Computational and Experimental Models Combine in Science on Saturday Program

LLNL scientists Nick Be (left) and Tim Carpenter describe the computational (in silico) methods and experimental (in vitro and in vivo) techniques used to speed up the therapeutic drug optimization process. The combination of these methods creates a pipeline to help scientists test more therapeutics at a lower cost. (Photo by Joanna Albala/LLNL Science Education Program.)

Lawrence Livermore National Laboratory’s (LLNL’s) educational outreach program Science on Saturday returned in February with a look at neurotechnology and the brain. Held annually at the Bankhead Theater in downtown Livermore, the lecture series offers local students and the public a peek into LLNL’s recent research.

In a February 23 presentation entitled “Crossing the Blood–Brain Barrier: One Byte at a Time,” LLNL scientists Tim Carpenter and Nick Be and Tracy High School teacher Erin McKay talked about the combination of computational and experimental models used at LLNL to determine how therapeutics can cross the blood–brain barrier and effectively treat the brain.

Treating the Brain

Therapeutic drugs typically travel though the bloodstream, accessing target areas through spaces and gaps between cells. However, the blood–brain barrier—a special structure in the body that helps to protect the brain from unwanted toxins and germs—includes proteins that block these gaps, making it extremely difficult for therapeutics to reach their intended target within the brain. Successfully modifying drugs to cross this barrier requires an understanding of the cell membrane structure and how drugs interact with this structure.

Cell membranes include two lipid layers, each with a hydrophilic (water-loving) and hydrophobic (water-hating) region. The two hydrophilic regions face outward, keeping substances on either side of the barrier, while the two hydrophobic regions face inward, repelling or slowing anything that tries to cross the barrier.

In the first of two demonstrations, McKay illustrated hydrophilic and hydrophobic interactions with a container filled with water (dyed blue) and oil (dyed yellow). She explained that oil is hydrophobic, causing a clear separation of water and oil in the container. She then poured another liquid (dyed red) into the container and mixed the contents. Though the water and oil briefly combined, they separated again with the oil on top and the now-purple liquid on the bottom. She pointed out that the red liquid must have been hydrophilic because it combined with the blue water to make purple—and revealed that it was, in fact, water.

If an effective drug is hydrophilic, it is more likely to stay outside the barrier and unlikely to permeate, or cross, the barrier. Therefore, researchers must add hydrophobic components to increase the drug’s permeability. But too many hydrophobic components create interactions with the hydrophobic layers, making it difficult for the drug to reach the other side of the barrier.

The second demonstration illustrated permeability with a race across the auditorium. McKay ran through the front row, giving each person a high five as she passed. Be ran across a larger aisle, and merely waved at each person as he passed. Be had fewer interactions, so he was able to cross the auditorium faster.

Modeling the Blood–Brain Barrier

With these concepts established, Carpenter explained that successful drug design strikes the optimal balance between effectiveness and permeability, but testing drugs to find the right balance can be a long and expensive process. Scientists at LLNL have sped up the process by creating a pipeline of three modeling methods, allowing them to test more drugs at a lower cost.

First, Carpenter, a computational biophysicist, uses some of the fastest supercomputers in the world to test hundreds of drugs using in silico (computational) models. He makes a model of the biological system at the scale of an atom, running quadrillions of calculations to force the atoms to move. He then uses advanced sampling methods to test many drugs and their respective permeability, predicting which are most likely to cross the blood–brain barrier.

Next, Be, an experimental biologist, verifies Carpenter’s predictions using in vitro (cell/assay) models. Be uses artificial lipid membranes to test the permeability of the drugs predicted to have the best results in terms of their ability to diffuse across these experimental membranes. He also tests candidate drugs in an experimental platform composed of brain endothelial cells to determine whether these compounds are capable of crossing the actual cells that compose the blood–brain barrier.

Finally, additional team members test only the drugs with the highest permeability using in vivo (animal) models. This experimental technique is the slowest and most expensive method; however, it is also the most accurate, providing confirmation that the drug can both access the area needing treatment and effectively treat the problem.

Before taking questions, Carpenter and Be explained the need to calibrate the computational models by running drugs with known permeability through in silico and in vitro models and comparing the results. This calibration helps them have confidence that the initial predictions are accurate.

Taking the Stage

For Be and Carpenter, both first-time Science on Saturday presenters, community outreach is important. “It’s a fundamentally important part of our job to communicate clearly and advocate for scientific inquiry in all parts of society,” Be said. “It was exciting to see all the students in the audience, taking time on their Saturday to learn about something that will hopefully excite them.”

Carpenter especially enjoyed the demonstrations and the Q&A sessions. “It’s encouraging to see people understanding the concepts and asking follow-up questions,” he said.

Be, Carpenter, and McKay delivered the talk twice, with both sessions well attended. Carpenter and Be returned to the stage on March 16 at the “Science on Screen” event in Modesto.

Watch previous Science on Saturday lectures on LLNL’s YouTube channel.

—Amanda Lewis