Abstract
Quantification of nucleic acids has become a common procedure in many toxicology laboratories. Among the technologies that accomplish this is the fluorogenic 5 ′ -nuclease assay, commonly known as TaqMan. Three TaqMan applications for genetic and molecular toxicology are presented in this article: quantification of gene expression, detection of genetic polymorphisms, and quantification of chromosomal DNA deletions. Of these, quantification of gene expression is the most widely used, and established TaqMan as a benchmark technology for nucleic acid quantification. Two additional applications, polymorphism detection and quantification of DNA deletions, demonstrate the flexibility and quantitative strengths that make TaqMan so powerful, including high precision, excellent sensitivity, and broad linear dynamic range. These and similar applications improve our ability to investigate genetic and molecular dimensions of toxicological phenomena, and have promoted the widespread use of TaqMan in toxicology departments in the pharmaceutical industry. In addition to presenting these applications, the authors discuss some of the challenges of integrating TaqMan and other new technologies into the drug development process.
TaqMan is a quantitative real-time polymerase chain reaction (PCR) that uses fluorescent probes to measure amounts of target nucleic acid. A fluorogenic (TaqMan) probe that contains both a reporter dye and a quencher dye anneals to the target of interest. During the extension phase of the PCR cycle, the 5′-exonuclease activity of Taq polymerase cleaves the hybridized fluorogenic probe, causing dissociation of the reporter dye from the quencher, resulting in an increase in detectable fluorescence. The increase in detectable fluorescence is proportional to the amount of PCR product in the reaction.
An amplification plot (Figure 1A ) containing raw data from TaqMan experiments is presented as fluorescence (y-axis) versus PCR cycle number (x-axis). As with conventional PCR, each round of amplification doubles the number of copies of target sequence in the reaction, such that an ∼1000-fold increase in PCR product is produced every 10 rounds of amplification. This process of target amplification results in the production of large amounts of fluorescence from small numbers of target in the original sample. Quantification of very small amounts of genetic material has proven useful in many fields, including those in which sample is limited in abundance (e.g., forensics), and when nucleic acids are at low but significant levels (e.g., food-borne pathogens).
TOXICOGENOMICS: QUANTIFICATION OF GENE EXPRESSION
Gene expression analysis has become a widely used approach to identify biomarkers and understand mechanisms of toxicity (Searfoss et al. 2003; Bailey and Ulrich 2004; Huang et al. 2004). Among the toxicological processes that can be investigated is cell proliferation, which plays an important role in carcinogenesis. Molecular events underlying cell proliferation include changes in expression of genes such as the cyclins and cyclin-dependent kinases, among others. In one drug development program at Lilly, a set of kinase inhibitors caused cell proliferation in liver, kidney, and adrenals in female F344 rats. TaqMan analysis (Figure 1) was used as part of the experiments that investigated this toxicity.
Cyclin D1 is a transcriptional biomarker of cell proliferation (Juan et al. 1996). Moreover, pathway analysis using the published literature indicated that cyclin D1 would be upregulated as a result of the intended pharmacology of these kinase inhibitors. We therefore measured cyclin D1 transcript levels in the livers of rats at different times following exposure to one of the kinase inhibitors. As expected, we observed a large induction in cyclin D1 transcript level (Table 1). Interestingly, the kinetics of transcriptional activation of cyclin D1 were very similar to the pharmacokinetics of the small molecule kinase inhibitor itself: T max ∼ 20 min; t 1 / 2 <1 h. These and additional data collected for this project supported the hypothesis that the cell proliferation phenomenon was the result of an on-target effect of the kinase inhibitors, and therefore structural modifications to this class of inhibitors was unlikely to produce a molecule that achieved the desired pharmacology without causing the cell proliferation. Lilly’s toxicology organization therefore deprioritized development of this drug target for the intended indication.
Changes in gene expression precede or accompany many toxicologic pathologies, as for the induction of cyclin D1 leading to cell proliferation. Results such as these have encouraged researchers to investigate changes in gene expression to identify biomarkers of toxicity, improve our understanding of mechanisms of toxicity, and predict toxicities that will arise as a result of longer-duration exposures. However, many challenges confront scientists trying to incorporate such technologies into the drug development process. The cyclin D1 example is illustrative of one such challenge: that molecular events can occur on a different time scale than do histological changes that currently drive safety assessments. As stated above, the kinetics of transcriptional activation of cyclin D1 were very similar to the pharmacokinetics of the kinase inhibitor itself (T max ∼ 20 min; t 1 / 2 <1 h). Consequently, we failed to detect an increase in cyclin D1 transcript in our first study, in which cyclin D1 transcript levels had returned to control levels by the time the tissues were sampled: 24 h following the last exposure. An additional study, using a range of time points, revealed this transient but profound molecular event that was integral to the mechanism of multiorgan cell proliferation. In contrast, analysis of tissues 24 h following treatment in a multidose study is usually suitable for histological analysis.
This difference in kinetics between molecular and histological phenomena poses a challenge for new technology deployment in the drug development pipeline: to increase the probability of detecting molecular events that predict or help us understand toxicity will require alterations to standardized toxicology tests, particularly with regards to exposure regimens and tissue collection protocols.
TOXICOGENETICS: DETECTION OF GENETIC POLYMORPHISMS
Determining the best medication and dose for each patient is an overarching goal of modern therapeutics. This process requires knowledge not only of the disease, but also of the patient. Genetics experiments have revealed many genetic polymorphisms that modify both the desirable and undesirable effects of pharmaceutical agents in man. Among the most well documented polymorphisms that modulate drug action are those that affect drug metabolism and clearance (Roden and George 2001). CYP2C9*2, for example, is a biallelic polymorphism that alters a patient’s exposure to the drugs warfarin, phenytoin, and tolbutamide, among others.
Observations such as these have encouraged the development of genotyping methods, including two that use the TaqMan technology: one based on allele-specific probes, the other based on allele-specific primers. Allelic discrimination with probes requires a separate probe for each allele in a polymorphism. Each probe forms a perfect match to the polymorphism it detects and contains a single base mismatch with the allele(s) to be discriminated against (Figure 2A ). Presence of the allele of interest leads to hybridization, cleavage, and signal by the probe specific for that allele. Strong signal from one probe indicates the person is homozygous for that allele. Individuals heterozygous at the locus produce signal from both probes (Figure 2B ).
TaqMAMA is another TaqMan-based genotyping technique. TaqMAMA combines the strengths of TaqMan with the allelic discrimination of the mismatch amplification mutation assay (MAMA; Cha et al. 1993; Glaab and Skopek 1999; Li et al. 2004). Allelic discrimination by MAMA is accomplished using primers that contain two 3′ mismatches between the MAMA primer and the allele to be discriminated against, whereas only a single mismatch occurs with the allele of interest (Figure 3A ). Two mismatches dramatically decrease PCR efficiency, whereas the single mismatch that occurs with the allele of interest has little or no effect on PCR efficiency, and thus strong signal is produced in the presence of the targeted allele (Li et al. 2004).
TaqMAMA genotyping assays have been described for a number of human polymorphisms, including CYP2C9*2 (Figure 3B ). Design of primers for additional TaqMAMA assays has been simplified by publication of a guide for MAMA primer selection (Li et al. 2004). TaqMAMA assays can be developed quickly, conducted relatively inexpensively, and provide a relatively high throughput assay format that will be attractive to many different types of laboratories. The limiting factor for toxicological research in this area, whether using TaqMAMA or one of the many other genotyping techniques, is the lack of availability of samples from patients that have experienced adverse drug reactions. As pointed out in the case of biomarker evaluation, a tissue banking effort would go a long way in increasing our knowledge in this important field of toxicogenetics.
GENETIC TOXICOLOGY: QUANTIFICATION OF CHROMOSOMAL DNA DELETIONS
DNA double-strand breaks (DSBs) are caused by physical and chemical mutagens, and can produce genome rearrangements and/or cytotoxicity. Repair of DSBs occurs by many pathways in eukaryotes, including homologous recombination. The process of homologous recombination requires several hundred base pairs of nearly identical sequence between the broken strand and a homologous partner. The partner strand provides a template for new DNA synthesis that ultimately restores continuity to the broken strand. Structural changes to DNA resulting from homologous recombination include small and large deletions, duplications, and translocations. Each of these mutagenic events is known to contribute to carcinogenesis.
RS112 is a yeast strain created by targeted integration of a plasmid into the HIS3 locus of Saccharomyces cerevisiae, producing two nonfunctional his3 heteroalleles. The upstream his3 allele is 3′ truncated; the downstream allele is 5′ truncated. The alleles are separated by ∼6.8 kb of plasmid DNA. The ∼400 bp homology between the two alleles is a hotspot for interchro-mosomal recombination events that deletes the plasmid DNA and thereby restores a single functional HIS3 allele (Figure 4A ). A broad range of genotoxic agents increases recombinant frequency (RF) at the HIS3 locus of RS112.
RF is ordinarily determined by counting the number of His+ yeast colonies that grow on His− medium after 2 to 3 days. An alternative technique for detecting mutations uses TaqMan to directly detect DNA molecules that have lost the ∼6.8 kb of plasmid DNA (Li, Cise, and Watson 2003). This is accomplished using primers whose annealing sites are outside the region of DNA lost due to recombination (Figure 4A ). These primers are too far apart to produce PCR products using wild-type DNA, but produced a robust TaqMan signal using DNA from recombinants (Figure 4B ). Spontaneous and chemical-induced recombinant frequencies (RFs) have been measured in time course and dose-response experiments with a number of direct-acting mutagens (Li, Cise, and Watson 2003). Chemical-induced increases in RFs detected using TaqMan after 17 h of exposure are similar to those observed by counting colonies on plates after 3 days’ growth on selective medium. The TaqMan assay can therefore be used to screen compounds earlier and more quickly than can be done by plating for selective growth. This strategy could be applied to any source of DNA.
DISCUSSION
TaqMan is one of an increasing number of technologies that provides valuable molecular data to toxicologists. Attributes that have made TaqMan a valuable quantitative tool are its excellent sensitivity, broad linear dynamic range, and high precision. The technology also has the flexibility of PCR itself, as exemplified by the three TaqMan applications presented in this article: gene expression analysis, genotyping, and detection of DNA deletions. These applications improve our ability to investigate biomarkers and mechanisms of toxicity, genetic contributors to biological activity of compounds, and increased throughput for quantification of adverse genetic events in chromosomal DNA.
How does one incorporate such a technology into drug development? Of the TaqMan applications presented here, gene expression analysis is the most commonly used at this time and is therefore most important to consider for integration in drug development. In the example given for cyclin D1 expression, TaqMan was used to provide mechanistic information that supported a decision to terminate development of a series of kinase inhibitors, and to deprioritize a drug target for one important indication. This is an example of targeted use of TaqMan for a defined gene selected by the investigator to answer a specific question. The technology can also be used in a screening mode for drug development, such as to answer the question “Does my compound cause transcriptional induction of drug metabolizing enzymes (DMEs) in a short-term rodent toxicity test?”. TaqMan is well suited for this question because the assays can be automated and the single analytical platform delivers data for all DMEs, regardless of the enzymatic reaction they catalyze (e.g., oxidation, conjugation, and transport). In contrast, a biochemical approach requires different assays that make use of different subcellular fractions, different substrates, and different analytical equipment.
What are the challenges of implementing TaqMan and other new technologies in the drug development pipeline? As pointed out in the cyclin D1 example, experiments must be designed and executed to measure molecular events that can be short-lived. Thus, incorporating exposure regimens and sampling times that differ from traditional toxicology studies is requisite. Also essential is the availability of technical staff that can process tissues in ways that assure sample quality: time is again an issue.
At a higher level, toxicology organizations must be willing to commit resources in key areas that go beyond acquisition of the instruments, consumables, and personnel expert in these new technologies. There must also be a system in place that drives toward a return on investment. One strategy that can speed this process is a sample archiving and retrieval system that provides specimens (i.e., serum, tissues) from toxicology studies that can be used to test new biomarkers in a timely manner. Without such a sample archiving system, data collection would be more time-consuming and more expensive, and as a result biomarker development would be slower than is necessary.
SUMMARY
TaqMan is one of many new technologies recently adopted by toxicologists in the pharmaceutical industry. Like virtually all emerging technologies, TaqMan went through a period of experimentation and evaluation before becoming recognized as a benchmark technology. In spite of TaqMan’s reputation of being a reliable technology, there are many challenges to its implementation on a routine basis. The challenges are both technical and cultural. Changes in behavior will be required to fully exploit the power of this technology. Working through these issues is critical not only for the success of the TaqMan technology, but also for the many other technologies that are available, and for those that are just around the corner.