octcovTo be an effective drug, a compound must possess the proper characteristics when it comes to how it is absorbed, distributed, metabolized, and excreted—an area of pharmaceutical research known simply as ADME, and often combined with toxicology as ADME/Tox. With in silico–-based testing, researchers can quickly and economically study the properties of existing or even theoretical compounds, but the results get less reliable when dealing with structures beyond those used to train the software. For cell-based approaches, the results might carry more weight about what lies ahead in animal and human testing, but getting those results can cost more, take longer, and a scientist needs the compound of interest in hand.

The leaders in pharmaceutical research take advantage of both approaches. "It isn’t an either-or situation," says Lisa Shipley, PhD, vice president of pharmacokinetics, pharmacodynamics, and drug metabolism at Merck & Co. (Whitehouse Station, NJ). "They need to be used in tandem."

When an ADME issue arises with a discovery compound, Merck uses in silico techniques. "These broaden your thinking about a molecule, and you can weed out bad aspects or dial in good ones before synthesizing anything," says Shipley.

When the in silico approaches produce promising candidates, Merck scientists move to wet-lab tools. "Then, we test in cell-based technologies and ultimately in vivo to confirm the predictions of the in silico models," Shipley says.

Complementary components
Other large pharmas also turn to a combination of ADME approaches. "in silico tools complement in vitro and in vivo testing throughout the drug discovery and drug development continuum," says Franz Schuler, PhD, DABT, head of discovery ADME project support at Roche (Basel, Switzerland). For example, he notes that Roche scientists use in silico tools to—among other things—build structure activity relationships (SARs) or to interpret results, "such as the elaboration of pharmacokinetic-pharmacodynamic relationships."

Nonetheless, Schuler adds that "these in silico tools have their own strengths and weaknesses and often depend on experimental data including results from cell-based assays." For example, he says that the predictive power of applying SARs "is often qualitative in nature" and it does "not replace the actual in vitro testing of synthesized compounds."

The findings from in vitro assays, though, are fed back into the in silico models to refine them. As Schuler says, "in silico, in vitro, and in vivo data are used in a complementary way to maximize the value of experiments."

Simulating a suite


click to enlarge

Scientists at Merck & Co. combine in silico and wet-lab techniques to take advantage of both in ADME/Tox. (Source: Merck & Co.)

In describing the power of in silicomethods for ADME, Pranas Japertas, PhD, product manager of PhysChem/ADMET product development at ACD/Labs (Toronto, Canada), says, "When a compound resembles those used to train the ACD/ADME Suite, the software provides very accurate results." It’s also easy to use. "You just draw the structure and then you get the results," says Japertas.

In addition to results that predict a compound’s ADME traits, the user also gets a reliability index. Japertas says that the index indicates the credibility of the predictions. In cases where a structure is similar to ones used to train the software and similar compounds create consistent data, the reliability is high. When no similar compounds were used in training the software or similar compounds produce erratic results, the reliability is low.

Beyond looking for ADME properties of novel compounds, a researcher could also use this software to tweak the characteristics of existing drugs. For example, the drug’s structure can be altered in the software and the ADME properties reanalyzed. "We’re trying to develop models that suggest how to change a compound to generate a better ADME profile," Japertas says.

Going for the KO
To look for a compound’s potential interactions with drug transporters, Sigma-Aldrich (St. Louis) has developed Transporter Knockout Cell-Based Assays. As described by Michael Dale Mitchell, product manager at Sigma-Aldrich, "Using our CompoZr zinc-finger nuclease technology, we created functional knockouts of three key efflux transporters in intestinal cell lines." The transporters are the multidrug-resistant protein (MDR1), breast cancer–resistant protein (BCRP), and multidrug resistance–-associated protein (MRP2). "The transporters are knocked out individually and in combination," says Mitchell.

Mitchell expects that drug makers will be able to compare a compound’s impact on efflux in parental cells with fully functional transporters versus Sigma’s transporter knockout cell lines. "These assays will allow explicit identification of drug-transporter interactions without reliance on nonspecific substrates and/or inhibitors," he says.

For now, this technology remains in development, but Derek Douglas, product manager at Sigma-Aldrich, says that "assays should be available before the end of the year."

Hepatic help

According to the U.S. Food and Drug Administration (FDA) website: "In the United States, drug-induced liver injury (DILI) is now the leading cause of acute liver failure … exceeding all other causes combined."


click to enlarge
The ToxInsight DILI Assay Cartridge and ToxInsight IVT platform from Thermo Fisher Scientific detect drug-induced hepatotoxicity. These images show stained cells treated with rosiglitazone (left), which is not hepatotoxic, and troglitazone (right), which is hepatoxic. (Source: Thermo Fisher Scientific)  

Consequently, drug makers always seek easier and more effective assays for hepatotoxicity, such as the Thermo Scientific ToxInsight DILI Assay Cartridge from Thermo Fisher Scientific (Waltham, Mass.). Mark Collins, PhD, director of global marketing for their Cellomics business unit, describes this device as "the power of animal testing in a cell-based assay." Developed to work with the ToxInsight IVT automated image cytometry platform, this cartridge detects drug-induced hepatotoxicity.

This 96-well assay looks at five biomarkers, including simple cell death, mitochondrial potential, glutathione depletion, reactive-oxygen species and others. Working with a range of cells, the results reveal a drug’s toxicity index, which Collins describes as "a ranking of how liver liver-toxic your compound is with a rough idea of where the toxicity is most impactful." As a result, it gives an initial indication of the mechanism of action behind the toxicity.

Another option comes from the HepatoPac, a microliver platform made by Medford, Mass.-based Hepregen. Bernadette Fendrock, Hepregen’s president and CEO, describes this as "an in vitro, long-term, predictive, microengineered liver model." She adds that it comes in 24- and 96-well formats, including models for human and rat, andas well as monkey, mouse, and dog models, which are in development.

In addition, the HepatoPac serves as a chronic model. "It’s functional for several weeks, as compared to days with current in vitro hepatocyte technology," Fendrock says. "Over those weeks, you have a physiological liver model, instead of a dying cell culture." Consequently, says Fendrock, the HepatoPac is more effective. "Conventional in vitro technology catches 30 to 50% of the hepatotoxic compounds, but HepatoPac finds about 80% of them, largely due to the ability to look at the compounds over time and to study the metabolism and toxicity."

Customizing a balance
When it comes to selecting products for in silico or cell-based approaches, Shipley says, "There’s no reason to reinvent the wheel." So researchers at Merck use off-the-shelf products when they meet the company’s needs. Still, Shipley says that Merck scientists "must be able to customize an in silico tool and add our own data."

For cell-based approaches, Shipley says, "There are some very good products." Still, she says, "You often have a specific problem where you need to tweak an available assay or develop a new one."

Overall, getting the most from today’s ADME/Tox tools comes from finding the best balance of simulations and cells, custom and commercial. Then, a drug maker can optimize a compound’s characteristics faster and more economically.

About the Author
Mike May is a publishing consultant for science and technology based in Houston, Texas.