Current Pharmacogenomics - Volume 3, Issue 1, 2005

With the recent sequencing of its entire genome, the yeast Saccharomyces cerevisiae has gained considerable interest in the field of anticancer research. The genetic properties of the yeast allow easy selection of a variety of mutants, as their related phenotypes provide valuable information on anticancer drugs effects in vivo. Moreover, the yeast has been extensively studied as a model system to decipher basic cellular processes that are well conserved from yeast to human and implicated in cancer establishment and progression. We discuss the advantages of Saccharomyces cerevisiae for investigating the metabolism of pyrimidines, a major target for anticancer chemotherapy. We also emphasize the links between basic science in the pyrimidine metabolism and new anticancer antimetabolites as well as future prospects in this field.

The ADP-ribosyltransferase (ADPRT) gene encodes the poly(ADP-ribose)polymerase-1 (PARP-1) enzyme, which plays critical roles in DNA-damage signaling and repair, cell death, maintenance of genomic stability, and carcinogenesis. It may also serve as a potential target for cancer therapy. In this review, we evaluate findings from animal model systems and molecular epidemiological studies to demonstrate the potential role of ADPRT/PARP-1 in aging and carcinogenesis. With increasing interest in associating human cancer risk with single nucleotide polymorphisms (SNPs) and/or dysfunction of ADPRT/PARP-1, several important technical challenges will need to be overcome. These challenges include developing specific functional assays, selecting SNPs with potential functional impact, and exploring statistical methods for gene-gene and gene-environmental interactions. Therefore, this review also highlights strategies to evaluate the functional significance of ADPRT/PARP-1 SNPs in human cancer risk assessment. In summary, dysfunction of PARP-1 may play a critical role in abnormal cellular functions; its molecular mechanism in aging and cancer susceptibility is an issue which needs urgently to be elucidated.

Severe neonatal unconjugated hyperbilirubinemia, with the risk of bilirubin encephalopathy or kernicterus in severe, untreated cases, occurs when bilirubin production exceeds the body's ability to eliminate it. The causes of neonatal hyperbilirubinemia are multifactorial and comprise increased hemolysis on the one hand, and diminished bilirubin conjugation on the other. In recent years, many of these etiologies have been found to have a genetic origin. Sometimes hereditary factors act independently, but in other circumstances, single mutations which ordinarily do not produce disease in and of themselves, may result in severe hyperbilirubinemia as a result of interaction with other mutated genes. Of cardinal importance to this discussion is the concept of the human genome, whereby the thousands of genes of which it is comprised may interact one with the other, or with the environment, exacerbating the severity of jaundice in certain individuals, and protecting against hyperbilirubinemia in others. Genetic interactions have been demonstrated combining increased bilirubin production with diminished bilirubin conjugation, resulting in severe hyperbilirubinemia. Appreciation of the multiplicity of genetic interactions is of importance in our evaluation of the neonate with severe hyperbilirubinemia, in our attempts to prevent future cases of kernicterus, and in genetic counseling of families who have had an infant with severe neonatal hyperbilirubinemia. Gene therapy for the most severe of these inherited defects of bilirubin conjugation, Crigler-Najjar syndrome type 1, might become a reality in future years.

Pharmacogenetics does seem to play a key role in the use of so-called weak opioids. It has been shown for codeine, dihydrocodeine, oxycodone and hydrocodone, that their O-demethylation in the 3-position results in metabolites which have much stronger μ-receptor binding. These opioids may therefore exert their pharmacological actions predominantly through their O-demethylated metabolites. However, this metabolic step is under genetic control of the polymorphic cytochrome P450 2D6 isozyme (CYP2D6). Poor metabolisers of CYP2D6 (∼10% of the Caucasian population) do not express this enzyme and hence can only form trace amounts of the O-demethylated metabolites of these four opioids. This might put these persons on risk of reduced or even abolished analgesic effects when given these weak opioids. From this point of view there are two major issues why weak opioids cannot wholeheartedly be recommended: large interindividual variability of the analgesic effect due to CYP2D6 polymorphism and 10% of patients with no benefit from these drugs. On the other hand it might be advantageous to use the O-demethylated metabolites morphine, oxymorphone and hydromorphone which are all strong opioids and have a smaller interindividual variability of the opioid effects. Instead of using weak opioids, small doses and controlled release formulations of strong opioids might be the future way to in analgesic therapy despite the fear of addiction and bureaucratic efforts involved with these compounds.

Nature or nurture? To what extent genetics play a role in drug efficacy and safety? These questions are not new. They are however gaining increasing prominence with the implementation of pharmacogenomics in various facets of medicine ranging from therapeutics, drug development and regulatory science to research funding decisions. For predisposition to common complex diseases, twin and family studies have been the mainstay for estimating genetic components of the attendant risk. On the other hand, the rapid pace of drug development in the pharmaceutical industry and the need for faster regulatory decisions call for an approach of higher throughput to identify the compounds for which heritability is likely to play a significant role in their pharmacokinetics and/or pharmacodynamics. A second predicament related to multifactorial nature of drug effects is that one typically observes a considerable overlap in the distribution of drug response phenotypes among subpopulations identified by each pharmacogenomic biomarker. This is in sharp contrast to monogenic pharmacological traits wherein it is feasible to partition the patient populations into discrete subgroups by analysis of a single gene. Hence, as pharmacogenomic investigations progress from monogenic to increasingly multigenic or multifactorial drug response phenotypes, the regulatory decision-makers are faced with a dilemma: How can a reviewer or a clinician determine if a given separation of a drug response profile by a pharmacogenomic biomarker is worthwhile for clinical implementation? The present manuscript makes an attempt to address these broad and emerging issues in pharmacogenomics and regulatory science. We propose that a comparison of inter- versus intra-subject variability in drug response under minimal environmental exposure may provide an upperbound estimate of heritability of drug efficacy and safety. It is also argued that seemingly modest changes in population averages may underestimate the dramatic impact of a genetic biomarker at the tails of a population. To this end, a conceptual framework for graded risk assessment among subpopulations with overlapping quantitative phenotypes is presented. We conclude with a broader discussion of the evolution of genetic biomarkers from monogenic to multigenic traits in pharmacology, the associated ethical, social and therapeutic policy corollaries and the challenges lying ahead.

Some forms of drug-induced lupus may be due to inhibition of T cell DNA methylation. DNA methylation modifies gene expression. In general, methylation of regulatory elements suppresses gene expression, while hypomethylation promotes gene expression. Methylation patterns are replicated during mitosis by a family of DNA methyltransferases, whose expression is regulated in part by signals transmitted through the extracellular signal-regulated kinase (ERK) pathway. Inhibition of DNA methylation during mitosis results in aberrant gene expression in the daughter cells, sometimes with pathologic consequences. A recent series of reports demonstrates that treating T lymphocytes with DNA methyltransferase inhibitors such as 5-azacytidine and procainamide, or ERK pathway inhibitors including hydralazine and the MEK inhibitor U0126, inhibits DNA methylation, alters gene expression, and induces autoreactivity. The autoreactive cells cause a lupus-like disease in animal models. Importantly, the same DNA sequences are demethylated in T cells from patients with active lupus, with identical effects on gene expression. These observations suggest that certain drugs and chemicals, and possibly as yet unidentified environmental toxins, can modify DNA methylation patterns through effects on DNA methyltransferase activity or expression, resulting in disordered gene expression and the subsequent development of a lupus-like disease.