|dc.description.abstract||New opportunities for better understanding of drug responses and adverse drug reactions have been created through the rapid evolution of high-throughput genotyping and next-generation DNA sequencing technologies. The main goals of this thesis were to establish a framework for collecting patient samples to enable the study of unusual drug responses and adverse drug reactions, and to explore the application of next-generation DNA sequencing methods in a range of pharmacogenetic and pharmacogenomic studies.
At the outset of this project, the pharmacogenetic aspects of international, direct-to-consumer genetic testing companies were reviewed. This revealed that the services provided by these companies were patchy and each company covered a varied range of drug-gene pairs. The chip- or PCR-based methodology employed by the companies also meant that only the most common variants were genotyped and this could affect the accuracy of phenotype prediction across different ethnic groups. The value of these services was also limited by a lack of clear evidence supporting their utility in the clinical setting. Nevertheless, it was concluded that these direct-to-consumer companies have played a useful role in educating the public - and indirectly the health professionals - with regard to the potential, clinical application of pharmacogenetics.
A prototype biobank (called Understanding Adverse Drug Reactions Using Genomic Sequencing, or UDRUGS) for the study of adverse drug reactions and unusual drug responses was established. Via this biobank, samples were sourced by various routes and subjects were consented for genetic analysis that ranged from candidate-gene to whole-exome sequencing. High-throughput genomic sequencing generates incidental findings that could be of health significance; thus a method for returning such findings to patients who agreed to it was established. UDRUGS cases formed the basis of the genetic and genomic studies in this thesis. Subject recruitment is an actively ongoing process. At the time of writing this thesis, a total of 37 participants had been recruited and in the majority of the cases, metabolic deficiency (for instance defects in the cytochrome P450 enzymes) appeared to be culpable for the reported adverse reactions.
Targeted Sanger DNA sequencing analysis proved revealing in a case of poor tolerance towards the administration of venlafaxine monotherapy, and that of a combined regimen of nortriptyline and fluoxetine. Novel CYP2D6*81 and CYP2C19*34 alleles were identified by Sanger sequencing. CYP2D6*81 creates a premature stop signal in the fifth exon of the CYP2D6 gene, probably causing the formation of a non-functional, truncated protein. The effect of CYP2C19*34 was less clear-cut, as the allele is composed of an upstream variant and two non-synonymous variants located near the translation start site. This allele was co-identified alongside the common CYP2C19*2 null allele. Long-range haplotype analysis, spanning a 19.5-kb region, showed that the three variants occurred on a separate allele from CYP2C19*2. It was noteworthy that this participant was of Gujarati descent and it is possible that the identified variants may represent CYP2D6 and CYP2C19 alleles specific to this population.
Recognising the important roles of CYP2D6 and CYP2C19 in drug response, a next-generation amplicon sequencing assay was designed for the comprehensive genotyping of the two genes in large numbers of clinical samples. CYP2D6 was contiguously amplified by long-range PCRs (6.6-kb) whereas three triplex PCRs were set up to amplify nine short amplicons covering the promoter region and all exons of the CYP2C19 gene. A total of 96 samples were included in the first assessment of the assay, which was run on the MiSeq® platform. High-quality data were generated with 100% agreement with the Sanger sequencing data for 19 control samples that were also included. However, several drawbacks of the assay, namely the tedious and costly nature of the PCR-based library construction steps and the coverage variation observed for CYP2D6, need to be addressed in follow-up experiments. The compatibility of this assay with a novel sequencing device, the MinIONTM (Oxford Nanopore Technologies, Oxford, UK), which detects DNA bases by disruptions of ionic current as the DNA strands are drawn through an array of nano-scale membrane pores, was also partially tested.
The choice of PCR microtubes was found to have a significant impact on the success of long-range PCRs used for examining the CYP2D6 gene. Coloured microtubes purchased from Neptune Scientific (Biotix, Inc., San Diego, CA, USA) performed poorly in comparison with those obtained from other suppliers. It is possible that chemical compound(s) leached from the microtubes upon heating and inhibited PCR amplification, although a definitive conclusion was not possible.
Whole-exome sequencing and array-based comparative genomic hybridisation were applied to explore the hypothesis that rare coding variants or structural variants contribute to a thiopurine hypermethylation phenotype seen in a proportion of inflammatory bowel disease patients treated with thiopurine drugs. In the exome data of twelve patients, four genes (SLC17A4, RCC2, NFS1 and ENOSF1) were noted to harbour rare and potentially deleterious variants that occurred at a higher-than-expected frequency (p <0.05). The role of pathogenic copy-number aberrations and the contribution of 52 a priori candidate genes to the phenotype was found to be insignificant. Further understanding of the genetic architecture of this trait would likely require the collection of many more cases and controls, in the context of a genome-wide association study.
In a New Zealand cohort of acute coronary syndrome patients, CYP2C19*2 was shown to have a substantive influence on clopidogrel responsiveness whereas the effect of *17 was less obvious. Eleven outliers with at least one intact CYP2C19 gene copy but very high residual platelet reactivity (≥90 U) were selected for whole-gene sequencing; no additional mutations were identified. Overall this suggested that in certain instances, non-genetic factors predominated over the effect of CYP2C19 polymorphisms.
This work has demonstrated that a variety of DNA sequencing technologies can be valuable tools for elucidating the genetic underpinnings of unusual ADRs or complex drug responses. It is foreseeable that the progressive advancement of next-generation sequencing technologies will continue to generate new knowledge in pharmacogenetics or pharmacogenomics, and perhaps make possible the pre-emptive identification of genetic variants that predispose to unusual drug responses or adverse reactions.||