Have you taken a medicine that hasn't worked, leaving a headache or some other ailment to linger on? Well, pharmacogeneticists have been telling us for some time that this apparent ineffectiveness of certain medicines might have its roots in how we metabolize drugs, but there has been little hard evidence to support this claim. However, proof now comes from a recent study, which has discovered single nucleotide polymorphisms ( SNPs) that affect the activity of a key metabolic enzyme and that occur at different frequencies in people from different ethnic groups.

This study, published in Nature Genetics, focused on the key enzymes involved in drug detoxification, the cytochrome P450 (CYP), and specifically the CYP3A family. CYP3A activity is the sum of the activity of several genes, including CYP3A5, the expression of which varies substantially in a minority of Caucasians. The CYP3A proteins metabolize and inactivate many drugs, including cancer chemotherapeutics, immunosupressants, steroids and HIV inhibitors — other CYP3A substrates include carcinogens and oestrogen. Kuehl et al. have now found that a SNP in a CYP3A5 intron leads to aberrant splicing of CYP3A5 and a truncated non-functional CYP3A5 protein. This is important because the full-length CYP3A5 product accounts for over 50% of total CYP3A activity, making it a principal component of the CYP3A complex. And their finding that African-Americans express higher levels of the full-length CYP3A5 allele than Caucasians implies that this ethnic group is less likely to experience dose-related drug toxicities that arise from inefficient drug clearance.

Kuehl et al. point out that studies such as theirs show that individual variation in drug metabolism can be easily determined for drugs that are metabolized by enzymes with well-defined and testable functions. But identifying SNPs that affect drug metabolism has been a barrier to progress in this field. How could this be done more efficiently in the future? The authors envisage that DNA from patients with known drug metabolism disorders could be screened against SNP databases as a way to identify novel loci of pharmacogenetic interest.