Ingrid Nobels

Eco-, geno- and human toxicology of bio-active nanoparticles for biomedical applications

Gene delivery has become an increasingly important strategy for treating a variety of human diseases, including infections, genetic disorders and tumours. To avoid the difficulties of using viral carriers, more and more non-viral gene delivery nanoparticles are developed. Among these new approaches polyethylene imine (PEI) is currently considered as one of the most effective polymer based method solution and considered as the gold standard. The toxicity of nanoparticles is a major concern when used for medical application. In this study we chose two nanoparticles for an in depth toxicological and ecotoxicological evaluation, one well characterized, PEI, and another novel polymer, poly(2-(dimethylamino)ethyl methacrylate) (PDMAEMA). In the present study we have assessed the toxicity of these cation nanoparticles as such and of the polyplexes - nanoparticles covered with DNA. As these nanoparticles are also frequently used in high volumes in various industries and as such may enter in the environment, we also made an initial assessment of ecotoxicological effects assessment. The following nanoparticles related aspects have been studied during the project: development and characterization, ecotoxicity, general toxicity and specific toxicity. To this end a battery of different tests was used. The conclusion of these tests is that toxicity is varying between different nanoparticles and between different DNA covering ratios. In general, in the different systems tested, the PEI polymer is more toxic than the PDMAEMA polymer. The same difference is seen for the polyplexes and the higher the charge ratio, the more toxic are the polyplexes. Our study also clearly shows the need for a broad spectrum of toxicity assays for a comprehensive risk assessment. Our study has performed such a comprehensive analysis of two biomedical nanoparticles.

Toxicity Ranking and Toxic Mode of Action Evaluation of Commonly Used Agricultural Adjuvants on the Basis of Bacterial Gene Expression Profiles

The omnipresent group of pesticide adjuvants are often referred to as “inert” ingredients, a rather misleading term since consumers associate this term with “safe”. The upcoming new EU regulation concerning the introduction of plant protection products on the market (EC1107/2009) includes for the first time the demand for information on the possible negative effects of not only the active ingredients but also the used adjuvants. This new regulation requires basic toxicological information that allows decisions on the use/ban or preference of use of available adjuvants. In this study we obtained toxicological relevant information through a multiple endpoint reporter assay for a broad selection of commonly used adjuvants including several solvents (e.g. isophorone) and non-ionic surfactants (e.g. ethoxylated alcohols). The used assay allows the toxicity screening in a mechanistic way, with direct measurement of specific toxicological responses (e.g. oxidative stress, DNA damage, membrane damage and general cell lesions). The results show that the selected solvents are less toxic than the surfactants, suggesting that solvents may have a preference of use, but further research on more compounds is needed to confirm this observation. The gene expression profiles of the selected surfactants reveal that a phenol (ethoxylated tristyrylphenol) and an organosilicone surfactant (ethoxylated trisiloxane) show little or no inductions at EC20 concentrations, making them preferred surfactants for use in different applications. The organosilicone surfactant shows little or no toxicity and good adjuvant properties. However, this study also illustrates possible genotoxicity (induction of the bacterial SOS response) for several surfactants (POEA, AE, tri-EO, EO FA and EO NP) and one solvent (gamma-butyrolactone). Although the number of compounds that were evaluated is rather limited (13), the results show that the used reporter assay is a promising tool to rank commonly used agricultural adjuvants based on toxicity and toxic mode of action data.

Aquatic multi-species acute toxicity of (chlorinated) anilines: experimental versus predicted data.

Aquatic toxicity information is essential in environmental risk assessment to determine the potential hazards and risks of new and existing chemicals. Prediction and modelling techniques, such as quantitative structure activity relationships (QSAR) and species sensitivity distributions (SSDs), are applied to fill data gaps and to predict, assess and extrapolate the toxicity of chemicals. In this study, both techniques (i.e. the ECOSAR programme as QSAR tool and SSDs) were assessed for a set of polar narcotic structural analogues that differ in their degree of chloro-substitution (aniline, 4-chloroaniline, 3,5-dichloroaniline and 2,3,4-chloroaniline). The acute toxicity of these compounds was tested in one prokaryote species (Escherichia coli) and three eukaryote aquatic species (Pseudokirchneriella subcapitata, Daphnia magna and Danio rerio). Consequently, the experimental acute toxicity data were compared to the QSAR predictions made by the ECOSAR programme and compared to the species sensitivity modelling results. Large interspecies differences in sensitivity were observed (D. magna>P. subcapitata>D. rerio>E. coli). 4-Chloroaniline acted as an outlier in P. subcapitata toxicity. Whereas in D. magna, toxicity decreased rather than increased with increasing logK(ow) of the test compounds. In general, large interchemical and interspecies differences in toxicity of these relatively simple chemical structures were observed. Moreover, this species variation could not entirely be characterized by the ECOSAR tool. SSD modelling is particularly focussed on species variations and emphasis is put on protecting those species that are most affected by chemical exposure. Compared to QSARs, SSDs offer broader perspectives regarding species sensitivity ranking, however, in this study they could only be applied for aniline and 4-chloroaniline.

Application of a multiple endpoint bacterial reporter assay to evaluate toxicological relevant endpoints of perfluorinated compounds with different functional groups and varying chain length.

Perfluorinated compounds are widely distributed in the environment; good knowledge about the toxic mode of action of these compounds can contribute to improved molecular design and risk assessment. The studied compounds were evaluated with a bacterial multiple endpoint reporter assay for responses in four different mode of action classes (oxidative damage, DNA damage, general cell lesions and membrane damage). The results of our study clearly demonstrate that inductions of stress responsive genes occur for the different compounds and confirm some of the known mechanisms of work for well studied compounds like PFOA and PFOS, and in addition provide new information for less studied compounds. Few inductions were observed after exposure to the low carbon number carboxylic acids, PFBtA (CF(3)(CF(2))(2)C(O)O(-)), PFPtA (CF(3)(CF(2))(3)C(O)O(-)), PFHxA (CF(3)(CF(2))(4)C(O)O(-)) and PFHpA (CF(3)(CF(2))(5)C(O)O(-)) at equimolar concentrations (0.0156-1 mM). The induction of membrane damage markers (MicF and OsmY) is prominently present after exposure to PFOS (CF(3)(CF(2))(7)SO(3)(-)) and even more after exposure to PFNA (CF(3)(CF(2))(7)C(O)O(-)). This is the first report describing the mode of action of carboxylic acids with 11 and 12 carbon atoms; they are equally potent inducers relative to PFOS and PFNA. Overall, the effects seen at the level of gene expression were higher for the sulfonic acids than for the carboxylic acids, but the effect of the chain length is more important than the effect of the functional group.

Dose–response relationships and statistical performance of a battery of bacterial gene profiling assays

Because of increasing awareness and legislative demands, there is a demand for the development and use of biological assays for the assessment of the toxicity of chemicals, environmental samples. Recently, a growing number of bacterial reporter assays have been developed and implemented. Nevertheless, little data is published on the performance of these assays in terms of analytical parameters. We present results on a battery of 14 transgenic Escherichia coli strains carrying different promoter::reporter fusions. Growth characteristics and basal expression levels were modeled and fitted, data show that growth curves should be taken into account during test development. Our study shows that the induction profiles reflect the mode of action, e.g., paraquat clearly induces the soxRS operon. The sensitivity of the assay compares well to that of whole organism tests, e.g., fish and Daphnia for polar organics. Metal toxicity is detected less efficiently, e.g., cadmium is detected near the LC50 of carp, considered a relatively insensitive species towards cadmium. The assay variability ranges from 10 to 40% depending on the strain, comparable to that of other bioassays. The variability was shown to be determined by the intrinsic traits of the promoter–strain combination, not by operating conditions.

MIXTURE TOXICITY AND GENE INDUCTIONS: CAN WE PREDICT THE OUTCOME?

As a consequence of the nature of most real-life exposure scenarios, the last decade of ecotoxicological research has seen increasing interest in the assessment of mixture ecotoxicology. Often, mixtures are considered to follow one of two models, concentration addition (CA) or response addition (RA), both of which have been described in the literature. Nevertheless, mixtures that deviate from either or both models exist; they typically exhibit phenomena like synergism, ratio or concentration dependency, or inhibition. Moreover, both CA and RA have been challenged and evaluated mainly for acute responses at relatively high levels of biological organization (e.g., whole-organism mortality), and applicability to genetic responses has not received much attention. Genetic responses are considered to be the primary reaction in case of toxicant exposure and carry valuable mechanistic information. Effects at the gene-expression level are at the heart of the mode of action by toxicants and mixtures. The ability to predict mixture responses at this primary response level is an important asset in predicting and understanding mixture effects at different levels of biological organization. The present study evaluated the applicability of mixture models to stress gene inductions in Escherichia coli employing model toxicants with known modes of action in binary combinations. The results showed that even if the maximum of the dose-response curve is not known, making a classical ECx (concentration causing x% effect) approach impossible, mixture models can predict responses to the binary mixtures based on the single-toxicant response curves. In most cases, the mode of action of the toxicants does not determine the optimal choice of model (i.e., CA, RA, or a deviation thereof).

Modulation of CYP1A1 activity by a Ginkgo biloba extract in the human intestinal Caco-2 cells

Ginkgo biloba is a widely consumed dietary supplement. Some dietary active compounds modulate the activity of biotransformation enzymes inside the enterocytes and more interestingly of cytochrome P-450 1A1 (CYP1A1). This enzyme is of a particular interest because of its implication in the metabolism of some exogenous pro-carcinogens or endogenous molecules. In the present work, we have used Caco-2 cells to study the effect of a standard reference material of a Ginkgo biloba extract (GBE) (10-400 μg/ml), as well as of its major individual active compounds (kaempferol, quercetin, isorhamnetin, ginkgolides and bilobalide), alone or in mixtures, at realistic intestinal concentrations, on the induction of CYP1A1 activity, in the presence or absence of benzo[a]pyrene (B[a]P) (0.1 μg/ml), a well-known CYP1A1 inducer. 3-O-rutinosides of kaempferol, quercetin and isorhamnetin were also tested. We have demonstrated a strong induction (p < 0.005) of CYP1A1 activity and a slight, but significant (p < 0.005), decrease of this activity in the presence of B[a]P by the GBE at the realistic exposure level of 100 μg/ml. The inductive effect was explained, in part, by quercetin and kaempferol after 24h exposure while unknown compounds seem to be responsible for the strong CYP1A1 induction observed after 6h exposure. The inhibitory potency of flavonols on CYP1A1 activity in presence of B[a]P was much stronger for the aglycones than for the 3-O-rutinosides, explaining the slight effect observed with the GBE, mainly composed of glycosylated flavonoids. These results indicate that GBEs may disturb intestinal CYP1A1 activity and, in turn, affect the metabolism of other compounds. The present paper thus highlights the necessity to take these side effects into account when administrating Ginkgo biloba herbal supplements.

Bacterial gene profiling assay applied as an alternative method for mode of action classification: Pilot study using chlorinated anilines

Polar narcotic structural analogues (e.g., chlorinated anilines with a differing degree of chlorosubstitution, such as aniline, 4-chloroaniline, 3,5-dichloroaniline, and 2,3,4-trichloroaniline) are assumed to induce their toxic effects via the same predominant mode of action (MOA; membrane damage) at equitoxic exposure concentrations. In this study, a bacterial gene profiling assay consisting of 14 general stress genes was used to test this hypothesis for these four compounds. Although we found a consistent induction of membrane damage, the response cascade and the extent of the response differed among the different chemical treatments. The higher chlorosubstituted anilines also triggered significantly more genes involved in other general stress MOA classes (oxidative stress and protein perturbation). These findings illustrate that, along with the commonly used physicochemistry-based MOA categorization methods, alternative tests such as the bacterial gene profiling assay can yield valuable biological information on the MOA of a certain chemical or group of chemicals that is crucial in high-quality environmental risk assessment. In a second phase, the experimental gene profiling data sets of the chlorinated anilines were analyzed and weighed against existing data on other polar and non polar narcotic compounds to obtain a broader comparison in which the predefined chemical MOAs (narcosis and polar narcosis) were contrasted with the biological MOAs (gene expression profiles). Although additional optimization of the assay is needed, our results show that the bacterial gene profiling assay opens new perspectives for biology-based chemical grouping, thereby further enabling targeted MOA-based risk assessment.

Potential of an in vitro toolbox combined with exposure data as a first step for the risk assessment of dietary chemical contaminants.

In vitro risk assessment of dietary contaminants has become a priority in human food safety. This paper proposes an in vitro approach associating different complementary tools in an original toolbox and aims to improve the assessment of the toxicological impact of dietary contaminants at realistic human exposure levels, with a special focus on the intestinal compartment. The system is based on the use of four complementary cellular tools, namely stress gene induction in transgenic strains of Escherichia coli, modulation of the activity of key biotransformation enzymes (cytochrome P-450 (CYP) 1A1 and 3A4) in a human intestinal cell line, and activation of aryl hydrocarbon receptor (AhR) and oestrogenic receptor (ER)-dependent genes in agonistic and antagonistic assays with luciferase reporter cells. It was applied to four chosen model molecules: ochratoxin A (OTA) and deoxynivalenol (DON), two common food-borne mycotoxins, and imazalil (IMA) and benomyl (BEN), two fungicides widely occurring in foodstuffs. All these assays were performed at or around a realistic intestinal concentration, determined through a deterministic approach based on the calculation of a theoretical maximum daily intake (TMDI). Using the four model molecules, it is clearly highlighted that induction of CYP1A1 activity and inhibition of CYP3A4 activity occurred in Caco-2 cells at a realistic intestinal concentration of IMA. Furthermore, some bacterial stress genes were induced in a range of realistic concentrations, following exposure to DON and IMA. In addition, BEN clearly provoked an ER agonistic activity in a human oestrogen sensitive reporter cell line. All these results are in accordance with the literature, suggesting that the in vitro toolbox constitutes an interesting approach in order to obtain a first ‘fingerprint’ of dietary contaminants at realistic human exposure for further risk assessment.