30
Z. Wang and J. Chen
After all, computational toxicology is only an emerging interdisciplinary field. It
is not in its nature obliged to decode the mystery of life. Therefore, development of
computational toxicology would definitely rely on the advances of exterior fields,
especially on the thrilling breakthrough of life sciences.
2.4.3 The Everlasting List of Interlinked Chemicals
as Mixtures
It should be noted that some chemicals will always be interlinked with certain other
chemicals, such as impurities, plasticizers, cosolvents, etc., particularly in industrial
products or in pharmaceutical or cosmetic formulations. It should also be noticed that
a CAS- or REACH-registered chemical could have multiple reaction products that
could be generated during major environmental processes such as photolysis, hydrolysis, and transformation by reactive oxygen species or by biochemical metabolism.
These latent products are not necessarily registered chemicals, and they could be
more hazardous to human and ecological health than their parent compounds. For
example, polybrominated diphenyl esters (PBDEs) can possibly be transformed into
HO-PBDEs or even notorious dioxins with the help of P450 enzymes [90, 91]. Evidences from in vitro assay have proved that the HO-PBDEs are more efficacious on
disrupting thyroid receptor β than the parent PBDEs [92].
No matter how interlinked the chemicals are, they always appear as mixtures in the
real environment. In an inorganic environment, mixtures of chemicals could result in
phenomena such as catalysis, inter-reaction, adsorption, etc. Experimentalists who
study the environmental fate of a queried chemical, usually adopt so-called central
composite design to probe the influence of environmental factors on the chemical
[93, 94]. Note that some of the environmental factors such as anions and dissolved
organic matters are also components of the mixture. If the queried chemical is readily
reactive, then spectra of downstream chemicals of associated reactions would also
contribute to the composition of the mixture. Theoretically, these central composite
designs or full-factorial experiments might characterize a response surface or hypersurface that quantitatively reflects influences from components of the mixture. In
biological systems, similar experiments could be conducted for mixtures in order
to determine so-called joint toxic effects or combined effects of chemicals [95].
The most vital defect of the response hypersurface obtained via wet experiments
is that it just describes the phenotypic response or apparent phenomena and cannot
reveal underlying mechanisms. Therefore, in its nature, the response hypersurface
can never be used to predict quantitative impact of a chemical that is not among the
original components/factors during the experimental determination of the response
hypersurface. In addition, scale of these joint effect-determining experiments would
grow exponentially as the number of factors increases. It seems that, mechanismbased models might be the only promising tools to predict the joint toxic effects of
mixtures with arbitrary composition. Nonetheless, the response hypersurface indeed
Z. Wang and J. Chen
After all, computational toxicology is only an emerging interdisciplinary field. It
is not in its nature obliged to decode the mystery of life. Therefore, development of
computational toxicology would definitely rely on the advances of exterior fields,
especially on the thrilling breakthrough of life sciences.
2.4.3 The Everlasting List of Interlinked Chemicals
as Mixtures
It should be noted that some chemicals will always be interlinked with certain other
chemicals, such as impurities, plasticizers, cosolvents, etc., particularly in industrial
products or in pharmaceutical or cosmetic formulations. It should also be noticed that
a CAS- or REACH-registered chemical could have multiple reaction products that
could be generated during major environmental processes such as photolysis, hydrolysis, and transformation by reactive oxygen species or by biochemical metabolism.
These latent products are not necessarily registered chemicals, and they could be
more hazardous to human and ecological health than their parent compounds. For
example, polybrominated diphenyl esters (PBDEs) can possibly be transformed into
HO-PBDEs or even notorious dioxins with the help of P450 enzymes [90, 91]. Evidences from in vitro assay have proved that the HO-PBDEs are more efficacious on
disrupting thyroid receptor β than the parent PBDEs [92].
No matter how interlinked the chemicals are, they always appear as mixtures in the
real environment. In an inorganic environment, mixtures of chemicals could result in
phenomena such as catalysis, inter-reaction, adsorption, etc. Experimentalists who
study the environmental fate of a queried chemical, usually adopt so-called central
composite design to probe the influence of environmental factors on the chemical
[93, 94]. Note that some of the environmental factors such as anions and dissolved
organic matters are also components of the mixture. If the queried chemical is readily
reactive, then spectra of downstream chemicals of associated reactions would also
contribute to the composition of the mixture. Theoretically, these central composite
designs or full-factorial experiments might characterize a response surface or hypersurface that quantitatively reflects influences from components of the mixture. In
biological systems, similar experiments could be conducted for mixtures in order
to determine so-called joint toxic effects or combined effects of chemicals [95].
The most vital defect of the response hypersurface obtained via wet experiments
is that it just describes the phenotypic response or apparent phenomena and cannot
reveal underlying mechanisms. Therefore, in its nature, the response hypersurface
can never be used to predict quantitative impact of a chemical that is not among the
original components/factors during the experimental determination of the response
hypersurface. In addition, scale of these joint effect-determining experiments would
grow exponentially as the number of factors increases. It seems that, mechanismbased models might be the only promising tools to predict the joint toxic effects of
mixtures with arbitrary composition. Nonetheless, the response hypersurface indeed