17 Approaches to the Toxicological Testing of Particles KenDonaldson MRC/University of Edinburgh Centre for Inflammation Research, Queen’sMedical ResearchInstitute SteveFaux MRC/University of Edinburgh Centre for Inflammation Research, Queen’sMedical ResearchInstitute Paul J. A. Borm Centre of ExpertiseinLife Sciences (CEL),Hogeschool Zuyd VickiStone School of Life Sciences, Napier University CONTENTS 17.1 Background to Testing of Particles 300 17.2 Factors Affecting Particle Toxicity 300 17.2.1 Particle Size 300 17.2.1.1 Particle Shape 301 17.2.1.2 Particle-Derived Transition Metals 302 17.2.1.3 Particle-Derived Free Radicals 303 17.2.1.4 Particle Biopersistence 303 17.3 Approaches to Testing 304 17.3.1 Characterizing the Particles 304 17.4 Assessment of Toxicity In Vitro 304 17.4.1 Cytotoxicity 305 17.4.2 Cell Stimulation 305 17.4.3 Cell Proliferation 305 17.4.4 Cytokine Measurement 306 17.4.5 Oxidative Stress Measurement 306 17.4.6 Alterations in Cell Signaling Cascades 306 17.4.6.1 Intracellular Calcium 306 17.4.6.2 Mitogen-Activated Protein Kinase Cascade 307 17.4.6.3 Transcription Factor Activation 307 17.4.7 Effects on Blood 308 17.4.7.1 Endothelium 308 17.4.7.2 Platelets 308 17.4.7.3 Atherosclerotic Plaques 308 17.4.8 Effects on the Brain 308 17.4.9 In Vitro Tests for Genotoxicity 309 299 © 2007 by Taylor & Francis Group, LLC 17.5 Animal Studies 309 17.5.1 Bronchoalveolar Lavage 309 17.5.2 Intratracheal Instillation 309 17.5.3 Inhalation Studies 309 17.5.4 Monitoring Exposure, Dose, and Response 310 17.5.5 ImmunologicalEffects of Particles 310 17.5.6 Effects on the Microbicidal Activity of the Lungs 310 17.5.7 Animal ModelstoStudy the CardiovascularEffects of Particles 311 17.5.8 Animal ModelsofNeurological Effects of Particles 311 17.5.9 Conclusion—A Tiered ApproachtoTesting of aNew Particle 311 References 312 17.1 BACKGROUND TO TESTING OF PARTICLES There is awide spectrumofdifferent approaches to the testing of particles that range from long- term inhalation studies,with the final endpoint of cancer, to short-term tests in vitro determining the ability of the particle to modulate cellular functions. The valueofeach of thesetests in predicting pathogenicity varies; but in general, there is aplayoffbetweenthe extended timescale and high cost of long-term pathogenicity experiments in animals which can be used in risk assessment and the short timescale and relative inexpensiveness of in vitro data,which can, at best,beused in hazard assessment. If epidemiological or clinical data were availableonthe pathogenicoutcome of exposure to anovel particle,then this would form the basis of arational test strategy. However, in the absence of such information, astrategy based on knowledge of structure, chemistry,and shape of the particle could allow benchmarking to similar particles of knownpathogenicity in order to decide on the mostappropriate endpoint. Thisspectrumofpossiblepathology arisingfromparticle exposure posesanimmediate problemfor atesting strategy. Astrategydesignedtodetect acarcinogenicendpoint, for example, wouldbeentirely different from onethat wouldbechosentodetect the potential to cause asthma.Sothere is aneed for some knowledge of the likelypathology that would arise. This couldbeprovided by: 1. Apriori knowledge from clinical observations or epidemiological studies indicating that aparticular disease or manifestationoftoxicityisassociated with exposureto the particle. 2. In the absence of such information, there could be benchmarking to particles of known toxicity. For example, if the particle type was amineral and contained some quartz, then the endpointsoffibrosis and cancer could be selected. If the particle was organic or contained heavy metals, then sensitization might be considered. Fibrous particleswould be suspected of causing mesothelioma, etc. 17.2 FACTORS AFFECTING PARTICLE TOXICITY 17.2.1 P ARTICLE S IZE Given the different diseases caused by particles, we assume that there are different target cells and tissues(e.g., airways, alveolar cells, immune cells) that are affected by different particles. This can be understood from the point of view of differencesindose to these different target cells caused by Particle Toxicology300 © 2007 by Taylor & Francis Group, LLC differencesinthe distributionofdosethroughout therespiratory tractbecause of variationin particle size. The fractionaldeposition of various sized particles for different pulmonary compartmentsis dictated by the aerodynamic diameter, D ae ,asshownTable 17.1.This is defined as the diameter of a particle of unit density with the same falling speed as the particle of interest. Different-sized particles would be expected to elicit different typesofpathogenicresponse because they depositindifferent regions, as shown in Table 17.2. From the foregoing, it can be concluded that the characterization of size of anovel particle is important in decidingwhat is the most likely endpoint to examine, since sizewill bear directly on the site of deposition and the subsequent response. When the site of the adverse health effect is not the locallung environment,itismore difficult to know which site of deposition is mostimportant because therecould be, in theory at least, translocation from any site of deposition—hence the questionmarks over the site of deposition of particles that are associated with mesothelioma and strokes/heart attacks. In the case of nanoparticles (see later) additional targets, such as the blood and brain may be relevant, as aconsequence of translocation. 17.2.1.1 ParticleShape Particles comeinanumber of different shapes, e.g., compact, platy, and fibrous.The importanceof particle shapeisbest understoodfor fibers, and fiber length is known to be amajor factorin pathogenicity as reviewed in Donaldson and Tran (2004).Long fibers of amosite asbestos were much more pathogenicthan asample of the same material milled so that the fiber length was drastically shortened, with much of the sample being so short that it was classified as non-fibrous (Davis et al. 1986). Forother particles, shape is less obvious as aparameter that mediates toxicity. TABLE 17.2 TheAdverse Effects Resulting from Different Sizes of Particles Depositing in Different Compartments Deposition Site Disease Thoracic COPD, asthma, central lung cancer Respirable Small airways disease, peripherallung cancer, emphysema, interstitial fibrosis, cardiovascular effects TABLE 17.1 Anatomical Site of Deposition for the Different Deposition Fractions DepositionFraction Approx Aerodynamic Diameter ( m m) Definition Site of Deposition Inhalable w 50–100 The fraction inhaled through the nose and mouth Mouth, larynx, pharynx Thoracic w 10–50 Fraction penetrating beyond the larynx The above plus airways Respirable ! w 10 Fraction penetrating beyond the ciliated airways The above plus terminal bronchioles and alveolar ducts Approaches to the Toxicological Testing of Particles 301 © 2007 by Taylor & Francis Group, LLC Shape may not be the only factor that dictates fiber pathogenicity—even amongst long and thin fibers thereare differencesinpathogenicity,especiallyinmesotheliomaproduction.Erionite (Maltoni, Minardi, and Morisi 1982)and silicon carbide fibers (Davis et al. 1996), for example were much more active in causing mesothelioma following inhalation exposure than would be expected from their dimensions. For this reason it is likely that another factor,surface reactivity (see below), could be important in mediating some of their pathogenicity. 17.2.1.2 Particle-Derived Transition Metals It has become evident that transition metals are key players in the pro-inflammatory effects of a range of particle types (Ghio et al. 2006). Iron is especially important because it has the ability to generate free radicals via Fenton chemistry that is well characterized. The state of the ironisall important;but in particular, the amount of Fe(II) is central since this is the directly harmfulspecies (Ghio and Cohen 2005). Consequently, total ironisnot necessarily informative as to the biologi- callyactive iron. The ironmustredox-cycle to be capable of causing majorinjuryto macromolecules. This is accomplished by areluctant in the region of the particle, e.g., glutathione, ascorbate, NADH, or even superoxide anion. This means that the presence of anti-oxidants in the lungs is a“double-edged sword.” By the sequence of eventsshown in Figure17.1, the highly toxic and reactive hydroxyl radical can be formed.The hydroxyl radical may be involved in diffusion-limited reactions which lead to the formation of various carbon-centered radicals, peroxyl, alkoxyl, and thiyl radicals, all of which have harmful consequences for cells. Becausedifferent reductants couldhave different potencies in causing reduction of Fe(III), and because of the knownrole of chelating agents, the micro-environ- mentofthe lung where the particle is present could be all-important in determining howmuch reactive iron is present at the surface. In addition, the particle may accumulate biological iron, which can also have freeradical-generating activity(Ghio, Jaskot, and Hatch 1994). In various models, the biological effects of several different types of ambient particle, including PM 10 and ROFA,have been suggested to be driven by their transition metal content (Dreher et al. 1997; Jimenezetal. 2000; Rice et al. 2001; Dai, Xie, and Churg 2002; Molinelli et al. 2002; McNeilly et al. in press).The measurement of the sum potential for aparticle to generate oxidative stress has been advanced as akey parameter that might dictateoverall toxicity. Particle Transition metal Reducing environment of the lungs Altered redox status Inflammation Direct tissue damage Redox cycling O 2 − and OH − FIGURE 17.1 Generation of oxidative stress by transition metals. Particle Toxicology302 © 2007 by Taylor & Francis Group, LLC 17.2.1.3 Particle-Derived Free Radicals Quartzisone of the most toxic particles and it is knowntohave ahighly reactive surface. The quartz surface can generate reactive oxygen species (ROS) in several ways following interactions of the quartzparticles with pulmonary cells or lungfluids (Castranova, Dalal,and Vallyathan 1995). PM 10 hasbeen demonstrated to generate freeradicalsand cause oxidative stress (Ghio et al. 1996; Gilmour et al. 1996; Squadrito et al. 2001; Austetal. 2002)via transition metals andorganics that redox cycle. Nanoparticlesare also capable of generating ROS in cell-free systems(Wilson et al. 2002). In fact, almost all pathogenicparticlesstudied, including asbestos (Lund and Aust 1991), glassfibers (Gilmour et al. 1997), and coalmine dust (Dalal et al. 1995)are capable of generating ROS in cellfree systems. Oxidativestress has been suggested as ageneric mechanism for the action of particles (Donaldson, Beswick,and Gilmour 1996), and more recently for nanoparticles (Donaldson et al. 2005; Oberdo ¨ rster, Oberdo ¨ rster, and Oberdo ¨ rster 2005). Many systemsare available formeasuring thefee-radical-generating potential of aparticle sample, includingpurely chemical methods such as HPLC (Brown, Fisher, and Donaldson 1998), the use of super-coiled plasmid DNA as asensoroffreeradicals (Gilmour et al. 1997), the use of EPR and spin traps to capture theseshort-lived moieties(Shi et al. 2003), and calfthymus DNA to detect 8-hydroxydeoxyguanosine (8-OHdG). 17.2.1.4 ParticleBiopersistence Biopersistence is the capacity of particlestopersist in the lungs.Biopersistence is limited by the potential of particles to dissolveorlose elements,break, or be mechanically clearedfrom the lungs by macrophages. Thepotential for aparticle to dissolveinthe lungs would seem intuitively to be an important factor,since the dose of particle would not be expected to build up in the case of soluble particles. However, little is know about the biochemical conditions that pertain in the lungs—the lung is largelya“black box” in this regard, althoughdifferences in pH and the impact of coating of the particles is to be anticipated. The best case where this property is seen as being important is with fibers. Long fibers are not well clearedfrom the respiratory region of the lungs (Coin, Roggli, and Brody 1994; Searl et al. 1999), presumably because of the difficulties of the alveolar macrophage to successfully phagocy- tose and then move with them to the mucociliary escalator. Thus the ability of long fibers to persist in the lungs without being either dissolved away or weakenedsothat they break into smaller fibers which can be easily clearedisseen as an important factor contributing to pathogenicity (Hesterberg et al. 1994). Forany particle, its ability to biopersist will be an important factor in modifying its pathogenicity. TABLE 17.3 Important ParticleCharacteristics to Be Ascertained in Samples of Particles of Unknown Toxicity Physical Characteristics Chemical Characteristics Dimensions Transition metals Surface area/unit mass Quartz or cristobalite content Biopersistence Heavy metals Free radical activity PAH Durability Endotoxin Approaches to the Toxicological Testing of Particles 303 © 2007 by Taylor & Francis Group, LLC 17.3 APPROACHES TO TESTING 17.3.1 C HARACTERIZING THE P ARTICLES An understanding of the nature of the test particle with regard to shape, size, elemental compo- sition, transition metal content, endotoxincontamination, etc., is vital to the testing strategy since it will allow the particle to be benchmarked. There are anumber of parameters that could be assessed. Once again, theconcept of benchmarking is ausefulone andthe source of theparticle, e.g., mineral-derived, man-made fiber, ash, etc., can be used to decide which is the mostlikelyparameter that should be determined. Table 17.3 showssomeparticle characteristicsthat can be quantified, which might shed light on its likely pathogenicity. Endotoxin is apotentialconfounder in allstudies with particlesand itspresenceshouldberigor- ously monitoredsince it mayexplain allthe toxicity of adustsample(Brown and Donaldson 1996). Endotoxincan be measuredbyspecific ELISAorfunctionally usingthe amoebocytelysateassay. 17.4 ASSESSMENT OF TOXICITY IN VITRO The European Centre for the Validation of Alternative Methods (ECVAM)publishedareport on “Nonanimal tests for evaluatingthe toxicity of solidxenobiotics,” and this contains recommen- dations for in vitro testswithparticles(Fubini et al. 1998). Themajority of in vitro testsfor detecting particle toxicity are aimed at detecting either direct or indirect pro-inflammatory effects. Acute and chronic inflammation is thought to be central to the etiology of manylung disorders, such as asthma and chronic obstructive pulmonary disease (COPD).The specificcharacteristics of the inflammatory response may be different,but all are characterized by the recruitment of inflammatory cells into the lung. These activated cells, such as alveolar macrophages and neutrophils, produce cytokines and ROS and many other mediatorsinvolved in inflammation. Oncetriggered, the inflam- matoryresponse will persist in these conditions leading to lung injury. The intracellular mechanism in the lung epithelium and the macrophages leading to lung injury in response to environmental particulates will involve the activation and upregulation of transcription factors, such as activator protein-1(AP-1)and nuclear factor- k B(NF- k B),leading to increased gene expression andthe biosynthesis of proinflammatorymediators. Particles may stimulate inflammation by anumber of pathways such as: 1. Cell death, apotentstimulus for inflammation 2. Nonspecific stimulation of cell receptors 3. Oxidativestress 4. Calcium flux 5. Via the immune system (see Figure17.2) Particles Cell necrosis Calcium flux Oxidative stress Non-specific stimulation of receptors Pro-inflammatory gene expression Inflammation Immune system Hypersensitivity FIGURE 17.2 Pathways for particles to cause inflammation. There can be direct stimulation of target cells for pro-inflammatory gene expression, or inflammation can arise by more complex indirect routes that involve cytotoxicity or the immune system. Particle Toxicology304 © 2007 by Taylor & Francis Group, LLC In vitro models are usefultools at twostages, namely the assessment of toxicity as the second tier of the testing system and the elucidation of the mechanism(s) of action. Areduction in the use of experimental animals is aclear advantage of such studies, but in vitro systems alsoprovide a “simplified” model in which the details of the mechanism of action may be morereadily examined. The potential toxicity of particles is tested in vitro usingeither primary cells or cell lines in culture. There are anumber of cell types that are of obvious interest when investigatingthe potential effects of particles, including type Iand type II epithelial cells, Clara cells, alveolar macrophages,and neutrophils. Primary cells are obviously an advantage in that theyare not transformed, hence they will correspond more closely to the cells found in vivo.Primary cells, however,often have alimited life span in culture, and for thisreason, cell lines such as the A549 human type II cell line and THP-1monocytes are widelyused, due to their readyavailability and the fact that the pathways under study are similar in these permanent celllines to freshly derived cells of thesametype.The acquisition of human primarycellsremains difficult formany researchers, and for this reason primary rat cells are frequently used as an alternative. The use of cells has both benefits and drawbacks, both of which are well-known. Thebenefits include the ability to dissect out the sub-cellular pathways and responses and to isolatethe responses specific to the cell type in question. Thedrawbacks are that there is no influence of the other cell typesand the circulation that ordinarily plays an important role in the responses of any single celltype. 17.4.1 C YTOTOXICITY Anumber of reliable and well-documented techniques are available to assess viability following particle exposure, including the MTT assay, which measures the metabolic competence of cells by assessing the activityofsuccinate dehydrogenase enzymeactivity, akey enzymeincellular respir- ation. Assessment of lactate dehydrogenase(LDH) enzymeleakage from the cells measures plasma membrane integrity. The MTT assay and measurement of LDH leakage are appropriate for death via necrosis. Anumber of particle types have been proposed to induce programmed celldeath or apoptosis(Be ´ ruBe ´ et al. 1996; Iyer and Holian1997). Thetechniques available for the detection of apoptosisare numerous,from detection of DNA fragmentation to commercially available cell death ELISA kits and fluorescent dyes that label the DNA,such as propidium iodide and Hoechst 33342. 17.4.2 C ELL S TIMULATION Many particles are thought to induceeffects on the lung by mechanismsotherthan toxicity. Some particlesmay in fact cause the stimulation of various cell types by, e.g., increasing entry of calcium or stimulatingkinase activity,leading to cell proliferationoranincrease in theproduction of cytokinesand otherpro-inflammatory mediators. Measurement of theproductionofROS and cytokines, such as TNF-a and IL-1b by target cells, may form part of atesting strategy to discrimi- nate between non-pathogenicand pathogenicparticulates by theability of various particulate preparations to differentially producethese mediators. There are anumber of assays that can be used to assess cell stimulation in vitro and theseare outlined below. 17.4.3 C ELL P ROLIFERATION Increased cell proliferation has been notedonexposure of various cell typestodifferent particles. For example, treatment of primary rat type II epithelial cells with silicafor 24 hhas been shownto induce cell proliferation, as assessed by the incorporation of tritiated thymidine into the DNA of dividing cells. This simpletechnique has the disadvantage of usingradioactivity, although at alow level. Asimilar technique involves the incorporation of 5-bromo-2 0 -deoxyuridine (BrdU) into DNA whichisthenassessedbyimmunostaining(Timblin, Janssen, andMossman 1995).Thisnon- radioactive technique hasthe advantage that BrdU can be used for observation by microscopy as well as quantification through either spectroscopy or fluorimetry. Approaches to the Toxicological Testing of Particles 305 © 2007 by Taylor & Francis Group, LLC 17.4.4 C YTOKINE M EASUREMENT One of the mostobvious ways to assess the potential inflammogenic activity of aparticle is by measuring the outputofcytokines. Secreted cytokinesare frequently assayed in the culturemedia through the use of ELISA. In addition, the quantification of specificmRNA sequences, through either Northern Blotting or RT-PCR, allows further investigation of gene regulation on exposure to particles. The sametechniques are also applicable to other pro-inflammatory mediators. 17.4.5 O XIDATIVE S TRESS M EASUREMENT There is abundant data to suggest that manyparticle typesinducetheir effects on the lungs in part through ROS, leadingtooxidative stress (Tao,Gonzalez-Flecha,and Kobzik 2003).The free radicals produced at the surface of avariety of particle typesalong with the ROS released by leukocytes during phagocytosis and inflammation induce an oxidative stress within the lung leading to arange of events from oxidative damage to bio-molecules,suchasDNA andprotein,to activation of oxidative stress-responsivetranscription factorsthatleadtotranscription of pro- inflammatory genes (Rahman and MacNee 2000; Gilmour et al. 2003; Brownetal. 2004a). The measurement of intracellular glutathione in its reduced (GSH) and oxidized (GSSG) forms remains asensitive means by which the induction of oxidative stress can be assessed. GSH is one of the majorintracellular antioxidants(reviewed in Karin1998). In actingasanantioxidant,two molecules of GSH areoxidizedtoformGSSG, which is then reduced back to GSH by the enzyme glutathione reductase using NADPH as asource of reductant. When the cell is exposed to high levels of oxidants,NADPH within the cell is decreased, allowing depletion of GSH and an increaseinGSSG.The depletion of GSH is often used as amarker of oxidative stress in response to particles. Enzyme assays exist to measure the activityofenzymes such as g -glutamyl transpepti- dase ( g -GT), the enzyme responsible for the uptake of the components of GSH across the plasma membrane, and g -glutamylcysteine synthetase ( g -GCS),the rate-limiting enzyme in the synthesis of GSH. 17.4.6 A LTERATIONS IN C ELL S IGNALING C ASCADES 17.4.6.1 Intracellular Calcium Alterations in intracellular calcium homeostasis have been implicated following oxidative stress.In the resting nonstimulated cell, aCa 2 C ATPasepump in the plasma membrane actively extrudes Ca 2 C from the cell while adifferent Ca 2 C ATPase pump in the endoplasmic reticulum (ER) actively sequesters Ca 2 C into this intracellular store. The ER Ca 2 C store is released on activation of the cell by stimulants,which resultsinthe production of inositol 1,4,5-trisphosphate(IP 3 ). This sharp increase in cytosolic calciumconcentration stimulatesthe openingofCa 2 C channels in the plasma membrane (calcium releaseactivated calcium channels; CRAC channels) allowing Ca 2 C to enter the cell down it concentration gradient (calcium releaseactivated calcium current,I CRAC ) resultinginasustainedincrease in cytosolic Ca 2 C concentration(Parekh andPenner1997; Berridge,Bootman, and Lipp 1998; Berridge 2001). Anumber of the transport proteins involved in the maintenance of Ca 2 C homeostasisare sensitive to oxidative stress. Forexample, the Ca 2 C - ATPaseofthe ER containsacysteine residue that is susceptible to oxidation, as are the IP 3 receptor calcium channels in the ER. Cytosolic Ca 2 C can be measured using fluorescent dyes such as fura-2 (Grynkiewicz, Poenie, andTsien 1985)which altertheir fluorescent propertiesonbinding to Ca 2 C .Fura-2has the advantageofbeing aratiodye,which permitsalterations in background fluorescence, for exampledue to the introduction of particles, while measuring calcium. Ultrafine carbon black (CB) has been shown to increase the resting cytosolic calcium concen- tration of ahuman monocytic cellline MonoMac 6(MM6) (Stone et al. 2000). This effect was not Particle Toxicology306 © 2007 by Taylor & Francis Group, LLC observed with the same dose of larger,respirable CB particlesorwith pathogenic a -quartz (DQ12). PM 10 was shown to producethe sametype of calcium influx with associated TNFa gene expression (Brown, Donaldson, and Stone 2004b). Thapsigargin is auseful tool to investigatethe potential effects of particles on Ca 2 C signaling (Thastrup et al. 1990). Thapsigargin works by inhibiting the ER Ca 2 C ATPase, resulting in leak of the ER store contentsinto the cytosol.Treatment with thapsigarginresults in asharp increasein cytosolic Ca 2 C (comparable to the effect of IP 3 )followed by astimulationofthe I CRAC .Treatment of amacrophage cell line(MonoMac 6) with ufCB for 30 min induced an increaseinthe I CRAC observed on treatment with thapsigargin, through an increased opening of the plasma membrane Ca 2 C channels (Stone et al. 2000). Similar effects were seen in response to PM 10 —the increase in calcium was involved in TNFa gene expression (Brown, Donaldson, and Stone 2004b). 17.4.6.2 Mitogen-Activated Protein Kinase Cascade The mitogen-activated protein kinase (MAPK)cascade includes the extra cellular signal-related kinase (ERK1,ERK2) activated in response to growth factors, oxidative stress or phorbolesters via aRas-dependentmechanism,c-Jun amino terminal kinase/stress activated protein kinase (JNK1, JNK2) activatedbyTNF-a in aRas-independent manner and p38 (Seger and Krebs 1995). Activation of the MAPK cascade involving phosphorylation and dephosphorylation of anumber of proteins leads to the transactivation of c-fos and c-jun and anumber of interrelated transcription factors(Seger and Krebs 1995). Moreover,inanumber of cellular systems, the balance between the activation of several arms of this pathwayappears to governwhetherapoptosisorcellproliferation occurs (Xia et al. 1995). Limited studies have been carriedout investigatingthe influence of particulates on the MAPK pathwayand these have been reviewed by the authors (Brownetal. 2004a; Donaldson et al. 2004). One recent studyhas shownthat exposure of normal humanbronchial epithelial (NHBE) cells to ultrafineelemental carbon particlesinduced the phosphorylation and activation of p38 MAPK. In addition, inhibition of p38 MAPK activity blockedthe interleukin-8mRNAexpressionin these cells. 17.4.6.3 Transcription Factor Activation ROS and inflammatory cytokines both cause activation of the transcription factorsNF-k Band AP-1 (Meyer, Schreck, and Baeuerle 1993). In addition, ROS have been suggested to act as second messenger moleculeswithinthe cell (Sen et al.1997).The transcription factorsNF-k Band AP-1 have been showntoberegulated by the intracellular redox status (Piette et al. 1997; Ginn- Pease and Whisler 1998). NF-k Bisatranscription factor important in the regulationofanumber of genes intrinsic to inflammation, proliferation, and lung defenses (Schins and Donaldson2000) includingcytokines, nitric oxide synthase,adhesion molecules, and protooncogenes,such as c- myc.The process of NF-k Bactivation involves the cytoplasmic phosphorylation, ubiquitination and subsequent proteolytic degradation of the I k Binhibitory subunits from NF-k B. Release of NF- k Bfrom I k Ballowsuncovering of the nuclear localization site on the NF-k Bsubunits so that it can migrate to the nucleus. Onceinthe nucleus, the activated transcription factorcomplex, which include p65protein subunits(Donaldsonetal. 2004), then bind to promoter regions of genes that have consensus NF-k BDNA binding sequences. NF-k Bhas been shown to be activated by anumber of particles (Jimenezetal. 2000; Shukla et al. 2000; Hubbard et al. 2002; McNeilly et al. 2005). AP-1isafamily of accessory transcription factorsthat interact with otherregulatory sequences calledTPA-response element (TRE) or AP-1sites (Donaldson et al. 2004). AP-1 transcription factorsinclude homo- (Jun/Jun) andheterodimer(Fos/Jun) complexesencodedbyvarious members of the c-fos and c-jun families of protooncogenes. The functional ramifications of c-fos Approaches to the Toxicological Testing of Particles 307 © 2007 by Taylor & Francis Group, LLC and c-jun transactivation may be cell type specific, butFos and Jun proteins may regulate the expression of othergenes requiredfor the progression through the cell cycle, apoptosis, or cell transformation (Angel and Karin 1991). Anumber of different pathogenicparticleshave been found to activate AP-1 (Timblin, Be ´ ruBe ´ , and Mossman 1998; Jimenez et al. 2002; Marwick et al. 2004; Shukla et al. 2004; Brown et al. 2004a; McNeilly et al. in press). Activation of transcription factorscan be investigated via anumber of methods. These include the gel mobility shift or retardation assayoftranscription factor DNA binding activity, immuno- histochemicalanalysis of protein localizationincells, gene transactivation assays using reporter gene constructs measuring luciferaseactivity, and western blotting of proteinlevels. 17.4.7 E FFECTS ON B LOOD With growing investigations into the effects of nanoparticles, there is concern that particles might gain access to the blood and thereby exertdirect pathogenic effects on the cardiovascular system.It is therefore cogent to discuss testing systems for the effects of particles on the blood. 17.4.7.1 Endothelium Theendothelium is an extremely importantcell typeintimately involvedinthe regulationof vascular tone,clotting, fibrinolysis, andinflammation,soitisahighly relevant celltostudy. HUVECsand variants are available that allow themeasurementofthe effectsthatparticles gaining access to them might have, such as gene expression for clotting factorsand inflammatory mediators (Gilmour et al. 2005). 17.4.7.2 Platelets Platelets are akey cell in the initiation of thrombus formation.Ifparticlesgaining access to the blood could activate platelets, then there would likely be thrombus formation. Interactions between plateletsand various nanoparticles have been reported (Nemmar et al. 2004; Radomski et al. 2005) with evidence that some particles can causeplateletaggregationand up-regulation of surface adhesion molecules on the platelets. 17.4.7.3 Atherosclerotic Plaques If particlesbecome blood borne, they are likely to be deposited in the vessel wall in the same way as LDL and by the same forces of turbulence. They couldthen interact directly with the cells in the atherosclerotic lesion. There are currently no specific in vitro models to investigatethis aspect. 17.4.8 E FFECTS ON THE B RAIN Becauseofinterest in the braintranslocation of particles, studies of the effects of particles on neurons are warranted. These can address the penetration of the blood brain barrier, such as: 1. Evaluation of toxicity leading to increased permeability and opening of tight junctions between endothelial cells. If particlesinthe blood are able to have this effect, they may be able to leave the blood and traverse the BBB. 2. Testing the ability of particles to undergo endocytosis or transcytosis in endothelium of the BBB. 3. Assessment of the influence of NP on cellmembrane fluidity, which may leadtoinhi- bition of the brain efflux system. Particle Toxicology308 © 2007 by Taylor & Francis Group, LLC [...]... V., Calcium and ROS-mediated activation of transcription factors and TNF-alpha cytokine gene expression in macrophages exposed to ultra-fine particles, Am J Physiol Lung Cell Mol Physiol., 286, L344–L353, 2004 Calderon-Garciduenas, L., Reed, W., Maronpot, R R., Henriquez-Roldan, C., Gado-Chavez, R., CalderonGarciduenas, A., Dragustinovis, I et al., Brain inflammation and Alzheimer’s-like pathology in... problem of “local overload” of lung defenses when non-toxic particles are used in this assay Therefore, the test particles should be compared against control particles of known high toxicity and of low toxicity 17. 5.3 INHALATION STUDIES Inhalation studies are commonly carried out in conventional toxicology protocols to determine the likely toxicity of particles Standard protocols are available for inhalation... detecting particle effects It is important to note that the carcinogenic effects detected for a number of particles in animal models, e.g., quartz and some non-toxic particles at overload, are a consequence of inflammation and so would not be detected by “classical” assays of genotoxicity 17. 5 ANIMAL STUDIES 17. 5.1 BRONCHOALVEOLAR LAVAGE Inflammation lies at the center of the adverse effects of particles... the OECD in Europe (see http://www.oecd.org/document/55/0,2340,en_2649_34377_2349687_1_1_1_1,00 html) and OPPTS in the U.S.A (see http://www.epa.gov/opptsfrs/publications/OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/Series/) © 2007 by Taylor & Francis Group, LLC 310 Particle Toxicology TABLE 17. 4 Different Lengths of Toxicological Study Study Timescale Subacute Subchronic Chronic Carcinogenicity... Testing of Particles 311 Zelikoff 1999) Impairment of pulmonary defenses against bacteria is described following exposure to metals and to wood smoke particles This is an important area where more information is needed and where specific particles may have considerable impact 17. 5.7 ANIMAL MODELS TO STUDY THE CARDIOVASCULAR EFFECTS OF PARTICLES Animal models can be used to monitor effects of particles... diesel particles was associated with endothelial dysfunction that was very likely oxidative stress-mediated (Mills et al 2005) 17. 5.8 ANIMAL MODELS OF NEUROLOGICAL EFFECTS OF PARTICLES ¨ Following on from the report by Oberdorster that nanoparticles can translocate from the lungs to the ¨ brain (Oberdorster et al 2004), reports of brain abnormalities in humans exposed to high pollution (Calderon-Garciduenas... and inhalation for particles has demonstrated that the same types of qualitative response were seen (Henderson et al 1995) The technique can be used for histopathology studies, but it is more often used in combination with BAL to study the short to medium-term inflammatory response to a suspect particle in relation to a known pathogenic particle such as quartz and a non-pathogenic particle such as pigment... IL-8 release mediated by oxidative stress from environmental particles, Am J Physiol Lung Cell Mol Physiol., 284, L533–L540, 2003 Gilmour, P S., Morrison, E R., Vickers, M A., Ford, I., Ludlam, C A., Greaves, M., Donaldson, K., and MacNee, W., The procoagulant potential of environmental particles (PM10), Occup Environ Med., 62, 164 171 , 2005 Ginn-Pease, M E and Whisler, R L., Redox signals and NF-kappaB... Donaldson, K., Soluble transition metals in welding fumes cause inflammation via activation of NF-kappaB and AP-1, Toxicol Lett., 158, 152–157, 2005 Meyer, M., Schreck, R., and Baeuerle, P A., H2O2 and antioxidants have opposite effects on activation of NF-kappa B and AP-1 in intact cells: AP-1 as secondary antioxidant-responsive factor, EMBO J., 12, 2005–2015, 1993 Mills, N L., Tornqvist, H., Robinson, S... particulate matter causes expression of nuclear factor (NF)-kappaB-related genes and oxidantdependent NF-kappaB activation in vitro, Am J Respir Cell Mol Biol., 23, 182–187, 2000 Shukla, A., Flanders, T., Lounsbury, K M., and Mossman, B T., The gamma-glutamylcysteine synthetase and glutathione regulate asbestos-induced expression of activator protein-1 family members and activity, Cancer Res., 64, 7780–7786, . University CONTENTS 17. 1 Background to Testing of Particles 300 17. 2 Factors Affecting Particle Toxicity 300 17. 2.1 Particle Size 300 17. 2.1.1 Particle Shape 301 17. 2.1.2 Particle- Derived Transition Metals 302 17. 2.1.3. 302 17. 2.1.3 Particle- Derived Free Radicals 303 17. 2.1.4 Particle Biopersistence 303 17. 3 Approaches to Testing 304 17. 3.1 Characterizing the Particles 304 17. 4 Assessment of Toxicity In Vitro 304 17. 4.1. Calcium 306 17. 4.6.2 Mitogen-Activated Protein Kinase Cascade 307 17. 4.6.3 Transcription Factor Activation 307 17. 4.7 Effects on Blood 308 17. 4.7.1 Endothelium 308 17. 4.7.2 Platelets 308 17. 4.7.3