Papa, E and P Gramatica. 2010. QSPR as a support for the EU REACH regulation and rational design of environmentally safer chemicals: PBT identification from molecular structure. Green Chemistry 12:836-843.
A new computer model may be a significant step forward in predicting the cumulative environmental risk of new and existing chemicals, say researchers who developed it. The model uses a compound’s chemical structure to classify its toxicity, persistence and bioaccumulation – three major traits regulators use to flag a chemical’s potential environmental hazards.
The benefits are significant: working computer models could minimize the need for animal testing, identify highly hazardous compounds already in use and tag the most harmful ones before they are manufactured or introduced to the market.
Globally, thousands of new chemicals are produced each year. The United States alone introduces between 700 and 800. New regulations – such as the Toxic Substances Control Act (ToSCA) in the United States and the Registration, Evaluation, Authorization and Restriction of Chemicals (REACH) in Europe – will require toxicity and persistence testing for all new chemicals. However, given the high cost of standard animal toxicity testing, companies may be unable to comply. In addition, current tests can miss some crucial human health risks, such as endocrine activity and fetal toxicity. Additional costly and time-consuming animal studies need to be performed to test for these effects.
The ongoing research into alternative, computer-based testing spans decades. Regulatory agencies in the United States, Canada, Europe and Asia have made progress developing models, but the scope is usually limited. Most models can accurately predict a particular hazard for only a small class of structurally-related chemicals.
The new model developed by researchers at Italy’s University of Insubria may go beyond a one-hazard, one-chemical type of approach. The researchers combined information from 180 organic chemicals – including some notorious environmental pollutants such as dioxins, PCBs and polyaromatic hydrocarbons (PAHs) – to build a robust computational model. This model ranked the compounds based on their cumulative harm, such that the highest ranked chemicals would have the greatest potential to be toxic, cumulative and persistent.
According to the paper, the model successfully predicted the simultaneous persistence, bioaccumulation and toxicity (PBT) behavior of chemicals based on their chemical structure. The model is applicable to a wide variety of chemicals but was based on traditional, organic-type compounds. It could act as a qualifying tool and may be useful in helping chemists design less hazardous industrial chemicals and materials
Some limitations exist, however, to this approach. The model might not work as well for new, distinct chemicals designed with different chemical properties from those used to build the model. This is because models are based on what is known about how certain chemicals behave in people and the environment. Little information may be known about how new types of chemicals behave, making good hazard predictions challenging. For example, nanomaterials are used in numerous innovative products – such as self-cleaning fabrics, ultra-strong alloys and bacteria-resistant surfaces. But, their toxicity is puzzling and does not always conform to toxicity or environmental behavior patterns that could be predicted by their chemical structure.