When a toxic material decays in the ground, it seems obvious that it would steadily weaken.
Not necessarily. Some chemical compounds can actually become more toxic while they decay, according to researcher Trevor Sleight. A graph showing the progress of their decay sometimes looks like a bell curve rather than a downward slope – the highest level of toxicity is not at the beginning of the compound’s degradation, but at a point somewhere during the process. Sleight has created a model to better understand what chemical compounds are most likely to degrade into more toxic daughter compounds. His research, done using CRC resources, was published this summer in the journal Environmental Science & Technology.
“Some compounds spend a long time not degrading in soil,” says Sleight, a graduate student in the labs of Carla Ng and Leanne Gilbertson in the Department of Civil and Environmental Engineering. “For this paper, we looked at PAHs – polycyclic aromatic hydrocarbons, chemical compounds that are released during the combustion of organic materials and fossil fuels. As they slowly degrade, they can increase in impact.”
This is a challenge for regulators and remediators, who normally measure the top compound – the original material – rather than the daughter compounds into which the material transforms as it decays. The toxic impact of those daughter compounds is hard to identify or predict – each PAH molecule can potentially degrade into tens of thousands of smaller metabolites, each with its own properties, each following its own pathway of degradation.
Sleight used CRC resources to develop methods based on network theory to better predict which of those thousands of pathways a compound may follow. He began with a molecular data tool provided by the Swiss Federal Institute of Aquatic Science and Technology (EAWAG), which models thousands of possible daughter compounds of a decaying parent compound. The data set produced by the Swiss tool was unmanageably large to analyze with anything available.
Needing computing power and specialized software, Sleight turned to CRC. Network theory rests on the concept of edges and nodes – roughly lines and intersections. In Sleight’s model, nodes represent metabolites produced as a compound degrades, and edges represent biodegradation reactions. Sleight built a network that models the relationships between metabolites and reactions, using a range of CRC resources from high volume chemical informatics software to customizable Python scripts.
The study focused on compounds of four “bad actors” on the EPA’s Priority Pollutants List: acenaphthene, anthracene, fluorene, and phenanthrene. He calculated the number of possible molecular structures in the four compounds based on the EAWAG data and constructed a network of degrading metabolites one layer at a time, combining the layers using Python.
An important part of the analysis was determining which biodegradation reactions associated with one compound are also associated with two or more other compounds, indicating that the first compound likely inﬂuences the pathway other compounds follow while degrading. Sleight’s algorithm highlights compounds and reactions with high likelihood to appear in within the network – and then trims the size of the network to include only compounds of interest.
Sleight trained the network analysis model against a literature survey comprising all available empirical studies on the behavior of individual PAH compounds when degrading in freshwater and soil – a total of 176 studies. Ultimately, the model represents which metabolites of the four PAHs are likely to decay into more toxic compounds, and the paths they would likely follow in decaying.
Gilbertson believes Sleight’s model could represent a scalable tool for regulators and remediators to evaluate thousands of substances.
“Trevor’s research is filling a need that impacts direct monitoring and bio-remediation techniques,” says Gilbertson, assistant professor of civil and environmental engineering. “The standard approach is looking for the disappearance of the parent compound – but the awareness that a compound can transform into something more harmful adds to the full picture of what daughter products to look for. It could also apply to a compound that does not appear in the empirical literature. Without his model, the time to accomplish that would be immense. He’s been able to narrow down the search.”
“Trevor is addressing the persistent challenge of kinetics – understanding the rate of change of compounds– in environmental science,” explains Ng, also assistant professor of civil and environmental engineering. “Modeling is one way to attack that challenge – empirical data only shows points of time in an experiment while modeling lets you run the experiment forever. But Trevor’s approach is unique. He develops methods to predict long-term changes, but he is not running biomolecular simulations – he’s purely getting down into the chemical structure.”
“The network analysis shows that different bacteria strains degrading different compounds can make the environmental problems worse,” says Ng. “If you could know which compounds were mutagenic –– you could avoid the compounds or speed up their degradation. It is a revelation for bioremediation.”
Read the full paper at: https://pubs.acs.org/doi/10.1021/acs.est.0c02217