Newsletter #2 June 2008

ATHON International Open Workshop on Mixture Toxicity at IVM

cbOn April 1st, Dr. Timo Hamers and Dr. Heather Leslie of the Chemistry & Biology Department of IVM organized an open workshop on mixture toxicity. The workshop was part of the annual meeting of the ATHON project, an EU-sponsored FP6 program on the toxicity and hazard of non-dioxin-like polychlorinated biphenyls (NDL-PCBs) in food, in which IVM is a partner.

The ATHON project ( focuses on a better toxicological characterization of NDL-PCBs. PCBs are prevalent contaminants in fatty food of animal origin such as meat, certain fish and diary products. Unlike dioxin-like PCBs, no health based guidance value for human exposure has been established for NDL-PCBs because their toxicity is insufficiently understood. Since NDL-PCBs constitute a major part of the PCBs found in food and human tissues, several regulatory advisory boards have recommended that more information be gathered about the toxicity of NDL-PCBs. ATHON focuses on the effects of NDL-PCBs on neurobehaviour, reproduction, foetal development, immunology, the endocrine system, liver functioning, and tumour promotion.
The ATHON consortium consists of scientists from 16 different research institutes in Europe and Canada who gather for annual meetings. It was an honour for IVM to host the third annual meeting of 2008. Traditionally, the annual meeting is preceded by a thematic workshop open to both ATHON and non-ATHON participants.
The common theme of interest for this year’s open workshop was “mixture toxicity”, which is a subject that many toxicologists need to deal with. After all, contaminants in food or in any other environmental matrix are rarely present alone, but typically occur together in a complex mixture of contaminants. Risk assessment of contamination, however, is traditionally performed on a compound-by-compound basis, which does not take into account that the effects of compounds with a similar mode of toxic action can be summed up, or that one compound may affect the toxicity of another compound.


The open workshop was chaired by IVM’s Dr. Juliette Legler and consisted of two parts: a plenary session consisting of six keynote presentations and an interactive session in which the participants divided into three groups each discussing another topic within the issue of mixture toxicity. In total, 45 scientists participated in the open workshop of which about half were ATHON participants.
The plenary session was opened by the coordinator of ATHON – Dr. Helen Håkansson of the Karolinska Institutet in Sweden – who gave a brief introduction on the ATHON project and the relevance of mixture toxicity for NDL-PCB risk assessment. Next, Dr. Thomas Backhaus (University of Gothenburg, Sweden), gave an introduction to the general principles of “concentration addition” (CA) and “independent action” (IA) within mixture toxicity, which was illustrated with an abundance of practical examples. The general conclusion from his presentation was that for hazard and risk assessment, the selection between a CA or IA model seems to be of minor importance. Keeping in mind the adage “seek simplicity but distrust it”, Dr. Backhaus advocated that CA might be a simple default assumption to start with and to work further from. The third plenary speaker was Dr. Niklas Johansson (Swedish EPA and Karolinska Institutet) who demonstrated the use of a Central Composite, Face centered (CCF) statistical test design. Results from CCF-designed tests are very useful for multivariate statistical analyses (such as PCA, PLS) in mixture toxicity. Since the CCF design does not include a non-exposed control group, however, regulatory acceptance of CCF-designed tests may be problematic.
After these theoretical talks with examples, the plenary session continued with an in vitro and an in vivo case study. Prof. Dr. Albert Duschl (University of Salzburg, Austria) showed an impressive set of binary mixture toxicity experiments in five different in vitro reporter gene bioassays for immunotoxicity. Many test compounds were able to suppress TNFalpha induced activation of cytokine promoter constructs. In some cases binary combinations of compounds provoked a suppression of the immuno-response, whereas the individual compounds did not, suggesting that compounds acted synergistically. Dr. Wayne Bowers (Health Canada) presented results from an in vivo study with rats exposed to a reconstituted mixture reflecting the composition of persistent organic pollutants (POPs) as found in human plasma samples. One of the important results from his studies is that levels of individual compounds in tissues differ for single compound exposures and mixture exposure. For instance, co-exposure to POPs lead to lower tissue levels of methyl-mercury (MeHg) than expected based on a single exposure to MeHg. Consequently, reduced exposure to other POPs leads to increased levels of MeHg in tissue.
Dr. Ad Ragas from Radboud University Nijmegen, The Netherlands presented an overview of risk assessment procedures for mixtures. Dr. Ragas concluded that guidance is needed for risk assessment of mixtures, requiring reconsideration of current legislation and paradigms. His study identified major challenges for the future, i.e. (1) to identify those cases in which mixture toxicity deviates from the useful CA and IA concepts, (2) to come up with a comprehensive classification of “modes of action” underlying the CA and IA concepts, and (3) to develop concepts that account for mixture interaction not only at the level of toxic effect, but also at the level of absorption, distribution, and metabolism.
In the interactive session, three topics related to mixture toxicity were addressed in three different break-out sessions:

  1. Can we use an indicator molecule for risk assessment of mixtures of NDL-PCBs or should we aim   for a concept of Toxicity Equivalency Factors (TEF)?

  2. What should (new) biomarkers for mixtures containing NDL-PCBs look like?

  3. What are the future challenges in mixture toxicity?

The participants of the first break-out group rejected the concept of using an indicator molecule to predict the mixture toxicity of NDL-PCBs because too many questions cannot be answered: Can one molecule represent all others in all relevant matrices? Is each toxic mode of action represented by the same molecule? Is the marker molecule and the composition of the mixture stable in time? Why analyze one PCB when it is just as easy to analyze them all? The participants agreed that a TEF-like approach makes more sense, although toxicological knowledge on NDL-PCB toxicity is currently insufficient to apply such an approach. The challenges for ATHON should be to identify the most important modes of action of NDL-PCBs, to identify the most sensitive endpoints for these modes of action, and to develop sensitive bioassays for these endpoints.
The participants of the second group agreed that biomarkers for NDL-PCBs should represent early signalling responses for later neurotoxic, reprotoxic, and hepatotoxic effects. The responses should be mechanism based, non-invasive, applicable in a battery, quantifiable, and specific. In addition, they should distinguish between dioxin-like and non-dioxin-like PCBs. Finally, the same mode of action should be used for developing in vitro bioassays.
The third interactive session about the major challenges for future mixture toxicity research resulted in ideas and discussion about finding answers to some of the following questions: How should we account for other stressors than chemicals? What is the contribution of chemical mixtures to the overall risk? Can we develop a concept for toxicokinetics of mixtures? What are the biological mechanisms behind mixture effects that deviate from current concepts (‘back to biology’)? How can “omics” be applied to mixture toxicity research? How should we take into account that different endpoints respond differently to the same mixture? How can “time” be treated as a variable in mixture toxicity, especially for compounds with different kinetics?
The large number of thought-provoking questions that were raised and the enthusiasm of the workshop participants in both the plenary and the interactive sessions made clear that mixture toxicity is a hot research topic. General principles for mixture toxicity are being discovered and seem to be applicable for practical use. Nevertheless, the application of mixture toxicity concepts in risk assessment still seems to be in its infancy. For this latter purpose, a paradigm shift is needed that is probably the biggest challenge for future (mixture) toxicity researchers.
Timo Hamers
Juliette Legler
Heather Leslie