Prior to the advent of recombinant techniques, knowledge of the biochemical or pharmacological defect for a few traits had led to the discovery of the responsible genes. Certain disorders of hemoglobin and other blood-borne proteins had been attributed to the presence of missense, nonsense, or frameshift mutations that occurred within the coding region of the gene, but recombinant DNA studies revealed that genetic diversity could be created by mutation at various regulatory sites as well as within the coding region of the gene. One entirely unanticipated but particularly notable consequence of these studies was the demonstration that higher organisms had evolved the ability to generate different proteins from a single gene (by alternative splicing mechanisms).
As the pace of mutation detection quickened, it soon became evident that mutations underlying hereditary disorders were most often revealed in one of two ways: (1) by the production of a functionally altered gene product (protein), or (2) by the production of a normal gene product in an altered amount. In the latter case, the amount of the variant protein was most often found to be reduced. There are, however, reports of several traits including serum butyryl cholinesterase sensitivity,49-51 CYP2D6 polymorphism,48 and glutathione-S-transferase52 attributed to increased amounts of the variant protein. To a first approximation, mutations that cause the synthesis of structurally altered polypeptides or proteins usually occur within the coding region of the gene; they include point mutations, frameshift mutations, large deletions or insertions, and chain termination mutations, whereas those that result in altered amounts of protein product include mutations of the transcriptional machinery such as deletions, insertions, mutations of the promoter regions and other regulatory regions, mutations of RNA processing such as those that alter splicing in the 5' untranslated region or 3' adenylation signals, and mutations of the translational machinery that are responsible for the initiation, elongation, and termination of polypeptide chains. The potential for epigenetic modifications that alter gene expression has been recognized more recently as discussed below (see Chapter 6).
These rough guidelines are intended to provide some perspective into the relationship of genetic changes to protein variation that may alter drug responses. It is prudent, however, to keep an open mind about any attempt to classify muta-tional mechanisms too rigidly from such limited information as it may not be entirely satisfactory. For instance, some traits that appear to be due to a quantitative defect in protein synthesis may on further analysis be found to be due to a structurally abnormal protein that undergoes rapid proteolysis, and others that are genetically heterogeneous could be due to more than a single type of mutational, recombinational, or epigenetic event.
Many pharmacogenetic traits are due to point mutations or other small genic lesions that lead to a functional change in the protein encoded, or to a virtual absence of the protein if the gene that encodes it is deleted entirely or is truncated by the introduction of a premature stop codon. Such alterations occur at specific genetic loci, and hence are useful as genetic markers for diagnosing specific traits. When DNA mutates, recognition sites for restriction enzymes may also change, and such changes can be detected by the difference in the RFLP patterns. The ability to detect minute lesions in DNA, as well as to determine the specific structural change, allows for the diagnosis of the traits in question. Expression and characterization of the mutant and normal proteins may also suggest a plausible explanation of the unusual response.
PCR can provide additional strategies for detecting minute lesions of DNA. Consider, for example, the following experiment. A given gene is amplified in one test tube with "wild-type" primers and cleaved with an appropriate restriction enzyme. In another tube, allele-specific amplification of an aberrant gene is performed with primers made to the mutated site and cleaved with the same restriction enzyme. DNA is needed only for the first amplification because additional amplifications can be performed with PCR-generated template DNA. The resulting fragments are separated on agarose gels (or alternatively on polyacryl-amide gels for improved resolution of very faint bands), and the RFLPs are compared for pattern differences. In many applications, allele-specific amplification by PCR is preferable to Southern analysis for comprehensive screening of populations because the latter technique is unable to discriminate many alleles and because PCR analysis is at least an order of magnitude more sensitive than Southern blotting for the detection/amplification of novel sequences, and requires only a small amount of DNA that can be obtained from leukocytes, buccal epithelium, or even dried blood spots.
A number of genes have been found to incorporate microsatellites. Microsatellites are stretches of repetitive DNA sequences, and if mutated, these elements appear to be capable of disrupting cell function. The importance of microsatellite instability has been recognized as a prognostic marker of familial and sporadic colorectal cancer.53 Microsatellite instability, a striking molecular feature of colorectal cancer seen in more than 90% of cases, refers to somatically acquired variations in the length of repetitive nucleotide sequences in DNA, such as (CA)n or (GATA)n. Gryfe and colleagues54 tested the hypothesis that colorectal cancers arising from microsatellite instability have distinctive clinical features that affect outcome. In a population-based series of 607 patients (50 years of age and younger) they found a high frequency of microsatellite instability in 17% of colorectal cancers in these patients. Additionally, they found that these cancers had a decreased likelihood of metastasizing to regional lymph nodes regardless of the depth of tumor invasion. They concluded that high-frequency microsatellite instability reduced the chances of metastases and is an independent predictor of a relatively favorable outcome. In another report, Datta and co-workers55 describe genetic and phenotypic correlates of colorectal cancer in young patients of age 21 years or younger. Datta's report indicates that there is microsatellite instability in almost half the colorectal cancers in children and young adults.
The genome is well covered by microsatellites, but polymorphisms with these repeats within genes are less abundant. Two methods have been applied successfully to the identification of alleles that cause disease: single-strand conformational polymorphism (SSCP) because of its simplicity, and density gradient gel electrophoresis (DGGE) because of its high, near 100%, sensitivity for allelic differences. Unfortunately, SSCP requires two or three gel conditions for detection and DGGE needs special equipment and more expensive GC-clamped primers. Another method has been described, enzymatic mutation detection (EMD), that promises to be superior to existing techniques in the search for elusive mutations.56 The technique employs bacteriophage resolvases that are capable of recognizing mismatched bases in double-stranded DNA and cutting the DNA at the mismatch. EMD takes advantage of the mismatch to detect individuals who are heterozygous at the given site—the presence and the estimated position of the mutation are both revealed. In simpler terms, the resolvase can be thought of as a restriction enzyme that recognizes only mutations.
Was this article helpful?