Air Date: Week of March 30, 2012

Nuclei are blue and a protein called kinase C involved in cell division and tumor genesis is green in this image made using Vala Sciences software.

Advances in computing are dramatically changing the field of chemical testing. And, as Living on Earth’s Ingrid Lobet reports, one day there may be far less need for scientists to use laboratory animals to test chemicals for health and safety.

Transcript

GELLERMAN: It's Living on Earth, I'm Bruce Gellerman.When scientist test things in vivo, they use living organisms. When they test in vitro, they use test tubes and culture dishes. Now comes the new technique scientists call in silica. It’s just one of the new ways researchers are using computers to find out more about chemicals. Living on Earth’s Ingrid Lobet reports.

[SOUND OF AUTOMATION]

LOBET: At Vala Sciences in San Diego, a machine takes pictures of cells that have undergone experiments where they’re exposed to different chemicals. But no one will ever look at these pictures. Instead, they’ll be analyzed by a computer. James Evans is director of automated technology at Vala.

EVANS: We get much higher levels of information from each experiment we do because we collect images and then those images are quantified and we extract lots of numbers from those images.

LOBET: This machine acts as microscope, camera and data-cruncher, all in one. And it’s one example of the way computers are yielding immense amounts of useful scientific information about chemicals.

EVANS: These systems are collecting images 24-7. So we end up with terabytes and terabytes of images.

LOBET: The Environmental Protection Agency recognized the advances in computing and created a National Center for Computational Toxicology. Right away founding director Bob Kavlock found there was catch-up to do. The data from many important older studies done on animals were in pdf format or on microfiche, not searchable.

KAVLOCK: We sent someone up there to Washington for three weeks and they just scrubbed the file cabinets of data. We spent the next two years extracting that into a database we could use. We wound up inputting about 6,000 reports.

LOBET: Now those studies can be searched. And that enables a next step: computers can use that base of data to begin to predict how chemical compounds will work.

KAVLOCK: What we basically do now is we study what chemicals do and we have a whole series of “whats” that we’re concerned about. We’re concerned do they cause birth defects, do they cause nervous systems damage, do they cause effects on reproduction? Do they cause cancer? So we get pretty good at knowing what the chemical might do, but we don’t know how it does it.

LOBET: Researchers have been focusing on the how, discovering the sequences of reactions chemicals set in motion within the body. As scientists understand these sequences or webs better, they are building computer models, simulations, of how cells work, and actually begin to do experiments in the computer, or ‘in silica.’ Researchers are beginning to build virtual tissues.

KAVLOCK: And so in our research program here we’ve got the virtual liver which is trying to develop a model of the functional liver and we have one on the virtual embryo which is trying to look at the developing processes that happen during fetal development.

ANASTAS: It is providing just treasure troves of data, more rapidly and in greater quantity than ever imagined before.

LOBET: Paul Anastas is considered the father of green chemistry and teaches at Yale. He’s also enthusiastic about these advances in what’s called computational toxicology or predictive toxicology.

ANASTAS: If you understand the molecular basis of hazard, if you understand what the fundamental chemical properties are that make a substance bioavailable, made it be able to get into your body, if you understand the molecular mechanisms that allow it to have a toxic effect, how it bound to particular receptor, then that is giving you rules on how you design a molecule so that it can't get into your body, so it can't cross particular biological membranes, so that it can't bind to particular receptor and therefore make it less toxic.

LOBET: Chemists are integrating this predictive ability into their design of new chemicals. But there’s still a gaping hole in knowledge of how existing chemicals work. An old pesticide may be known to cause nerve damage. But we often don’t know whether it upsets subtle balances in immunity. Here, faster testing, aided by computers, offers hope.

[SOUND OF MACHINE]

LOBET: Back at Vala Sciences in San Diego a robotic arm deposits tiny droplets of a chemical into minute compartments, one millimeter square. In the compartments are human fat cells saved from someone who underwent liposuction. The fat cells are then exposed to a chemical. What is this chemical? Pat McDonough, VP of Biology can’t tell you that.

McDONOUGH: The EPA gives us compounds but they don’t tell us what they are. We can’t bias our results in any way because we don’t know what the compounds are expected to do.

LOBET: The machine is able to simultaneously test the mystery chemical on the fat cells at six different concentrations. Three hundred eighty four experiments every half hour.

McDONOUGH: We expose the cells to hundreds of different compounds in different samples.

LOBET: After the cells are exposed and possibly changed by the chemical, an imager takes pictures of what’s happening in every compartment. There are images that show only the fat cells’ nuclei, or proteins holding the cells together. Again Pat Mcdonough.

McDONOUGH: We can also look for lipid droplets, that correspond to obesity and see whether our test compounds are altering those.

LOBET: Tests like these are shedding new light on chemicals, especially pesticides, that have been with us for years. Animal testing is still considered the gold standard. But computers are digesting data from animal studies, cell studies and digital research with unprecedented efficiency, creating the ability to predict what compounds, old and new, shouldn’t be on the market.

[SOUNDS OF AUTOMATION]

LOBET: For Living On Earth, I’m Ingrid Lobet.

 

Links

National Center for Computational Toxicology (EPA)

American Society for Cellular and Computational Toxicology

Environmental Bioinformatics and Computational Toxicology Center

High Throughput Screening company in San Diego