have profound effects on the cardiovascular system. Few studies outside the air pollution field have illuminated the mechanics of cardiovascular events that result from respiratory exposures. Detection of ENPs remains an important issue.
  • Mechanistic determinants of the toxicity of specific ENPs or groups of ENPs are not well defined, and therefore the key characteristics of nanomaterials responsible for these effects are not clear. Surface area and surface chemistry have repeatedly emerged as important determinants and are often poorly reported in toxicology studies.
  • Studies comparing carefully controlled ENP and PM exposures in animals are not being widely conducted, potentially because access to such facilities is very limited due to prohibitive cost and sustainable science planning issues. Similarly, studies in genetically modified animal models have been limited, leading to a lack of studies in animals that relate to human toxicity and restricts adequate hazard identification in relation to human exposures.
  • ENP effects in susceptible groups to airborne particles (e.g. neonates, children, older adults, diabetics) are not understood.
  • Chronic exposure effects are not understood, limiting understanding of the role of ENP in cancer, neurodegenerative disease and fibrosis.

The key scientific priorities in determining ENP toxicity effects in humans may therefore be summarized thus:

  • Create nanomaterial toxicity research programmes to integrate human exposure and toxicity data as a necessity of commercial product release.
  • Develop a coordinated and standardized effort to quantify human exposures to ENPs using exposure models and key release/exposure data from certain products meant for human consumption, plus important occupational and environmental exposures.
  • Develop non-invasive/non-destructive detection techniques to identify ENPs in the environment and in tissue, especially those capable of distinguishing carbonaceous material at very low concentrations.
  • Develop and invest in in vivo testing for ENPs to establish hazard identification models and mechanistic pathways in susceptible in vivo models.
  • Develop toxicity assay tests to relate to in vivo exposures.
  • Utilize the positive characteristics of ENPs to study fundamental processes in cells, organs and the whole body.
  • Communicate these risks responsibly and effectively to society, industry and especially policy makers to establish credible communication between scientists at the cutting edge and the decision-makers contributing to decisions on widespread human exposures.
  • Integrate emerging information into industrial-production-processing decision-makers to develop products that are ‘safe-by-design’.
  • Ensure liability is clear and insurance of nanoproducts is provided. There remains significant uncertainty related to nanotechnologies, and definitions in law are kept open at this time to prevent excluding materials that may cause harm from protective legislature.
  • Emerging technologies such as nanomedicine needs to avoid the hyperbole and ensure that risk and benefits are carefully balanced.
  • Ensure that there is good public engagement in nanotechnology.

CONCLUSIONS

The application of ENPs is rapidly growing, and toxicology for key commercial ENPs is lacking because the law tends to treat nanomaterials as other materials, not accounting for their special properties. To establish the classic link between exposure and effects, much more evidence is required, recognizing that ENPs demonstrate unprecedented mobility for exogenous material in complex biological systems. Toxicity in some biological systems has been clearly shown, but this is highly dependent upon size, shape, material surface and corona by determining access to biological compartments that exogenous material does not normally reach. Effects may also be modified by individual susceptibility of the exposed host. At this moment, it is not clear that sufficient exposures will occur to merit concerns in the widest population. However, in high-dose groups, perhaps with susceptibility, one may expect to see ENP effects more regularly as release to market increases. The legal frameworks to establish adequate consumer, occupational and environmental protection require urgent attention.

ACKNOWLEDGEMENTS

The European Centre for the Environment and Human Health (part of the Peninsula College of Medicine and Dentistry which is a joint entity of the University of Exeter, the University of Plymouth and the NHS in the South West) is supported by investment from the European Regional Development Fund and the European Social Fund Convergence Programme for Cornwall and the Isles of Scilly. Stephen Holgate is a Medical Research Council Professor. The images in Figure 1 were produced by Dr Bjorn Stolpe of FENAC, a NERC-funded facility.

Ancillary