Authors: Michiel Kraak & Kees van Gestel
Reviewer: Thomas Backhaus
Learning objectives:
You should be able to
· explain the concepts involved in mixture toxicity testing, including Concentration Addition and Response Addition.
· design mixture toxicity experiments and to understand how the toxicity of (equitoxic) toxicant mixtures is assessed.
· interpret the results of mixture toxicity experiments and to understand the meaning of Concentration Addition, Response Addition, as well as antagonism and synergism as deviations from Concentration Addition and Response Addition.
Key words: Mixture toxicity, TU summation, equitoxicity, Concentration Addition, Response Addition, Independent Action
Introduction
In contaminated environments, organisms are generally exposed to complex mixtures of toxicants. Hence, there is an urgent need for assessing their joint toxic effects. In theory, there are four classes of joint effects of compounds in a mixture as depicted in Figure 1.
Four classes of joint effects |
No interaction (additive) |
Interaction (non-additive) |
Similar action |
Simple similar action/ Concentration Addition |
Complex similar action |
Dissimilar action |
Independent action/ Response Addition |
Dependent action |
Figure 1. The four classes of joint effects of compounds in a mixture, as proposed by Hewlett and Plackett (1959).
Simple similar action & Concentration Addition
The most simple case concerns compounds that share the same mode of action and do not interact (Figure 1 upper left panel: simple similar action). This holds for compounds acting on the same biological pathway, affecting strictly the same molecular target. Hence, the only difference is the relative potency of the compounds. In this case Concentration Addition is taken as the starting point, following the Toxic Unit (TU) approach. This approach expresses the toxic potency of a chemical as TU, which is calculated for each compound in the mixture as:
with c = the concentration of the compound in the mixture, and ECx = the concentration of the compound where the measured endpoint is affected by X % compared to the non-exposed control. Next, the toxic potency of the mixture is calculated as the sum of the TUs of the individual compounds:
Imagine that the EC50 of compound A is 300 μg.L-1 and that of compound B 60 μg.L-1. In a mixture of A+B 30 μg.L-1 A and 30 μg.L-1 B are added. These concentrations represent 30/300 = 0.1 TU of A and 30/60 = 0.5 TU of B. Hence, the mixture consists of 0.1 + 0.5 = 0.6 TU. Yet, the two compounds in this mixture are not represented at equal toxic strength, since this specific mixture is dominated by compound B. To compose mixtures in which the compounds are represented at equal toxic strength, the equitoxicity concept is applied:
1 Equitoxic TU A+B = 0.5 TU A + 0.5 TU B
1 Equitoxic TU A+B = 150 μg.L-1 A + 30 μg.L-1 B
As in traditional concentration-response relationships, survival or a sublethal endpoint is plotted against the mixture concentration from which the EC50 value and the corresponding 95% confidence limits can be derived (see section on Concentration-response relationships). If the upper and lower 95% confidence limits of the EC50 value of the mixture include 1 TU, the EC50 of the mixture does not differ from 1 TU and the toxicity of the compounds in the mixture is indeed concentration additive (Figure 2).
Figure 2. Concentration-response relationship for a mixture in which the toxicants have a concentration additive effect. The Y-axis shows the performance of the test organisms, e.g. their survival, reproduction or other endpoint measured. The horizontal dotted line represents the 50% effect level, the vertical dotted line represents 1 Toxic Unit (TU). The black dot represents the experimental EC50 value of the mixture with the 95% confidence limits.
An experiment appealing to the imagination was performed by Deneer et al. (1988), who tested a mixture of 50 narcotic compounds (see section on Toxicodynamics and Molecular interactions) and observed perfect concentration addition, even when the individual compounds were present at only 0.25% (0.0025 TU) of their EC50. This showed in particular that narcotic compounds present at concentrations way below their no effect level still contribute to the joint toxicity of a mixture (Deneer et al., 1988). This was also shown for metals (Kraak et al., 1999). This is alarming, since even nowadays environmental legislation is still based on a compound-by-compound approach. The study by Deneer et al. (1988) also clearly demonstrated the logistical challenges of mixture toxicity testing. Since for composing equitoxic mixtures the EC50 values of the individual compounds need to be known, testing an equitoxic mixture of 50 compounds requires 51 toxicity tests: 50 individual compounds and 1 mixture.
Independent Action & Response Addition
When chemicals have a different mode of action, act on different targets, but still contribute to the same biological endpoint, the mixture is expected to behave according to Response Addition (also termed Independent Action; Figure 1, lower left panel). Such a situation would occur, for example, if one compound inhibits photosynthesis, and a second one inhibits DNA-replication, but both inhibit the growth of an exposed algal population. To calculate the effect of a mixture of compounds with different modes of action, Response Addition is applied as follows: The probability that a compound, at the concentration at which it is present in the mixture, exerts a toxic effect (scaled from 0 to 1), differs per compound and the cumulative effect of the mixture is the result of combining these probabilities, according to:
E(mix) = E(A) + E(B) – E(A)E(B)
Where E(mix) is the fraction affected by the mixture, and E(A) and E(B) are the fractions affected by the individual compounds A and B at the concentrations at which they occur in the mixture. In fact, this equation sums the fraction affected by compound A and the fraction affected by compound B at the concentration at which they are present in the mixture, and then corrects for the fact that the fraction already affected by chemical A cannot be affected again by chemical B (or vice versa). The latter part of the equation is needed to account for the fact that the chemicals act independent from each other. This is visualised in Figure 3.
Figure 3. Illustration of stressors acting independent of each other, using the example given by Berenbaum (1981). Subsequently a handful of nails and a handful of pebbles are thrown to a collection of eggs. The nails break 5 eggs, and these 5 eggs broken by the nails cannot be broken again. The pebbles could break 4 eggs, but 1 egg was already broken by the nails. Hence, the pebbles break 3 additional eggs.
The equation: E(mix) = E(A) + E(B) – E(A)E(B)
can be rewritten as: 1-E(mix) = (1-EA)*(1-EB)
This means that the probability of not being affected by the mixture (1-E(mix)) is the product of the probabilities of not being affected by (the specific concentrations of) compound A and compound B. At the EC50, both the affected and the unaffected fraction are 50%, hence (1-EA)*(1-EB) = 0.5. If both compounds equally contribute to the effect of the mixture, (1-EA) = (1-EB) and thus (1-EA or B)2 = 0.5, so both (1-EA) and (1-EB) equal = 0.71. Since the probability of not being affected is 0.71 for compound A and compound B, the probability of being affected is 0.29. Thus at the EC50 of a mixture of two compounds acting according to Independent Action, both compounds should be present at a concentration equalling their EC29.
Interactions between the compounds in a mixture
Concentration Addition as well as Response Addition both assume that the compounds in a mixture do not interact (see Figure 1). However, in reality, such interactions can occur in all four steps of the toxic action of a mixture. The first step concerns chemical and physicochemical interactions. Compounds in the environment may interact, affecting each other’s bioavailability. For instance, excess of Zn causes Cd to be more available in the soil solution as a result of competition for the same binding sites. The second step involves physiological interactions during uptake by an organism, influencing the toxicokinetics of the compounds, for example by competition for uptake sites at the cell membrane. The third step refers to the internal processing of the compounds, e.g. involving effects on each other’s biotransformation or detoxification (toxicokinetics). The fourth step concerns interactions at the target site(s), i.e. the toxicodynamics during the actual intoxication process. The typical whole organism responses that are recorded in many ecotoxicity tests integrate the last three types of interactions, resulting in deviations from the toxicity predictions from Concentration Addition and Response Addition.
Deviations from Concentration Addition
If the EC50 of the mixture is higher than 1 TU and the lower 95% confidence limit is also above 1 TU, the toxicity of the compounds in the mixture is less than concentration additive, as more of the mixture is needed than anticipated to cause 50% effect (Figure 4, blue line; antagonism). Correspondingly, if the EC50 of the mixture is lower than 1 TU and the upper 95% confidence limit is also below 1 TU, the toxicity of the compounds in the mixture is more than concentration additive (Figure 4, red line; synergism).
Figure 4. Concentration-response relationships for mixtures in which the toxicants have a less than concentration additive effect (blue line), a concentration additive effect (black line) and a more than concentration additive effect (red line). The Y-axis shows the performance of the test organisms, e.g. their survival, reproduction or other endpoint measured. The horizontal dotted line represents the 50% effect level, the vertical dotted line represents 1 Toxic Unit (TU). The coloured dots represent the EC50 values with the corresponding 95% confidence limits.
When the toxicity of a mixture is more than concentration additive, the compounds enhance each other’s toxicity. When the toxicity of a mixture is less than concentration additive, the compounds reduce each other’s toxicity. Both types of deviation from additivity can have two different reasons: 1. The compounds have the same mode of action, but do interact (Figure 1, upper right panel: complex similar action). 2. The compounds have different modes of actions (Independent action/Response Addition; Figure 1, lower left panel).
Concentration-response surfaces and isoboles
Elaborating on Figure 4, concentration-response relationships for mixtures can also be presented as multi-dimensional figures, with different axes for the concentration of each of the chemicals included in the mixture (Figure 5A). In case of a mixture of two chemicals, such a dose-response surface can be shown in a two-dimensional plane using isoboles. Figure 5B shows isoboles for a mixture of two chemicals, under different assumptions for interactions according to Concentration Addition. If the interaction between the two compounds decreases the toxicity of the mixture, this is referred to as antagonism (Figure 5B, blue line). If the interaction between the two compounds increases the toxicity of the mixture, this is referred to as synergism (Figure 5B, red line). Thus both antagonism and synergism are terms to describe deviations from Concentration Addition due to interaction between the compounds. Yet, antagonism in relation to Concentration Addition (less than concentration additive; Figure 5B blue line) can simply be caused by the compounds behaving according to Response Addition, and not behaving antagonistically.
Figure 5. A. Dose-response surface showing the effect of chemicals A and B, single (sides of the surface), and in mixtures. B. Isoboles showing the toxicity of the same mixtures in a two-dimensional plane. The isoboles can be seen as a cross section through the dose-response surface. The isoboles show the 50% effect level according to Concentration Addition of mixtures of the two compounds in case they do not interact (black line), when they interact antagonistically (blue line) and when they interact synergistically (red line).
Synergism and antagonism evaluated by both concepts
The use of the terms synergism and antagonism may be problematic, because antagonism in relation to Concentration Addition (less than concentration additive; Figure 5B blue line) can simply be caused by the compounds behaving according to Response Addition, and not behaving antagonistically. Similarly, deviations from Response Addition could also mean that chemicals in the mixture do have the same mode of action, so act additively according to Concentration Addition. One can therefore only conclude on synergism/antagonism if the experimental observations are higher/lower than the predictions by both concepts.
Suggested further reading
Rider, C.V., Simmons, J.E. (2018). Chemical Mixtures and Combined Chemical and Nonchemical Stressors: Exposure, Toxicity, Analysis, and Risk, Springer International Publishing AG. ISBN-13: 978-3319562322.
Bopp, S.K., Kienzler, A., Van der Linden, S., Lamon, L., Paini, A., Parissis, N., Richarz, A.N., Triebe, J., Worth, A. (2016). Review of case studies on the human and environmental risk assessment of chemical mixtures. JRC Technical Reports EUR 27968 EN, European Union, doi:10.2788/272583.
References
Berenbaum, M.C. (1981). Criteria for analysing interactions between biologically active agents. Advances in Cancer Research 35, 269-335.
Deneer, J.W., Sinnige, T.L., Seinen, W., Hermens, J.L.M. (1988). The joint acute toxicity to Daphnia magna of industrial organic chemicals at low concentrations. Aquatic Toxicology 12, 33–38.
Hewlett, P.S., Plackett, R.L. (1959). A unified theory for quantal responses to mixtures of drugs: non-interactive action. Biometrics 15, 691 610.
Kraak, M.H.S., Stuijfzand, S.C., Admiraal, W. (1999). Short-term ecotoxicity of a mixture of five metals to the zebra mussel Dreissena polymorpha. Bulletin of Environmental Contamination and Toxicology 63, 805-812.
Van Gestel, C.A.M., Jonker, M.J., Kammenga, J.E., Laskowski, R., Svendsen, C. (Eds.) (2011). Mixture toxicity. Linking approaches from ecological and human toxicology. SETAC Press, Society of Environmental Toxicology and Chemistry, Pensacola.