21st Century Approaches for Evaluating Exposures, Biological Activity, and Risks of Complex Substances: Workshop highlights.

The June 2019 workshop 21st Century Approaches for Evaluating Exposures, Biological Activity, and Risks of Complex Substances, co-organised by the International Council of Chemical Association's Long-Range Research Initiative and the European Commission's Joint Research Centre, is summarised. Focus was the need for improved approaches to evaluate the safety of complex substances. Approximately 10% and 20% of substances registered under the EU chemicals legislation are 'multi-constituent substances' and 'substances of unknown or variable compositions, complex reaction products and biological substances' (UVCBs), respectively, and UVCBs comprise approximately 25% of the U.S. Toxic Substances Control Act Inventory. Workshop participants were asked to consider how the full promise of new approach methodologies (NAMs) could be brought to bear to evaluate complex substances. Sessions focused on using NAMs for screening, biological profiling, and in complex risk evaluations; improving read-across approaches employing new data streams; and methods to evaluate exposure and dosimetry. The workshop concluded with facilitated discussions to explore actionable steps forward. Given the diversity of complex substances, no single 'correct' approach was seen as workable. The path forward should focus on 'learning by doing' by developing and openly sharing NAM-based fit-for-purpose case examples for evaluating biological activity, exposures and risks of complex substances.

Updating physiologically based pharmacokinetic models for manganese by incorporating rapid association/dissociation processes in tissues.

Previously, we developed a series of physiologically based pharmacokinetic (PBPK) models for manganese (Mn) in which saturable tissue binding and dose-dependent increases in biliary excretion captured key aspects of Mn homeostasis biology. These models reproduced the non-linear behavior of Mn kinetics in different tissues, accounting for dose-dependent changes in Mn kinetics. The original model construct had relatively slow association and dissociation rate constants for Mn binding in tissues. In this updated model, both rates of entry into tissue and the interaction of Mn with binding sites are rapid, and the step limiting Mn accumulation is the saturation of tissue binding sites. This binding reflects general cellular requirements for Mn with high affinity but rapid exchange between bound and free forms, which we captured using a dissociation constant (KD) of ~ 0.5 µM across tissues while maintaining different maximum binding capacities in each tissue. Variability in the binding capacities accounted for different background levels of Mn in particular tissues. This alternative structure successfully described Mn kinetics in tissues in adult rats exposed to Mn either in their diet or by inhalation, indicating that both the original and the present models capture the dose-dependent and tissue-specific kinetic behavior of Mn in adult rats. Although the published models that emphasize the role of smaller tissue binding rate constants in non-linear behaviors capture all relevant dose-dependent kinetic behaviors of this metal, increasing biological relevance of the model structure and parameters should provide greater confidence in applying the Mn PBPK models to risk assessment.

Physiologically-based pharmacokinetic modeling suggests similar bioavailability of Mn from diet and drinking water.

Due to concerns for enhanced absorption of manganese (Mn) from drinking water compared to diet, bioavailability of Mn from drinking water remains a major data gap in understanding Mn kinetics. In this study, PBPK models for adult rats and humans were updated with a drinking water exposure route and were used to assess the homeostatic control of Mn uptake, excretion and tissue kinetics between the two different ingestion modes. Drinking water model parameters were estimated from tissue kinetic data from a drinking water study in rats. The published study included a 10 ppm-Mn diet with additional Mn added to drinking water to give a total ingested Mn dose equivalent to that from a 200 ppm diet. The 200 ppm diet and equivalent mixed drinking water/diet exposures provided Mn concentrations for brain (striatum, olfactory bulb, and cerebellum), liver and bone after 7 and 61 days of Mn exposure. Modeling of these data sets indicated that (1) the oral Mn bioavailability is similar for diet or drinking water and (2) homeostatic control of gut uptake of Mn occurs with either drinking water or dietary ingestion. This updated description for absorption and distribution of Mn from gut was added to a human Mn-PBPK model to simulate Mn exposure from multiple routes of exposure (i.e. dietary intake, drinking water, and inhalation). This increases the utility of the Mn PBPK model by allowing for the simulation of multiple Mn exposure scenarios, including variable daily food and drinking water exposures in a human population.

A tissue dose-based comparative exposure assessment of manganese using physiologically based pharmacokinetic modeling-The importance of homeostatic control for an essential metal.

A physiologically-based pharmacokinetic (PBPK) model (Schroeter et al., 2011) was applied to simulate target tissue manganese (Mn) concentrations following occupational and environmental exposures. These estimates of target tissue Mn concentrations were compared to determine margins of safety (MOS) and to evaluate the biological relevance of applying safety factors to derive acceptable Mn air concentrations. Mn blood concentrations measured in occupational studies permitted verification of the human PBPK models, increasing confidence in the resulting estimates. Mn exposure was determined based on measured ambient air Mn concentrations and dietary data in Canada and the United States (US). Incorporating dietary and inhalation exposures into the models indicated that increases in target tissue concentrations above endogenous levels only begin to occur when humans are exposed to levels of Mn in ambient air (i.e. >10µg/m3) that are far higher than those currently measured in Canada or the US. A MOS greater than three orders of magnitude was observed, indicating that current Mn air concentrations are far below concentrations that would be required to produce the target tissue Mn concentrations associated with subclinical neurological effects. This application of PBPK modeling for an essential element clearly demonstrates that the conventional application of default factors to "convert" an occupational exposure to an equivalent continuous environmental exposure, followed by the application of safety factors, is not appropriate in the case of Mn. PBPK modeling demonstrates that the relationship between ambient Mn exposures and dose-to-target tissue is not linear due to normal tissue background levels and homeostatic controls.

Case studies putting the decision-making framework for the grouping and testing of nanomaterials (DF4nanoGrouping) into practice.

Case studies covering carbonaceous nanomaterials, metal oxide and metal sulphate nanomaterials, amorphous silica and organic pigments were performed to assess the Decision-making framework for the grouping and testing of nanomaterials (DF4nanoGrouping). The usefulness of the DF4nanoGrouping for nanomaterial hazard assessment was confirmed. In two tiers that rely exclusively on non-animal test methods followed by a third tier, if necessary, in which data from rat short-term inhalation studies are evaluated, nanomaterials are assigned to one of four main groups (MGs). The DF4nanoGrouping proved efficient in sorting out nanomaterials that could undergo hazard assessment without further testing. These are soluble nanomaterials (MG1) whose further hazard assessment should rely on read-across to the dissolved materials, high aspect-ratio nanomaterials (MG2) which could be assessed according to their potential fibre toxicity and passive nanomaterials (MG3) that only elicit effects under pulmonary overload conditions. Thereby, the DF4nanoGrouping allows identifying active nanomaterials (MG4) that merit in-depth investigations, and it provides a solid rationale for their sub-grouping to specify the further information needs. Finally, the evaluated case study materials may be used as source nanomaterials in future read-across applications. Overall, the DF4nanoGrouping is a hazard assessment strategy that strictly uses animals as a last resort.

A decision-making framework for the grouping and testing of nanomaterials (DF4nanoGrouping).

The European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC) 'Nano Task Force' proposes a Decision-making framework for the grouping and testing of nanomaterials (DF4nanoGrouping) that consists of 3 tiers to assign nanomaterials to 4 main groups, to perform sub-grouping within the main groups and to determine and refine specific information needs. The DF4nanoGrouping covers all relevant aspects of a nanomaterial's life cycle and biological pathways, i.e. intrinsic material and system-dependent properties, biopersistence, uptake and biodistribution, cellular and apical toxic effects. Use (including manufacture), release and route of exposure are applied as 'qualifiers' within the DF4nanoGrouping to determine if, e.g. nanomaterials cannot be released from a product matrix, which may justify the waiving of testing. The four main groups encompass (1) soluble nanomaterials, (2) biopersistent high aspect ratio nanomaterials, (3) passive nanomaterials, and (4) active nanomaterials. The DF4nanoGrouping aims to group nanomaterials by their specific mode-of-action that results in an apical toxic effect. This is eventually directed by a nanomaterial's intrinsic properties. However, since the exact correlation of intrinsic material properties and apical toxic effect is not yet established, the DF4nanoGrouping uses the 'functionality' of nanomaterials for grouping rather than relying on intrinsic material properties alone. Such functionalities include system-dependent material properties (such as dissolution rate in biologically relevant media), bio-physical interactions, in vitro effects and release and exposure. The DF4nanoGrouping is a hazard and risk assessment tool that applies modern toxicology and contributes to the sustainable development of nanotechnological products. It ensures that no studies are performed that do not provide crucial data and therefore saves animals and resources.

A critical appraisal of existing concepts for the grouping of nanomaterials.

The grouping of substances serves to streamline testing for regulatory purposes. General grouping approaches for chemicals have been implemented in, e.g., the EU chemicals regulation. While specific regulatory frameworks for the grouping of nanomaterials are unavailable, this topic is addressed in different publications, and preliminary guidance is provided in the context of substance-related legislation or the occupational setting. The European Centre for Ecotoxicology and Toxicology of Chemicals Task Force on the Grouping of Nanomaterials reviewed available concepts for the grouping of nanomaterials for human health risk assessment. In their broad conceptual design, the evaluated approaches are consistent or complement each other. All go beyond the determination of mere structure-activity relationships and are founded on different aspects of the nanomaterial life cycle. These include the NM's material properties and biophysical interactions, specific types of use and exposure, uptake and kinetics, and possible early and apical biological effects. None of the evaluated grouping concepts fully take into account all of these aspects. Subsequent work of the Task Force will aim at combining the available concepts into a comprehensive 'multiple perspective' framework for the grouping of nanomaterials that will address all of the mentioned aspects of their life cycles.

Incorporation of rapid association/dissociation processes in tissues into the monkey and human physiologically based pharmacokinetic models for manganese.

In earlier physiologically based pharmacokinetic (PBPK) models for manganese (Mn), the kinetics of transport of Mn into and out of tissues were primarily driven by slow rates of association and dissociation of Mn with tissue binding sites. However, Mn is known to show rapidly reversible binding in tissues. An updated Mn model for primates, following similar work with rats, was developed that included rapid association/dissociation processes with tissue Mn-binding sites, accumulation of free Mn in tissues after saturation of these Mn-binding sites and rapid rates of entry into tissues. This alternative structure successfully described Mn kinetics in tissues in monkeys exposed to Mn via various routes including oral, inhalation, and intraperitoneal, subcutaneous, or intravenous injection and whole-body kinetics and tissue levels in humans. An important contribution of this effort is showing that the extension of the rate constants for binding and cellular uptake established in the monkey were also able to describe kinetic data from humans. With a consistent model structure for monkeys and humans, there is less need to rely on cadaver data and whole-body tracer studies alone to calibrate a human model. The increased biological relevance of the Mn model structure and parameters provides greater confidence in applying the Mn PBPK models to risk assessment. This model is also well-suited to explicitly incorporate emerging information on the role of transporters in tissue disposition, intestinal uptake, and hepatobiliary excretion of Mn.

Adsorption and conformation of serum albumin protein on gold nanoparticles investigated using dimensional measurements and in situ spectroscopic methods.

The adsorption and conformation of bovine serum albumin (BSA) on gold nanoparticles (AuNPs) were interrogated both qualitatively and quantitatively via complementary physicochemical characterization methods. Dynamic light scattering (DLS), asymmetric-flow field flow fractionation (AFFF), fluorescence spectrometry, and attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy were combined to characterize BSA-AuNP conjugates under fluid conditions, while conjugates in the aerosol state were characterized by electrospray-differential mobility analysis (ES-DMA). The presence of unbound BSA molecules interferes with DLS analysis of the conjugates, particularly as the AuNP size decreases (i.e., below 30 nm in diameter). Under conditions where the γ value is high, where γ is defined as the ratio of scattering intensity by AuNPs to the scattering intensity by unbound BSA, DLS size results are consistent with results obtained after fractionation by AFFF. Additionally, the AuNP hydrodynamic size exhibits a greater proportional increase due to BSA conjugation at pH values below 2.5 compared with less acidic pH values (3.4-7.3), corresponding with the reversibly denatured (E or F form) conformation of BSA below pH 2.5. Over the pH range from 3.4 to 7.3, the hydrodynamic size of the conjugate is nearly constant, suggesting conformational stability over this range. Because of the difference in the measurement environment, a larger increase of AuNP size is observed following BSA conjugation when measured in the wet state (i.e., by DLS and AFFF) compared to the dry state (by ES-DMA). Molecular surface density for BSA is estimated based on ES-DMA and fluorescence measurements. Results from the two techniques are consistent and similar, but slightly higher for ES-DMA, with an average adsorbate density of 0.015 nm(-2). Moreover, from the change of particle size, we determine the extent of adsorption for BSA on AuNPs using DLS and ES-DMA at 21 °C, which show that increasing the concentration of BSA increases the measured change in AuNP size. Using ES-DMA, we observe that the BSA surface density reaches 90% of saturation at a solution phase concentration between 10 and 30 µmol/L, which is roughly consistent with fluorescence and ATR-FTIR results. The equilibrium binding constant for BSA on AuNPs is calculated by applying the Langmuir equation, with resulting values ranging from 0.51 × 10(6) to 1.65 × 10(6) L/mol, suggesting a strong affinity due to bonding between the single free exterior thiol on N-form BSA (associated with a cysteine residue) and the AuNP surface. Moreover, the adsorption interaction induces a conformational change in BSA secondary structure, resulting in less α-helix content and more open structures (ß-sheet, random, or expanded).

Pharmacokinetic evaluation of the equivalency of gavage, dietary, and drinking water exposure to manganese in F344 rats.

Concerns exist as to whether individuals may be at greater risk for neurotoxicity following increased manganese (Mn) oral intake. The goals of this study were to determine the equivalence of 3 methods of oral exposure and the rate (mg Mn/kg/day) of exposure. Adult male rats were allocated to control diet (10 ppm), high manganese diet (200 ppm), manganese-supplemented drinking water, and manganese gavage treatment groups. Animals in the drinking water and gavage groups were given the 10 ppm manganese diet and supplemented with manganese chloride (MnCl(2)) in drinking water or once-daily gavage to provide a daily manganese intake equivalent to that seen in the high-manganese diet group. No statistically significant difference in body weight gain or terminal body weights was seen. Rats were anesthetized following 7 and 61 exposure days, and samples of bile and blood were collected. Rats were then euthanized and striatum, olfactory bulb, frontal cortex, cerebellum, liver, spleen, and femur samples were collected for chemical analysis. Hematocrit was unaffected by manganese exposure. Liver and bile manganese concentrations were elevated in all treatment groups on day 61 (relative to controls). Increased cerebellum manganese concentrations were seen in animals from the high-manganese diet group (day 61, relative to controls). Increased (relative to all treatment groups) femur, striatum, cerebellum, frontal cortex, and olfactory bulb manganese concentrations were also seen following gavage suggesting that dose rate is an important factor in the pharmacokinetics of oral manganese. These data will be used to refine physiologically based pharmacokinetic models, extending their utility for manganese risk assessment by including multiple dietary exposures.

Update on a Pharmacokinetic-Centric Alternative Tier II Program for MMT-Part II: Physiologically Based Pharmacokinetic Modeling and Manganese Risk Assessment.

Recently, a variety of physiologically based pharmacokinetic (PBPK) models have been developed for the essential element manganese. This paper reviews the development of PBPK models (e.g., adult, pregnant, lactating, and neonatal rats, nonhuman primates, and adult, pregnant, lactating, and neonatal humans) and relevant risk assessment applications. Each PBPK model incorporates critical features including dose-dependent saturable tissue capacities and asymmetrical diffusional flux of manganese into brain and other tissues. Varied influx and efflux diffusion rate and binding constants for different brain regions account for the differential increases in regional brain manganese concentrations observed experimentally. We also present novel PBPK simulations to predict manganese tissue concentrations in fetal, neonatal, pregnant, or aged individuals, as well as individuals with liver disease or chronic manganese inhalation. The results of these simulations could help guide risk assessors in the application of uncertainty factors as they establish exposure guidelines for the general public or workers.

Tissue and cellular distribution of gold nanoparticles varies based on aggregation/agglomeration status.

AIM: The ability of nanoparticles to form larger superstructures of aggregates and agglomerates has been extensively noted in the literature. The in vivo biological impact of these structures, however, has not been assessed. This knowledge gap is especially critical in the safety assessment of nanoparticles to be used for therapeutic purposes. METHOD/RESULTS: Here we show that when administered intravenously into a mouse model, gold nanoparticle superstructures of reversible agglomerates and irreversible aggregates demonstrate significant differences in organ and cellular distribution compared with the primary particle building blocks. In addition, different structures produced different blood serum chemistry data. CONCLUSION: These findings raise the possibility for different mechanisms of toxicity between the structures. Such a possibility necessitates complete characterization and stability assessment of nanomaterials prior to their in vivo administration.

Tumor necrosis factor interaction with gold nanoparticles.

We report on a systematic investigation of molecular conjugation of tumor necrosis factor-α (TNF) protein onto gold nanoparticles (AuNPs) and the subsequent binding behavior to its antibody (anti-TNF). We employ a combination of physical and spectroscopic characterization methods, including electrospray-differential mobility analysis, dynamic light scattering, polyacrylamide gel electrophoresis, attenuated total reflectance-Fourier transform infrared spectroscopy, fluorescence assay, and enzyme-linked immunosorbent assay. The native TNF used in this study exists in the active homotrimer configuration prior to conjugation. After binding to AuNPs, the maximum surface density of TNF is (0.09 ± 0.02) nm(-2) with a binding constant of 3 × 10(6) (mol L(-1))(-1). Dodecyl sulfate ions induce desorption of monomeric TNF from the AuNP surface, indicating a relatively weak intermolecular binding within the AuNP-bound TNF trimers. Anti-TNF binds to both TNF-conjugated and citrate-stabilized AuNPs, showing that non-specific binding is significant. Based on the number of anti-TNF molecules adsorbed, a substantially higher binding affinity was observed for the TNF-conjugated surface. The inclusion of thiolated polyethylene glycol (SH-PEG) on the AuNPs inhibits the binding of anti-TNF, and the amount of inhibition is related to the number ratio of surface bound SH-PEG to TNF and the way in which the ligands are introduced. This study highlights the challenges in quantitatively characterizing complex hybrid nanoscale conjugates, and provides insight on TNF-AuNP formation and activity.

Measuring agglomerate size distribution and dependence of localized surface plasmon resonance absorbance on gold nanoparticle agglomerate size using analytical ultracentrifugation.

Agglomeration of nanoparticles during measurements in relevant biological and environmental media is a frequent problem in nanomaterial property characterization. The primary problem is typically that any changes to the size distribution can dramatically affect the potential nanotoxicity or other size-determined properties, such as the absorbance signal in a biosensor measurement. Herein we demonstrate analytical ultracentrifugation (AUC) as a powerful method for measuring two critical characteristics of nanoparticle (NP) agglomerates in situ in biological media: the NP agglomerate size distribution, and the localized surface plasmon resonance (LSPR) absorbance spectrum of precise sizes of gold NP agglomerates. To characterize the size distribution, we present a theoretical framework for calculating the hydrodynamic diameter distribution of NP agglomerates from their sedimentation coefficient distribution. We measure sedimentation rates for monomers, dimers, and trimers, as well as for larger agglomerates with up to 600 NPs. The AUC size distributions were found generally to be broader than the size distributions estimated from dynamic light scattering and diffusion-limited colloidal aggregation theory, an alternative bulk measurement method that relies on several assumptions. In addition, the measured sedimentation coefficients can be used in nanotoxicity studies to predict how quickly the agglomerates sediment out of solution under normal gravitational forces, such as in the environment. We also calculate the absorbance spectra for monomer, dimer, trimer, and larger gold NP agglomerates up to 600 NPs, to enable a better understanding of LSPR biosensors. Finally, we validate a new method that uses these spectra to deconvolute the net absorbance spectrum of an unknown bulk sample and approximate the proportions of monomers, dimers, and trimers in a polydisperse sample of small agglomerates, so that every sample does not need to be measured by AUC. These results demonstrate the potential utility of AUC to characterize NP agglomeration and sedimentation for nanotoxicity and biosensor studies, as well as to characterize NP agglomerate size and absorbance to improve LSPR and surface-enhanced Raman spectroscopy based biosensors.

Gold nanoparticle trafficking of typically excluded compounds across the cell membrane in JB6 Cl 41-5a cells causes assay interference.

Nanoparticles (NP) often interfere with the mechanism and interpretation of high throughput in vitro toxicity assays. This interference may occur at any time during the assay and spans most NP systems. This study reports on a specific type of gold NP assay interference, where unmodified gold NPs were able to traffic certain assay molecules that contained primary amines across the cell membrane resulting in false positive results for toxicity assays. The enhanced assay molecule permeability was eliminated when the gold NP surface was both sterically and chemically blocked by polyethylene glycol (PEG). The results support the growing consensus that appropriate controls and assay validation should occur prior to interpretation of results of assays using NP.

Multivalent dendrimeric and monomeric adenosine agonists attenuate cell death in HL-1 mouse cardiomyocytes expressing the A(3) receptor.

Multivalent dendrimeric conjugates of GPCR ligands may have increased potency or selectivity in comparison to monomeric ligands, a phenomenon that was tested in a model of cytoprotection in mouse HL-1 cardiomyocytes. Quantitative RT-PCR indicated high expression levels of endogenous A(1) and A(2A) adenosine receptors (ARs), but not of A(2B) and A(3)ARs. Activation of the heterologously expressed human A(3)AR in HL-1 cells by AR agonists significantly attenuated cell damage following 4h exposure to H(2)O(2) (750 microM) but not in untransfected cells. The A(3) agonist IB-MECA (EC(50) 3.8 microM) and the non-selective agonist NECA (EC(50) 3.9 microM) protected A(3) AR-transfected cells against H(2)O(2) in a concentration-dependent manner, as determined by lactate dehydrogenase release. A generation 5.5 PAMAM (polyamidoamine) dendrimeric conjugate of a N(6)-chain-functionalized adenosine agonist was synthesized and its mass indicated an average of 60 amide-linked nucleoside moieties out of 256 theoretical attachment sites. It non-selectively activated the A(3)AR to inhibit forskolin-stimulated cAMP formation (IC(50) 66nM) and, similarly, protected A(3)-transfected HL-1 cells from apoptosis-inducing H(2)O(2) with greater potency (IC(50) 35nM) than monomeric nucleosides. Thus, a PAMAM conjugate retained AR binding affinity and displayed greatly enhanced cardioprotective potency.