June 15, 2008 (Vol. 28, No. 12)
Researchers Explore a Number of Strategies to Better Predict Drug Responses in the Clinic
As drug development costs continue to escalate, determining toxicity and possible interactions with foods, herbal supplements, and other drugs as early as possible is crucial.
Drug response is variable owing to individual differences in the expression of drug-metabolizing enzymes as well as transporters at sites of absorption, elimination, and/or tissue distribution. This variability can result in unpredictable exposure and tissue distribution of drugs, which may be manifest as adverse effects or therapeutic failure.
Kuo-Chi Cheng, Ph.D., a fellow at Schering (www.schering.com), notes that this has always been a chief area of concern. “Historically, enzyme induction was a major issue, and the only way to unearth whether this was happening was by implementing an animal study around the compound of interest,” he says.
The majority of pharmacokinetic drug-drug interactions occur during drug metabolism in two ways. Either one drug (the perpetrator) may inhibit the metabolism of another drug (the victim), or a perpetrator drug may increase the metabolism of a victim drug. The latter is caused by induction of drug-metabolizing enzymes.
Induction of the major drug-metabolizing enzymes in humans is through activation of nuclear receptors. Assessing receptor activation by potential perpetrator compounds can predict whether there is a potential for drug-drug interactions.
Puracyp’s (www.puracyp.com) CSO, Judy Raucy, Ph.D., agrees. “Potential drug-drug interactions are a concern,” she says. “They can result from activation of nuclear receptors leading not only to alterations in the expression of drug-metabolizing enzymes and transporters, but changes in other biochemically important genes.”
She points out that PXR is a regulator of over 60 genes in the rat and more than 23 genes in humans. Therefore, activation of PXR can alter the disposition of foreign (xenobiotics) and endogenous (endobiotics) chemicals. “We’re not just talking about drug-drug interactions, we’re also considering the consequences of altering homeostasis. There isn’t any federal mandate to track that, but screening should be done as soon as drugs are synthesized in order to assess the activation of nuclear receptors,” Dr. Raucy notes.
Recently, new insights have been gained relating to the regulatory mechanisms governing the expression of drug-metabolizing enzymes and transporters by ligand-activated nuclear receptors. These receptors are a class of proteins found within the interior of cells that are responsible for sensing the presence of hormones and certain other molecules.
These receptors work in concert with other proteins to regulate the expression of specific genes, thereby controlling the metabolism, development, and homeostasis of the organism.
PXR, CAR, FXR, LXR, VDR, HNF4a, and AhR form a battery of nuclear receptors that regulate the expression of many important drug-metabolizing enzymes, transporters, lipid, bile acid, and cholesterol-metabolizing enzymes. For example, PXR and CAR are involved in controlling the metabolism and basolateral transport of bile acids; in addition these two receptors contribute to hepatic protection in cholestasis.
Recent evidence also suggests a role for PXR in cholesterol and steroid metabolism, maintaining mineral corticoid levels, and vitamin D regulation of bone homeostasis. Thus, increased activation of PXR or CAR can lead to alterations in concentrations of any of these key regulatory endobiotics. “The technology is becoming more mature,” says Dr. Cheng. “We are just beginning to understand how much induction translates in vivo. But we don’t yet understand the whole picture.”
Nuclear-receptor activation technology has been around for awhile; because of its ability to predict drug-drug interactions, it is a rapidly growing field. It is also useful for directing chemistry programs away from molecular structures that have liabilities that would activate important biotransformation genes, thus causing clinically significant drug-drug interactions.
“Traditionally, activation of nuclear hormone receptors has been left to the late stages of development,” notes Norman Huebert, Ph.D., principal scientist and team leader of ADME at Johnson & Johnson (www.jnj.com). “It wasn’t the showstopper that other variables in the drug discovery process were. Now, most hepatocyte-based assays are getting a better handle on some of the safety issues involved.”
Dr. Huebert’s group started out by looking at whether some of the newer assays, including nuclear receptor assays such as PXR, would be able to predict adverse reactions. “We compared these to classical induction studies with human hepatocytes and found that there was a very good correlation.”
Prioritizing compounds is the main use of cell lines in Dr. Huebert’s group, which works in conjunction with the DMPK and mechanistic tox groups in development. Cell lines are also applied to late lead optimization and have been applied in earlier stages when warranted.
“We are not looking to advance also-ran molecules that didn’t pass solubility muster or have had other problems,” said Dr. Huebert. “We’re using these to make the chemists’ life easier. If they know upfront that there is going to be an issue with metabolism, distribution, solubility, or any other possible problem there might be, chemists can handle this as a known hurdle. In discovery, as opposed to development, chemists are still making choices, and if you can highlight the issues with a chemotype you can save a lot of time and money in the development of the compound.”
Weimin Tang, Ph.D., another principal scientist at Johnson & Johnson, says that using stable cell line-based technology provides several advantages over hepatocytes. Those advantages include reproducible results, unlimited cell supply, and easy adaptation to existing HTS program. The correlation between results from stable cell lines to that of hepatocytes is also satisfactory for most of the objects tested.
Gilead Sciences (www.gilead.com) focuses on antiviral therapies for a variety of targets. According to Bernard Murray, Ph.D., senior scientist, drug metabolism, “We don’t screen everything against everything. Instead, we attack problems in a focused manner. Induction liability is a known issue for some of our chemotypes, and so we screen those most widely. We also use the same technologies for benchmarking compounds in other compound series. In all situations, we need early test systems that are scientifically relevant.”
Dr. Murray likes nuclear receptors for their sheer reproducibility. “It is usually based in one cell line, and responses to the positive controls are consistent. We routinely use human hepatocytes as our second-line screen for drug-drug interaction potential, but, of course, they are vulnerable to donor-to-donor variability,” he adds. “We like working with nuclear receptors because the technology is predictive, relatively economical, and has a quick turnaround. Speed, economy, and reproducibility are all useful in helping us steer away from trouble in the form of clinical drug-drug interactions.”
Of interest to most pharmaceutical companies, Dr. Raucy notes, is screening drugs for PXR activation, which in turn increases the expression of proteins involved in drug metabolism including CYP3A4, CYP2Cs, P-Glycoprotein, and a number of other drug and bile acid transporters. Many companies are also interested in screening for Ah receptor (dioxin-response element: CYP1A) binding and activation. Genes controlled by AhR are primarily involved in xenobiotic metabolism but can also play a role in metabolism of some endobiotics.
Addressing the Problem
As a result of their role in controlling endogenous biochemical pathways, nuclear receptors are of vital importance in screening for adverse drug reactions caused by new chemical entities. Moreover, they have a proven prediction record of drug-drug interactions and thus are important to safety analysis.
There are several methods for assessing nuclear-receptor activation. One technique includes the use of carcinoma cells transiently transfected with the receptor of interest and the response element of a target gene. There is some degree of variability associated with this technique. In addition, there are several ligand-binding techniques that will assess whether a xenobiotic binds to the ligand-binding domain of the nuclear receptor. One of the drawbacks to this technique is data interpretation because not all ligands will produce a functional response.
Puracyp offers cell lines that contain reporter gene constructs to examine receptor activation. “The primary application of the cell lines is to screen for nuclear-receptor activation,” says Dr. Raucy. “The real value of using this technology is to direct the chemistry of the particular therapeutic program away from the liability of activating key nuclear receptors. In fact, the technology is so precise, it can be used to determine structure activity relationships regarding a particular nuclear receptor.”
One can’t have a discussion on drug-drug interactions without addressing cytochrome P450 (CYP). The metabolism and clearance of most drugs are primarily mediated by CYP. Adverse drug-drug interactions frequently occur because of altered P450 activities, explains Jim Cali, Ph.D., senior scientist at Promega (www.promega.com).
“For example, nuclear receptor-mediated induction of a P450 gene by a component of St. John’s Wort reduces efficacy of oral contraceptives by increasing their rate of metabolism.”
One of Promega’s areas of strength, he states, is in bioluminescent luciferase reporter gene assays, which are ideal for studying gene regulation by nuclear receptors. According to Dr. Cali, Promega has also developed luminogenic P450 enzyme assays that use luciferin derivatives as P450 substrates. By combining reporter gene assays with enzyme assays, one can predict P450 induction at both the transcriptional level and at the level of enzyme activity during the earliest stages of drug discovery.
“We have developed three assays for the CYP3A4 enzyme, which accounts for about 30% of the P450 in human liver and oxidizes about half of all drugs that are eliminated after prior metabolism,” explains Dr. Cali. “Luciferin-PFBE is ideal for cell-based measurement of CYP3A gene inductions by drugs. The cell-based assay is rapid and simple and can be performed in a nonlytic mode that leaves cells intact for additional analysis.”
In addition to its utility for measuring induction in cell-based assays each substrate can be used as a probe for in vitro biochemical assays of recombinant CYP3A4 to detect P450 inhibition by test compounds. Dr. Cali points out that running a reporter assay is only doing part of the job. “We get a more complete picture by pairing reporter assays with enzyme assays.”
The value of this approach is illustrated by considering the effects of bergamottin, a component of grapefruit juice, on CYP3A4 expression. Although bergamottin activates CYP3A4 gene transcription, its net effect is to reduce activity, because it is also an inhibitor of the CYP3A4 enzyme. This explains how grapefruit juice consumption slows clearance of CYP3A4-metabolized drugs such as the cholesterol-lowering statins.
“Reporter gene technology paired with P450 enzyme assays is a good way to get a full picture of how certain compounds impact drug disposition through nuclear receptor interactions,” notes Dr. Cali.
It’s a matter of public record that regulatory guidelines focus on the use of human hepatocytes as the gold standard of testing. Some comments appear on the FDA site saying that this needs to change and that the focus should not be solely on human hepatocytes, given the variability of reproducibility from donor to donor.
Dr. Cheng agrees. “We use nuclear receptors as the first line—they definitely have a lot of merit in drug discovery. Those hepatocytes derived from human donors serve as our second-tier assays, only because there is so much variability in the cell lines from donor to donor.”
The FDA acknowledges the evolution of reporter technology, explains Dr. Murray. “They will allow the use of data from nuclear receptors, certainly, but they will also want more information, which will require using human hepatocytes to follow up.” He notes that in dealing with some therapeutic areas, drug-drug interactions may not be quite as big an issue. However, in dealing in his field of antiviral discovery for HIV, “you could get into trouble pretty quickly unless you keep your eyes open.”
Dr. Murray is working with human PXR, CAR AhR and also rat and monkey PXR and also “keeps an eye on what vendors have available. Dog PXR, for example, would be something we’d use, as well. When picking final candidates to nominate for development, we are often screening multiple compounds against every receptor that is available.”
As for overall reproducibility, Dr. Tang emphasizes that the FDA guideline is the standard to follow. “So far, based on our results, we have not seen any significant outliers when comparing human hepatocytes to the Puracyp cell lines. But whoever implements the results of these finds has to understand the potential drawbacks.”
Looking into the Future
Dr. Murray notes that some of the receptors already have well-characterized systems in place—human PXR and AhR in particular. Assays for CAR activation, he notes, are somewhat less mature, “but when assays for all three of these important receptors are in place, we’ll feel more comfortable. The more information we have, the lower the risk when moving forward. I do see receptor-based technology as being strongly predictive of the potential for clinical drug-drug interactions.”