A research team at the University of Illinois Urbana-Champaign has used advanced computer simulations to better understand why some synthetic cannabinoids produce severe side effects, while others with similar structures do not. Led by Diwakar Shukla, professor of chemical and biomolecular engineering, the study combines deep learning, large-scale molecular simulations, and distributed computing to examine how different cannabinoid-like compounds interact with receptors in the human brain. The findings suggest new strategies for designing safer cannabinoid-based medicines.
Dutta, S., & Shukla, D. (2025). Characterization of binding kinetics and intracellular signaling of new psychoactive substances targeting cannabinoid receptor using transition-based reweighting method. ELife, 13. https://doi.org/10.7554/eLife.98798.3
Synthetic cannabinoids fall under a broader category known as new psychoactive substances, many of which were initially explored as potential pain treatments before being abandoned due to safety concerns. These compounds are now more commonly associated with illicit drugs sold under names such as Fubinaca, Chimica, and Pinaca. Although they are designed to mimic classical cannabinoids, the Illinois team found that their behavior at the molecular level differs in important ways that help explain their stronger and often harmful effects.
Diwakar Shukla from University of Illinois Urbana-Champaign stated,
“The largest class of NPS are often sold as the street drugs Fubinaca, Chimica and Pinaca. In addition to the adverse side effects, the formulas used to produce NPS vary, making them challenging to detect in standard drug screenings.”
At the center of the study is the cannabinoid receptor, a protein embedded in brain cell membranes that helps regulate pain, mood, and perception. Classical cannabinoids tend to activate the receptor through a signaling route known as the G protein pathway, which is generally associated with therapeutic effects. In contrast, many synthetic cannabinoids preferentially trigger an alternative route called the beta arrestin pathway, which has been linked to more intense psychological and physiological responses.
One challenge in studying these compounds is that synthetic cannabinoids bind very tightly to the receptor and detach slowly. These binding and unbinding events are rare on the timescales accessible to standard laboratory experiments or conventional computer simulations. As a result, the processes that may be responsible for harmful side effects have been difficult to observe directly.
To overcome this limitation, the researchers applied a computational approach known as the Transition-Based Reweighting Method. This technique allows scientists to estimate both the thermodynamics and kinetics of slow molecular events by efficiently sampling transitions that would otherwise require enormous amounts of computing time. Graduate researcher Soumajit Dutta applied this method to model how synthetic cannabinoids disengage from the receptor, revealing differences in binding strength and duration compared to classical cannabinoids.
The work was further supported by the Folding@Home platform, which enables volunteers around the world to contribute spare computing power. By running thousands of simulations in parallel and combining the results, the team was able to capture long-timescale molecular behavior that would be impractical on a single computer system. Machine learning algorithms were then used to determine which simulations would provide the most useful additional data.
Together, these tools provided a clearer picture of how subtle structural features in synthetic cannabinoids influence receptor signaling. The simulations showed that compounds associated with stronger beta arrestin signaling tend to remain bound to the receptor for longer periods, increasing the likelihood of adverse effects. This insight points toward specific molecular characteristics that drug designers could avoid when developing new cannabinoid-based therapies.
Rather than dismissing synthetic cannabinoids entirely, the researchers argue that understanding their interaction mechanisms opens the door to safer alternatives. By designing molecules that bind less tightly, unbind more readily, or favor the G protein pathway over beta arrestin signaling, it may be possible to retain therapeutic benefits while reducing risk. The study demonstrates how modern computational methods can extend beyond basic modeling to inform practical decisions in drug design, especially for complex biological systems where experiments alone fall short.
Adrian graduated with a Masters Degree (1st Class Honours) in Chemical Engineering from Chester University along with Harris. His master’s research aimed to develop a standardadised clean water oxygenation transfer procedure to test bubble diffusers that are currently used in the wastewater industry commercial market. He has also undergone placments in both US and China primarely focused within the R&D department and is an associate member of the Institute of Chemical Engineers (IChemE).
