6. What are potential harmful effects of nanoparticles?
- 6.1 Can nanoparticles interact with living organisms?
- 6.2 Which characteristics of nanoparticles are relevant for health effects?
- 6.3 How can inhaled nanoparticles affect health?
- 6.4 What are the health implications of nanoparticles used as drug carriers?
- 6.5 How should harmful effects of nanoparticles be assessed?
- 6.6 What are the effects of nanoparticles on the environment?
6.1 Can nanoparticles interact with living organisms?
The SCENIHR opinion states:
3.7 The Potential for Interactions Between Nanoparticles and Living Systems
The hierarchical self organization of life spans from single molecules around 1 nm in size to large animals and plants (~10 m) and to very large organized populations of a species (~100 m). Nanoparticles may be of the same dimensions as some biological molecules such as proteins and nucleic acids. Many of these biomolecules consist of long macromolecular chains which are folded and shaped by cooperative and weak interaction between side groups, H-bridges and salt bridges. Here, functionalized nanoparticles, such as colloidal gold (Hayatt 1989), may intrude into the complex folded structures (Cheng et al 1999, Hainfeld and Powell 2000). Evidence for such interactions is seen from the experience with immunolabelling (Romano and Romano 1977) and related surface functionalisation techniques to target nanoparticles to biomolecules as markers for high resolution Transmission Electron Microscopy and optical imaging systems. Other nanoparticle systems which are established for research purposes in cell systems include quantum dots (Chan and Nie 1998) and magnetic nanoparticles (Josephson et al 1999). For a recent review see (Penn et al 2003). Surface active agents have been shown to alter the path of nanoparticles (Schurch 1990).
3.7.2 Nanoparticles in Living Systems – The Surface Effects
All nanoparticles, on exposure to tissues and fluids of the body, will immediately adsorb onto their surface some of the macromolecules that they encounter at their portal of entry. The specific features of this adsorption process will depend on the surface characteristics of the particles, including surface chemistry and surface energy, and may be modulated by intentional modification or functionalisation of the surfaces (Schellenberger et al 2004). This is well demonstrated through the use of specific biomolecular linkers that are anchored on the surface of nanoparticles or within vesicles and liposomes (Nardin 2000). In this way the affinity of a nanoparticle can be shaped to fit to a particular protein, and thus target a specific biomolecular assembly on a membrane, or within a specific organelle or cell surface. The specificity of such surface layers is used for analytical purposes (Elghanian et al 1997), for optical labelling of biomolecules in molecular libraries (Han et al 2001) and for drug or gene delivery to cells (Hood et al 2002). Thus, both the existence of passive surface layers and surface active agents compromise the risk evaluation of nanoparticles by mere chemical composition. In agreement with bulk surface chemistry, metallic nanoparticles are of considerable chemical reactivity while ionic crystal nanoparticles have been observed to accumulate protein layers when exposed to the cytoplasm or in the lymphatic fluid. This protein layer is possibly involved in the interaction of the nanoparticle by the cellular system.
3.7.3 The Effects of Size, Shape, Surface and Bulk Composition
The interaction of nanoparticles with living systems is also affected by the characteristic dimensions. As noted above, nanoparticles, of a few nm in size, may reach well inside biomolecules, a situation not possible for larger particles. It has been reported that inhaled nanoparticles reach the blood and may reach other target sites such as the liver, heart or blood cells (Oberdörster G et al 2002, MacNee et al 2000, Kreyling et al 2002).
Nanoparticles may translocate through membranes. There is little evidence for an intact cellular or sub-cellular protection mechanism. For humans, inhalation is the most frequent route of access, and therefore the process of aggregation of the nanoparticles in the inhaled air has to be taken into account.
In order to understand and categorize the mechanisms for nanoparticle toxicity, information is needed on the response of living systems to the presence of nanoparticles of varying size, shape, surface and bulk chemical composition, as well as the temporal fate of the nanoparticles that are subject to translocation and degradation processes. The typical path within the organ and / or cell, which may be the result of either diffusion or active intracellular transportation, is also of relevance. Very little information on these aspects is presently available and this implies that there is an urgent need for toxicokinetic data for nanoparticles.
3.7.4 Solubility and Persistence
In view of the active functionalisation and the possible interaction of nanoparticles with bio-molecular structures, it is important to consider the dose and dose rate of the particulate agent, its ability to spread within the body and ecosystem, the decay of number concentration and the erosion of individual particles. Many nanoparticles will have considerable solubility. For these materials the interaction with living systems remains close enough to the bulk chemical agent to justify the use of well established toxicological testing procedures and approaches. For biodegradable particles, the particle composition and degradation products will influence their biological effects. On the other hand, materials with very low solubility or degradability, could accumulate within biological systems and persist there for long durations. It is with nanoparticles of this character that the greatest concerns must arise, and attention will have to be paid to the comparison of the persistence of the particles and the time constants of the metabolic and cellular activities within the target host.
The major emerging issue to be discussed in the context of the biological interactions of nanoparticles is related to those particles with little or no solubility, or being non- degradable at the locality where accumulation is observed. There remain many unknown details about the interaction of nanoparticles and biological systems.
Source & ©: SCENIHR
3.7 […] Nanoparticles and Living Systems, p. 20
6.2 Which characteristics of nanoparticles are relevant for health effects?
The SCENIHR opinion states:
3.8 Toxicology of Nanoparticles
Studies specifically dealing with the toxicity of nanoparticles have only appeared recently and, although now emerging in the literature, are still rare. Data concerning the behaviour and toxicity of particles mainly comes from studies on inhaled nanoparticles (reviewed by Oberdörster G 1996, Oberdörster G et al 2005, Donaldson and Stone 2003, Borm 2002, Donaldson et al 2001a, 2004, Dreher 2004, Kreyling et al. 2004). Data on the behaviour of particles is also available from pharmaceutical studies in which formulations involving nanoscale components are used to solve problems dealing with insolubility of drug formulations and for drug delivery (Baran et al 2002, Cascone et al 2002, Duncan 2003, Kipp 2004).
Not all toxicological studies to date deal with nanoparticles as recently defined (size <100 nm) or have characterised the nanoparticles according to recent knowledge. However, this does not necessarily interfere with the conclusions reached in these studies.
3.8.1 The Mediators of the Toxicity of Particles
Reduction in size to the nanoscale level results in an enormous increase of surface to volume ratio, so relatively more molecules of the chemical are present on the surface, thus enhancing the intrinsic toxicity (Donaldson et al 2004). This may be one of the reasons why nanoparticles are generally more toxic than larger particles of the same insoluble material when compared on a mass dose base. The expression of a dose response relationship on the basis of particle size resulted in a similar dose response relationship between low solubility - low toxicity, particles of different sizes (Oberdörster G et al 2000). In studies of low toxicity particles, TiO2 induced a more severe lung inflammation and particle lymph node burden compared to BaSO4 when dosed at mass burden in milligrams (Tran et al 2000). Surface area was therefore a driver for inflammation for these materials; the differences in severity of the response disappeared when the dose was expressed as surface area. These examples emphasize the importance of particle size, and by implication, the amount of surface area presented to the biological system for particle toxicity.
The chemical composition and the intrinsic toxicological properties of the chemical are of importance for the toxicity of particles (Donaldson et al 2004). The effect of carbon black has been shown to be more severe than that of titanium dioxide (Renwick et al 2004), while for both compounds the nanoparticles induced lung inflammation and epithelial damage in rats at greater extent than their larger counterparts. In addition, chemicals adsorbed on the surface may affect the reactivity of nanoparticles. Fractions isolated from particulate air pollutants (diesel exhaust particles) were demonstrated to exert toxic effects on cells in vitro (Xia et al 2004). Nanoparticles in ambient air can have a very complex composition, and these components, such as organics and metals, can interact. Metallic iron was able to potentiate the effect of carbon black nanoparticles, resulting in enhanced reactivity, including oxidative stress (Wilson et al 2002). In contrast, surface modification of nanoparticles can also result in a diminishing of cytotoxicity. The in vitro cytotoxicity of superparamagnetic iron oxide nanoparticles could be abrogated by coating the nanoparticles with pullulan (Gupta and Gupta 2005). Also for dextran and albumin derivatised iron oxide nanoparticles, a reduction in in vitro cytotoxicity was noted (Berry et al 2003).
For several different nanoscale particles (polyvinyl chloride, TiO2, SiO2, Co, Ni), only Co induced toxicity in endothelial cells, which was accompanied by the production of the pro-inflammatory cytokine IL8 (Peters et al 2004). For other particles only TiO2 and SiO2 induced minor and profound IL8 releases, respectively. An explanation of the differences in cytotoxicity was not presented but might be due to both material differences and/or size difference at the nanoscale, as the particle size ranged from a mean diameter of 14nm to 120 nm and even clusters of 420 nm (Peters et al 2004).
For micron sized biomaterial particles, the in vivo distribution was dependent on the composition of the material. With two polymers, polymethylmethacrylate (PMMA) particles but not polystyrene (PS) particles could be recovered from the spleen after intraperitoneal administration (Tomazic-Jezic et al 2001). The PS particles regardless of size were accumulated primarily in the adipose tissue of the peritoneal cavity, with very few particles in the spleen.
Although nanoparticles in air can be used as an information source for particle toxicity, one has to be aware that particles in ambient air as part of pollution of combustion origin are coated with all kinds of reactive chemicals including biological compounds such as endotoxin (Carty et al 2003, Kreyling et al 2004, Schins et al 2004). Thus the information obtained from ambient air particles for nanoparticle toxicity should take into account the possible influence of particle composition and contamination.
Shape is also likely to be an important factor although there is little definitive evidence. Fibres provide a significant example of the debate about shape, especially in relation to inhalation, where the physical parameters of thinness and length appear to determine respirability and inflammatory potential. The biopersistence of fibres effectively determines their dose.
A special category of fibres are nanotubes, which may be of a few nanometres in diameter but with a length that could be several micrometers. Risks should be assessed bearing in mind the well known carcinogenic effects of certain asbestos fibres. In two recently published in vivo studies, single-wall carbon nanotubes (SWCNTs) were demonstrated to induce lung granulomas after intratracheal administration (Lam 2004, Warheit et al 2004), indicating that these nanotubes cannot be classified as a new form of graphite on material safety data sheets. On a dose per mass basis the nanotubes were more toxic than quartz particles, well known for their lung toxicity, although the mass dose was very high and mechanical blockage of some airways was noted. Carbon black, carbonyl iron and graphite produced no significant adverse effects. Multifocal granulomatous lesions were observed without accompanying inflammation, cell proliferation or cytotoxicity, which was suggested to indicate a potentially new mechanism of pulmonary toxicity and injury by the nanotubes, not following the normal paradigm of toxic dusts (Warheit et al 2004). In vitro studies using a human keratinocyte cell line showed that carbon nanotube exposure resulted in accelerated oxidative stress and cellular toxicity, which may be interpreted as potential for dermal toxicity (Shvedova et al 2003).
Source & ©: SCENIHR
3.8 Toxicology of Nanoparticles, p. 22
6.3 How can inhaled nanoparticles affect health?
The SCENIHR opinion states:
3.8.2 Inhaled Particles
The role of particulate matter as a component of air pollution with an influence on human health is well established, although the mechanisms of action are poorly understood (Englert 2004). Ambient particulate air pollution was found to be statistically associated with cardiovascular morbidity and mortality (Pope 2000, Samet et al 2000, Peters et al 2001). However, very little is known on the relationship between the specific exposure to nanoparticles and health effects, in contrast with the large number of epidemiological studies on larger particles.
Von Klot et al (2002) could not distinguish between ambient fine particles and nanoparticles with respect to the association with increased asthma medication use. In another study, fine particles were more strongly related to cardio-respiratory symptoms than were nanoparticles (de Hartog et al 2003, Pekkanen et al 2002). Peters et al (1997) demonstrated that the number of nanoparticles is more strongly associated with health effects than the mass. Epidemiological studies on ambient air pollution do not provide consistent evidence that nanoparticles are more hazardous than larger particles. It may well be that epidemiological studies are not well suited to demonstrating differences between the toxicity of the various components of particulate matter. The exposure-dose relationships depend so much on time and location, and epidemiological studies are hampered by the lack of appropriate measurement. There is some evidence that combustion-derived particles emanating from traffic are a key driver for adverse health effects.
Estimating the dose of inhaled particles requires the knowledge of several mechanisms including regional deposition, retention, solubility, redistribution, translocation into the circulation, metabolism, accumulation in certain organs and the excretion pathways via urine and faeces. The factors that control or affect particle deposition include the particle characteristics themselves, the respiratory tract geometry and individual features of ventilation such as the mode of breathing.
Figure 1. Non-monotonic relationships of particle deposition as function of diameter in a healthy adult
Inhaled particulate matter can be deposited throughout the human respiratory system including pharyngeal, nasal, tracheobronchial and alveolar regions, depending on particle size as described in one model , shown in Figure 1, after Price et al (2002). A similar model has been proposed by ICRP (1994). The fractional deposition efficiency of particles with a size below 100 nm is between 30 and 70 % in pulmonary regions, although the predictability becomes less accurate at the nanoscale. With decreasing size there is a major increase in alveolar deposition. Cassee et al demonstrated that the toxicity of various size of soluble aerosolized cadmium chloride (CdCl2) could be accurately predicted by calculating the dose rather than using the exposure concentration (Cassee et al. 2002).
Figure 2. Diagram of known and suspended mechanical clearance pathways for insoluble particles depositing in the pulmonary region
The general pathways for the mechanical clearance of insoluble particles in the pulmonary region, after McClelland (1998) is given in Figure 2. After deposition in the respiratory tract, translocation of nanoparticles may potentially occur to the lung interstitium, the brain, liver, spleen and possibly to the foetus in pregnant females (MacNee et al 2000, Oberdörster G et al 2000, 2002). It is emphasised that there is extremely limited data available on these pathways. With insoluble iridium nanoparticles in the size range of about 15 and 80 nm, clearance was found to be primarily via the airways into the gastrointestinal tract. Only a small fraction (<1%) of the particles was translocated into secondary organs like liver, spleen, heart and brain, of which the 80 nm sized particles were translocated an order of magnitude less that the 15 nm sized particles, indicating the importance of size even within the nanometre range (Kreyling et al 2002). The presence of low amounts of particles in the liver and spleen could be attributed to translocation from the lung to the blood and sequestration by the sinusoidal macrophages of these organs. Particles were not dissolved nor absorbed from the gut (Kreyling et al 2002). However, depending on the exposure time, the actual amount of translocated particles could be considerable. Circulating particles (after intravenous administration) were accumulated in liver and spleen and retained there. Passage of inhaled nanoparticles into the bloodstream was demonstrated in one human study (Nemmar et al 2002), but two other similar studies have failed to show such a translocation. Another potential route of translocation of inhaled nanoparticles is the olfactory nerve in the nose leading to the olfactory bulb of the brain. 13C nanoparticles with a size about 35 nm were detected in the brain olfactory bulb after inhalation exposure. The route of brain entry was suggested to be by migration along the olfactory nerve into the olfactory bulb of the brain after deposition on the olfactory mucosa in the nasal region (Oberdörster E et al 2004).
As noted above, several model nanoparticulate materials have been used for the evaluation of air particulate toxicity. These include, polystyrene, titanium dioxide, carbon black, cobalt, nickel and latex (Oberdörster G et al 2000, Donaldson et al 2000, Dick et al 2003, Nygaard et al 2005). Nanoparticles of titanium dioxide induced more bronchoalveolar inflammation than fine TiO2 when rats were exposed to an equal concentration (Ferin et al 1992, Oberdörster G et al 1994, 2000). Similar results were obtained for nanoparticulate polystyrene (Brown et al 2001). These studies indicate that materials which by themselves are low in toxicity could be toxic when administered in formulations involving nanoparticles. If these particles are of low solubility, this effect could be solely due to the increased surface area of the inhaled dose. Similar results were obtained for low doses of nanoscale and fine carbon black and latex particles, (Li et al 1999, Donaldson et al 2000, 2001a, Wilson et al 2002, Renwick et al 2004). For nickel nanoparticles, an enhanced lung inflammation and toxicity was observed compared to larger sized nickel (Zhang et al 2003). Thus, for inhalation exposure it can be concluded that nanoparticles may show an increased toxicity compared to larger particles of the same chemical composition. Besides size, the chemical nature itself has an impact on the induced lung inflammation after intratracheal instillation, as nanoparticulate Ni was more toxic than nanoparticulate Co, with nanoparticulate TiO2 being the least toxic (Zhang et al 1998). The ranking of toxicity was reflected in the capability of the materials to induce free radical damage to plasmid DNA, indicating that free radical generation may underlie these observed differences in toxicity.
As mentioned above nanoparticles may be able to translocate from the lung into the blood resulting in systemic exposure of internal organs, although the extent of this may vary. Another route of translocation from the airways may be by neuronal uptake. Inflammatory biomarkers such as Interleukin 1α (IL1α) and Tumour Necrosis Factor α (TNFα) were increased in the brain of mice exposed to ambient air particulate matter compared to controls (Campbell et al 2005). It is unknown whether this leads to potentially adverse consequences, but certainly warrants further studies. In view of the induction of inflammatory cytokines, a relation with a variety of neurological diseases might be considered.
Vascular effects in terms of thrombosis were observed for intratracheally administered 60 nm amine modified polystyrene particles but not for 400 nm sized particles (Nemmar et al 2003). Both 60 nm and 400 nm sized polystyrene particles induced pulmonary inflammation, so inflammation and thrombogenesis are not necessarily coupled. In a recent study in rats carbon nanoparticles (about 38 nm) were found to induce a mild but consistent increase in heart rate (Harder et al 2005) but only induced a low grade pulmonary inflammation. The effect on the heart rate could not be related to blood hypercoagulability, which is in discrepancy with other reports (Donaldson et al 2001b, Nemmar et al 2003). Studies with inhalation of diesel soot in human subjects indicate that the function of the endothelial cells in the forearm is impaired following inhalation, which was evident from impaired vasomotor and secretory responses to pharmacological stimulation. The available data are consistent with the occurrence of a systemic inflammatory response and an alteration of autonomic cardiac control, but there is little evidence of endothelial dysfunction, pro-coagulatory states or nanoparticle-related myocardial malfunction.
Mechanisms of Particulate Toxicity
Several possible mechanisms of action for the toxicity of particles in general have been postulated (see figure 7), including injury of epithelial tissue (Pagan et al 2003), inflammation, oxidative stress response (Nel et al 2001, Donaldson et al 2001, Donaldson and Stone 2003), and allergy (Dybing et al 2004). At the cellular level oxidative stress is considered to be of importance (Donaldson et al 2001a,b, Oberdörster G et al 2005). Nanoparticle induced oxidative stress responses in keratinocytes, macrophages and blood monocytes after in vitro exposure (Shvedova et al 2003, Brown et al 2004). In a recent study gene expression profiles indicated that pulmonary injury and inflammation are likely due to increased expression of an oxidative stress response and subsequent contributions from cytokines and chemokines after exposure to urban particulate matter (Kooter et al 2005). Enhancement of antioxidant enzymes of alveolar macrophages was demonstrated after in vivo exposure to TiO2 nanoparticles, but this was not sufficient to counteract the lipid peroxidation and hydrogen peroxide generation that occurred (Afaq et al 1998). Thus, the overall resultant effect appears to be induction of oxidative stress in the cells, although the extrapolation of this mechanism to all types of nanoparticles is not possible.
Source & ©: SCENIHR
3.8.2 Inhaled Particles, p. 24
6.4 What are the health implications of nanoparticles used as drug carriers?
The SCENIHR opinion states:
3.8.3 Particles for Drug Delivery
Carriers for Drug Delivery
Nanostructures and nanoparticles can be used for drug delivery purposes, either as the drug formulation itself or as the drug delivery carrier (Cascone et al 2002, Baran et al 2002, Duncan 2003, Kipp 2004). Current research focuses on cancer therapy, diagnostics and imaging, although many challenges still need to be solved (Ferrari 2005). In addition, nanostructures are being investigated for gene delivery purposes (Kneuer et al 2000, Salem et al 2003, Ravi Kumar et al 2004, Gemeinhart et al 2005, Yoo et al 2005, Roy et al 2005).
Many different formulations involving nanoparticles have been used for drug delivery purposes, including albumin (Damascelli et al 2003), poly(D,L-lactic-co-glycolide)acid (PLGA) (Panyam et al 2002, Weissenbock et al 2004), solid lipid formulations (Muller et al 1997, 2000, Wissing et al 2004), cetyl alcohol/polysorbate nanoparticles (Koziara et al 2004), hydrogels (Gupta and Gupta 2004), gold (Hainfeld et al 2004, Paciotti et al 2004), polyalkylcyanoacrylate composites (Cruz et al 1997, Olivier et al 1999, Kreuter et al 2003), magnetic iron oxide (Gupta and Gupta 2005), methoxy poly(ethylene glycol)/poly(ε-caprolactone) (Kim et al 2003), and gelatin (Cascone et al 2002). Albumin nanoparticles are already the subject of clinical studies for anticancer drug delivery purposes (Damascelli et al 2003).
Not all of the ‘nanoparticle formulations’ mentioned are strictly or solely nanoparticulate in the sense that their size is not always below 100 nm and some drug delivery systems include particles up to several hundreds of nanometres. In many cases, the technology to produce very small particles did not exist in the early stages of development, but now there is an increasing refinement in their size and it is relevant to this Opinion to discuss this group of products together. The route of administration may be oral, parental (subcutaneous, intramuscular, intra-arterial, intravenous) and via the skin.
The aims for nanoparticle entrapment of drugs are either enhanced delivery and uptake by cells and/or the reduction in toxicity of the free drug to non target organs. For intravenous administration, long circulating and target- specific nanoparticles are needed. One of the problems is evasion of the entrapment of nanoparticles in the mononuclear phagocytic system, as present in liver and spleen (Gibaud et al 1996, Moghimi et al 2001). Surface modification with poly(ethyleneglycol), (PEG) resulted in a prolonged presence in the circulation by avoiding recognition and phagocytosis by the mononuclear phagocytic system (Bazile et al 1995). Besides reduction of therapeutic efficacy, liver entrapment also may have an effect on liver function. For cyanoacrylate and polystyrene nanoparticles (around 214 nm and 128 nm, respectively) transient liver alterations were observed after a single and chronic intravenous administration (Fernandez-Urrusuno et al 1995, 1997). Inflammatory responses were characterized by secretion of acute phase protein α1-acid glycoprotein by hepatocytes. In addition, antioxidant defences of hepatocytes were depleted, probably as a result of local release of oxidative species. Although nanoscale formulation is aimed at enhancing drug delivery without loss of drug activity, a study comparing insulin-chitosan nanoparticles to chitosan solution and chitosan powder formulations showed that the insulin-chitosan nanoparticles were less effective in terms of bioavailability and lowering blood glucose level in both a rat and sheep model (Dyer et al 2002).
After oral administration, only 10% of 60 nm polystyrene particles were recovered from the tissue of the gastrointestinal tract. Most of these particles were present in lymphoid tissue such as Peyer’s Patches and lymphoid aggregates in the large intestine (Hillery et al 1994).
After dermal administration, negatively charged nanoparticles of about 50 and 500 nm were found to permeate the skin, while positively charged and neutral particles of all sizes did not. It was suggested that particle size was less important than the total charge, explaining why both 50 and 500 nm sized latex particles showed permeation and 100 or 200 nm negatively charged particles did not (Kohli and Alpar 2004). A greater concentration of charge was suggested to be responsible for overcoming the skin barrier, explained for the 50 nm particles as being the small size and large surface area, and for the 500 nm explained by the high number of charged groups. It can be expected that penetration of damaged skin will be easier. Thus, when nanoparticles are used in ointments there should be particular consideration of skin permeation.
Encapsulation in sub 130 nm size poly(lactic acid – glycolic acid), PLGA, particles increased cellular uptake of a photosensitizer, resulting in enhanced cytotoxicity in vitro (Konan et al 2003). Toxicity of free nanoparticles was not determined in this study. Chemical characteristics such as surface charge may determine the fate of nanoparticles in cells. PLGA nanoparticles were found to be ingested by cells by endocytosis (Panyam et al 2002, Konan et al 2003). The escape from these endosomes into the cellular cytoplasm was suggested to be caused by a change in surface charge from negative to positive, resulting in cytoplasmic delivery of the incorporated drug. The hypothesis concerning the influence of the positive surface charge for escaping the endosomes was supported by data obtained with negatively charged polystyrene nanoparticles which remained in the endosomal compartment of the smooth muscle cells used in this study.
Nanoparticles may be used for gene delivery, applications including plasmid DNA administration for vaccination (Salem et al 2003, Cui and Mumper 2002, 2003, Zhang et al 2005) and cancer therapy (Ramesh et al 2004, Gordon and Hall 2005). Gene transfer was accomplished in vitro and in vivo using various types of nanoparticles. With silica nanoparticles of about 42 nm, gene transfer was obtained with very low cell toxicity (Ravi Kumar et al 2004). A clinical trial with gene therapy aimed at determining safety and tolerability was performed in cystic fibrosis patients (Konstan et al 2004).
Specific targeting to retinal epithelium cells in the eye is possible (Bourges et al 2003). For very small quantum dots (<10 nm) specific targeting of peptide coated quantum dots to the vasculature of lungs and tumours has been reported (Åkerman et al 2002). PEG coating abrogated uptake by the reticuloendothelial system of liver and spleen. In contrast, about 40-50 nm magnetic nanoparticles coated with PEG were quite well taken up by endocytosis (Gupta and Curtis 2004)
For indomethacin loaded nanospheres (size below 200 nm) composed of methoxy poly(ethylene glycol)/poly(ε-caprolactone) polymers, the in vitro cytotoxicity was reduced when compared to free indomethacin, although some minor toxicity of 15-20% growth reduction was still present (Kim et al 2003). In vivo acute toxicity studies found a LD50 value of 1.47 g.kg-1 , and 50% of this LD50 value administered for 7 days did not induce acute toxicity in heart, lung, liver and kidney. It was concluded that these methoxy poly(ethylene glycol)/poly(ε-caprolactone) polymer nanospheres were non- toxic.
Surface modifications of nanoparticles offer possibilities for medical applications such as drug targeting in terms of cellular adhesion and invasion and transcellular transport. Carbohydrate binding ligands on the surface of biodegradable PLGA nanospheres were found to associate at higher rates with cell membranes (Weissenböck et al 2004). Such increased adherence may lead to an enhanced activity of the drug presented as or incorporated in nanoparticles. For solid lipid nanoparticles, (SLN), in vitro cytotoxicity was dependent on the surfactant used for stabilization of the nanoparticles with one of the investigated surfactants inducing cytotoxicity (Muller et al 1997, Olbrich et al 2004). The stabilizing surfactants showed the largest differences in toxicity, although toxicity could be markedly reduced by binding to the nanoparticles. SLNs of various composition were investigated for their use in skin application (Santos Maia et al 2002). In vitro studies showed an increased drug (glucocorticoid) penetration of skin and epidermal localization. For nanoparticles in the size of 200 – 400 nm, the composition of the lipid matrix was shown to have an impact on the cytotoxicity of SLN (Schöler et al 2002). Coupling specific proteins such as antibodies to the nanoparticle surface may enable a more specific immune directed targeting of the particles to certain cells or organs (Nobs et al 2004).
Organ Specific Targeting
One of the advantages of the use of nanoparticles for pharmaceutical formulations is the potential to cross the blood brain barrier (BBB). However, this also may be the major drawback for systemic administration of nanoparticles in terms of potential brain toxicity. Such passage was suggested to be possible by the toxic effect of nanoparticles (about 200nm) on cerebral endothelial cells (Olivier et al 1999), although for similar nanoparticles (about 300nm) this was contradicted and not found for a different type of nanoparticle (Lockman et al 2003). Physical association of the drug to the nanoparticles was necessary for drug delivery to occur into the brain (Kreuter et al 2003). When nanoparticles with different surface characteristics were evaluated, neutral nanoparticles and low concentrations of anionic nanoparticles were found to have no effect on BBB integrity, whereas high concentrations of anionic nanoparticles and cationic nanoparticles were toxic for the BBB. The brain uptake rates of anionic nanoparticles at lower concentrations were superior to neutral or cationic formulations at the same concentrations. Therefore, nanoparticle surface charge must be considered for toxicity and brain distribution profiles (Lockman et al 2004).
Specific migration into draining lymph node is of importance for both treatment and diagnostic purposes. Nanoparticle formulations of polyisobutylcyanoacrylate (Nishioka and Yoshino 2001) and fluorescent quantum dots (Kim et al 2004, Soltesz et al 2005) were shown to localize into such draining lymph nodes. Also uptake by endothelial cells can be used for diagnostic as well as therapeutic (prevention of cardiovascular restenosis) purposes (Davda and Labhasetwa 2002, Uwatoka et al 2003, Westedt et al 2004).
The use of nanoparticles as drug carriers may reduce the toxicity of the incorporated drug (Kim et al 2003), although discrimination between the drug and the nanoparticle toxicity cannot always be made. The structure and properties of gold nanoparticles make them useful for a wide array of biological applications. Toxicity, however, has been observed at high concentrations using these systems. Goodman et al (2004) demonstrated that for 2 nm gold particles cationic particles were moderately toxic, whereas anionic particles were relatively non-toxic. Such very small sized gold nanoparticles were found to be non toxic when administered to mice for tumour therapy (Hainfeld et al 2004).
For thiol derivatized PEG - colloidal gold nanoparticles with tumour necrosis factor (TNF) an enhanced anti-tumour activity was reported when compared to free TNF (Paciotti et al 2004). Topoisomerase inhibitors when formulated in lipid containing nanoparticles showed increased anti-tumour activity in an in vivo nude mouse xenograft human tumour model (Williams et al 2003). Although phagocytosis by macrophages does not seem to be necessary for the uptake of nanoparticles, the immune system is not totally inactive when dealing with nanoparticles. For 100 nm polystyrene particles, an IgE adjuvant activity was observed in an animal model system of ovalbumin allergy (Nygaard et al 2005). Antibodies against fullerenes could be induced after the intraperitoneal injection of C60 conjugated with serum proteins.
Source & ©: SCENIHR
3.8.3 Particles for Drug Delivery, p. 28
6.5 How should harmful effects of nanoparticles be assessed?
The SCENIHR opinion states:
How should harmful effects of nanoparticles be assessed?
3.8.4 Toxicological Testing
The consideration of dose response relationships in the toxicology of nanoparticles poses a significant problem. In toxicology the paradigm exists that health effects are correlated to the mass of the agent to which the individual is exposed, resulting in an accumulated mass as internal or organ dose/exposure. For nanoparticles the concentration number and the resulting total surface area determine the interactions with biological systems. Therefore the surface area and number concentration appear to be more reasonable parameters for doses in terms of exposure (Brown et al 2001, Oberdörster G et al 2000, Höhr et al 2002). The increase in lung inflammation for nanoparticles compared to fine particles was noted when doses were expressed as mass. In contrast when doses were expressed as surface area similar responses were observed for both fine and nanoparticles (Oberdörster G et al 2000). Also when comparing toxicity differences between TiO2 and BaSO4, dose response relationships were similar when compared at a dose expressed as surface area burden (Tran et al 2000). For TiO2 nanoparticles, significant species differences were noted after inhalation exposure with rats, mice and hamsters, the rat being the most sensitive (Bermudez et al 2004). Pulmonary responses and dosimetry (particle retention and overload) were considered to be responsible for these differences. Several hypotheses were proposed for the adverse health effects of nanoparticles as part of ambient air pollution (reviewed by Kreyling et al 2004). These hypotheses for adverse health effects of nanoparticles include:
- Importance of large surface area for interactions with cells and tissues
- Complex formation with biomolecules
- Formation of increased level of radical species compared to larger particles
- Increased induction of oxidative stress
- Induction of cellular DNA damage
- Induction of oxidative stress by lipid peroxidation
- Deposition characteristics dependent on size
- Uptake by cells of respiratory epithelium
- Increased access to interstitial spaces
- Access to systemic circulation
- Reduced function of macrophages, reduced phagocytosis of particles themselves, reduced macrophage mobility and cytoskeletal dysfunction
- Increased pro-inflammatory activity and induction of cytokines and other mediators
- Adverse effects on cardiac functions and vascular homeostasis
Some hypotheses raised for ambient air nanoparticles may be of limited or no relevance for engineered nanoparticles, such as adsorbance of toxic substances. Although such adsorbance cannot be ruled out, it is probably of less importance for production and handling facilities of large volumes of engineered nanoparticles compared to the particles in ambient air. Limitations of the studies cited may be the relatively high doses used, the short periods of time investigated, and/or artefacts occurring during sampling of the particles on filters (Wittmaack et al 2002).
In addition, the use of healthy animal models may hamper the interpretation of the results as some of the effects listed and may only be a risk for susceptible organisms and predisposed individuals, but not to healthy people (Kreyling et al 2004). Age, co- pollutants and a compromised respiratory tract can modify the pulmonary inflammation and oxidative stress induced by carbonaceous nanoparticles (Elder et al 2000). For chronic obstructive pulmonary disease, the generation of free radicals on the surface due to high reactivity of nanoparticles, and the induction of oxidative stress, might contribute to the induction of inflammation (MacNee et al 2003).
In vitro observations with keratinocytes, macrophages and blood monocytes revealed the induction of oxidative stress in these cells after exposure to nanoparticles (Shvedova et al 2003, Brown et al 2004). A role for free radicals and reactive oxygen species was also suggested by in vitro studies in which antioxidants were able to block the particle induced release of TNFα from alveolar macrophages (Dick et al 2003). Also macrophage phagocytosis was impaired by nanoparticles (Renwick et al 2001). For TiO2 and ZnO nanoparticles, oxidative damage to DNA was demonstrated (Dunford et al 1997, Rahman et al 2002), resulting in micronucleus formation and apoptosis.
The type of cell under investigation may also be of importance for the ultimate effect of the particles investigated. For epithelial lung cells, either as cell line or primary rat type 2 cells, the coarse fraction of urban ambient air showed similar or higher potency to induce cytokine release and cytotoxicity compared to the finer fractions (Hetland et al 2004). Human macrophages and osteoblasts showed a different behaviour towards nanotopography surfaces, macrophages showing preference for the nanosurface and being activated, while the osteoblasts moved away from the nanosurfaces (Rice et al 2003).
In view of the specific characteristics demonstrated for nanoparticles and nanoparticle formulations, the assays usually performed for determining toxicity of products may not be sufficient to detect all possible adverse effects of nanoparticles. However, this may not be the case for assessing potential environmental effects given the nature and the simplicity of standard, regulatory tests. Nanoparticles may differ in reactivity and solubility and may interact with all kinds of endogenous proteins, lipids, polysaccharides and cells. Based on experiences in inhalation toxicology, a series of tests was proposed for evaluation of the toxicity of nanoparticles used in drug delivery systems (Borm and Kreyling 2004). These included tests for blood cell damage after intravenous administration, acute phase responses of hepatocytes or lung cells, permeability tests of endothelial cells, for destabilization of atheromatous plaques in animal models for atherosclerosis, for effects on the autonomic nervous system, for adjuvant activity in an immunization model, for immune activation by measuring T cell activity and cytokine induction in lymph nodes, for the determination of surface activity and induction of oxidative stress in cell lines, for toxicity on various cell lines in vitro, and for biopersistance.
Reduction in size to the nanoscale changes the characteristics of particles, primarily due to the increased surface to volume ratio. There are as yet no paradigms to anticipate the significance of any of these changes in characteristics, so the safety evaluation of nanoparticles and nanostructures cannot rely on the toxicological and ecotoxicological profile of the bulk material that has been historically determined. The biological behaviour of nanoparticles is determined by the chemical composition, including coatings on the surface, the decrease in size and corresponding shifts in chemical and physical properties, the associated increase in surface to volume ratio, and the shape. In addition, aggregations of nanoparticles may have an effect on their biological behaviour as well. The dose expressed as surface area or number of particles administered shows a better relationship with biological and/or toxic effects than dose expressed as mass. The biological evaluation of nanoparticles and/or products incorporating nanoparticles should be performed on a case by case basis.
Epidemiological studies on ambient air pollution demonstrate the general adverse effects of particulate matter on humans. However, chemical absorbents on the particulates themselves can be partly responsible for modifying the toxic effects, which limits extrapolation of these results to nanoparticle toxicity.
One mechanism of toxicity of nanoparticles is likely to be induction of reactive oxygen species and the consequential oxidative stress in cells and organs. Testing for interaction of nanoparticles with proteins and various cell types should be considered as part of the toxicological evaluation. Nanoparticle translocation and uptake by the body occurs after inhalation exposure (neuronal uptake, translocation across lung epithelium, and ingestion), oral exposure (ingestion), and dermal exposure depending on the characteristics of the nanoparticle under investigation. With the exception of airborne particles delivered to the lung, information on the biological fate of nanoparticles including distribution, accumulation, metabolism, and organ specific toxicity is still minimal.
Source & ©: SCENIHR
3.8.4 Toxicological Testing, p. 32
6.6 What are the effects of nanoparticles on the environment?
The SCENIHR opinion states:
Colvin’s (2003) discussion on the potential impact of engineered materials demonstrates the lack of data on the exposure and effects of nanoparticles. To date, only a few studies have been carried out with species used for ecotoxicological testing. Oberdörster (2004b) showed the 48 hours LC50 in Daphnia magna for uncoated water soluble fullerenes nC60 is 800 ppb. E. Oberdörster (2004a) demonstrated a significant increase of lipid peroxidation in the brain and glutathione depletion in the gill of juvenile largemouth bass (Micropterus salmoides) after exposure for 48 hours to 0.5 ppm of fullerenes nC60., but the increase was not significant at 1 ppm.
In their follow-up studies, Oberdörster G et al. (2005) report the possible molecular mechanism of these observations. The bactericidal properties of fullerenes have been reported by Yamakoshi et al (2003). However, considering that a large number of the above cited human toxicology studies have examined the uptake and effects of nanoparticles at a cellular level, it can be hypothesized that these observations may also hold for species other than humans. As such the reports may be useful for the assessment of the effects on environmental species. Work to support this hypothesis is needed. Careful examination and interpretation of existing data and careful planning of new research is, however, required if we are to establish the true ecotoxicity of nanoparticles, and the differences with conventional forms of the substances.
Source & ©: SCENIHR
3.8.3 Particles for Drug Delivery, p. 31
Other Environmental Species
There is almost no published literature on the effects of nanoparticles on environmental species. Colvin’s (2003) discussion on the potential impact of engineered materials demonstrates the lack of data on the exposure and effects of nanoparticles. To date, only a few studies have been carried out with species used for ecotoxicological testing. Oberdörster (2004b) showed the 48 hours LC50 in Daphnia magna for uncoated water soluble fullerenes nC60 is 800 ppb. Oberdörster G (2004a) demonstrated a significant increase of lipid peroxidation in the brain and glutathione depletion in the gill of juvenile largemouth bass (Micropterus salmoides) after exposure for 48 hours to 0.5 ppm of fullerenes nC60. , but the increase was not significant at 1 ppm. In their follow-up studies, Oberdörster E et al. (2005) report the possible molecular mechanism of these observations. The bactericidal properties of fullerenes have been reported by Kai et al (2003). However, a number of the above cited human toxicology studies have examined the uptake and effects of nanoparticles at a cellular level, it is reasonable to assume that these observations can be extrapolated to environmental species. Work to support this hypothesis is needed. Careful examination and interpretation of existing data and careful planning of new research is required to establish the true ecotoxicity of nanoparticles and the differences with conventional forms of the substances.
Because of the inverse relationship between particle size and surface area, it is imperative that, for various environmental (model) species, (1) dose (or concentration) – effect relationships are established as a function of total surface area and/or number of particles (and surface charge) rather than mass units and (2) a comparison is made between the effects of the conventional and the nanoparticle form(s) of the substance. It should also be recognized that the potential problems associated with persistent insoluble nanoparticles in the environment may be considerably greater than with human health assessment. The protection goals and endpoints (i.e. protection of individuals vs. protection of populations) of an environmental effect assessment are clearly different than those used of the human health evaluation. As such, next to in vitro studies which may help to establish potential differences in the toxic action of nanoparticles and conventional forms of the substance, in vivo assays will have to be performed using model species representative of each of environmental compartment (terrestrial, aquatic etc.) and reflecting different exposure routes ( water, food-borne, etc.).
Source & ©: SCENIHR
3.10.3 Hazard Considerations, p. 45