Medical Device Link.

Originally Published IVD Technology November/December 2002

Molecular Diagnostics

Genetic analysis and future pharmacogenomic applications

A technique for sequencing single nucleotide polymorphisms holds promise for individualized drug therapies.

Kellie Watson and Jerry Williamson

The Human Genome Project provided biomedical researchers with access to nearly the complete record of DNA sequences found in the human genetic makeup. This geno-mics initiative was soon followed by a project to identify and map single nucleotide polymorphisms (SNPs), the most common form of genetic variation in humans. The latter effort was organized and funded by The SNP Consortium. These projects were directed at acquiring raw genetic data. But now, in order to reap medical benefits, the focus is shifting toward practical applications of this genetic information.

SNPs are expected to play an important role in this new era of applied genomics. Consequently, significant opportunities may be derived from SNP analysis of individuals and populations. It is anticipated that SNP analyses will be used to diagnose common diseases, to determine disease susceptibility, and to predict or monitor an individual's response to therapeutic treatment. The ability to tailor therapeutic treatment to meet the specific needs of a particular patient and still avoid adverse effects is the promise of pharmacogenomics.

SNPs in Pharmacogenomics

Single nucleotide polymorphisms represent as much as 90% of all human genetic variation. They are defined as single base-pair positions in genomic DNA that vary among individuals in one or several populations. The least-common base alternative, or allele, must occur in the population at a frequency of at least 1% for the variant to be considered an SNP. SNPs are believed to underlie susceptibility to such common diseases as cancer, diabetes, and heart disease and to contribute to the traits that make individuals unique. In addition, these polymorphisms may help explain why different people respond differently to the same drug.

SNPs are being used as genetic markers in numerous genomewide association studies of large sample populations to identify genes, or unknown genetic variants that contribute to common, genetically complex diseases and disorders.^1,2 The goal of these pharmaco-genomic efforts is to determine the frequency of specific alleles in various populations or in diseased or predisposed individuals. Publicly available databases of SNPs, which include their exact location in the human genome, have been generated by two large-scale genomics efforts. The SNP Consortium has mapped more than 1.25 million SNPs (available at http://snp.cshl.org/), and the National Center for Biotechnology Information has cataloged more than 4.1 million of them in a database called dbSNP (http://www.ncbi.nlm.nih.gov/SNP/index.html).

Besides foretelling the risk of disease, SNPs can provide information about a patient's likely response to drug therapy based on factors present in both host and pathogen genomes. By understanding the relationship between SNPs and their impact on patient health, physicians will be able to use genotypic information to diagnose disease, determine susceptibility or resistance to drugs, and select therapies to minimize side effects and offer the best potential outcomes. Routine genotyping may also be used in drug trials to screen volunteers and patients.

Pharmacogenomics. Pharmacogenomics (also called pharmacogenetics) is the study of how genes affect the physiological response to drug therapies. Personalized medicine may be regarded as the objective of this field of inquiry. Personalized medicine involves the use of diagnostic procedures to

• Determine an individual's predisposition to disease.
  
• Detect disease early while intervention is feasible.
  
• Assess the severity of disease and monitor disease progression.
  
• Select the most effective drug therapy and dose.
  
• Evaluate the effectiveness of treatment.
  
• Avoid adverse reactions to therapeutic drugs.

The integration of diagnostics and therapeutics has the potential to seriously reduce adverse drug events, which now cause more than 2 million hospitalizations and 100,000 deaths annually, at a cost of $100 billion to the healthcare system.^3,4

Although molecular diagnostic testing represents less than 4% of the total IVD market, it is the fastest-growing segment. Sales of molecular diagnostic tests are expected to reach $4.2 billion by 2006 as a result of increasing demand. In fact, industry analysts predict that by 2010 half of all prescribed therapeutics will have a corresponding diagnostic test.³

SNP Technologies. Technologies developed to read the long stretches of DNA that were important for the sequencing of the human and other genomes are not well suited for SNP and other short-read-sequence applications. For example, studies investigating relationships between SNPs and specific conditions, including disease states and genetic predisposition to specific conditions, require that large numbers of samples be tested to establish statistically significant results. This means that any sequencing technique used must be highly accurate, reliable, and fast. Sequence information about the SNP in question and, preferably, the base sequence surrounding the polymorphic position are also required. This is a significant challenge when large numbers of samples need to be tested. For SNP analyses to be useful in more-routine applications, these issues must be resolved. Improved SNP technologies are needed that offer ease of use as well as accuracy and reliability. However, many of the techniques introduced to perform SNP analysis demand a high degree of technical skill and experience in order to achieve accuracy and reliability. This could be an obstacle to the routine application of SNP analysis, particularly in clinical settings.

As listed below, several technologies have been developed to perform SNP genotyping. They are based on

• Sequencing (systems are available from Pyrosequencing AB [Uppsala, Sweden], Amersham Pharmacia Biotech Inc. [Piscataway, NJ], Visible Genetics Inc. [Toronto], and Applied Biosystems [Foster City, CA]).
  
• Microarray/microfabrication (developers include Affymetrix Inc. [Santa Clara, CA] and Nanogen [San Diego]).
  
• Mass spectrometry (Sequenom Inc. [San Diego] and Qiagen N.V. [Venlo, The Netherlands]).
  
• Single-base primer extension (Orchid Biosciences Inc. [East Princeton, NJ]).
  
• Probes, hybridization, or enzymatic reactions (Applied Biosystems [Foster City, CA], Third Wave Technologies Inc. [Madison, WI], Promega Corp. [Madison, WI], and Amersham Pharmacia Biotech [Buckinghamshire, UK]).
  
• Beads (Lynx Therapeutics Inc. [Hayward, CA], Illumina Inc. [San Diego], and Luminex Corp. [Austin, TX]).
  
• Molecular beacons.

This article uses one sequencing technology from this group to illustrate several potential pharmacogenomic applications.

Pyrosequencing Technology

Figure 1. How the Pyrosequencing technology works: Added deoxynucleotide triphosphates (dNTPs) are incorporated by DNA polymerase if they are complementary to the base in the template strand. Pyrophosphate (PPi) is released in a quantity equimolar to the nucleotides incorporated (a). Sulfurylase converts PPi to adenosine triphosphate (ATP) in the presence of adenosine 5'-phosphosulfate and drives the luciferase-mediated conversion of luciferin to oxyluciferin. Light produced by the luciferase-catalyzed reaction is quantitatively detected by a charge-coupled device camera and seen as a peak in a pyrogram (b). Apyrase, a nucleotide-degrading enzyme, removes unincorporated dNTPs and excess ATP, producing diphosphates, monophosphates, and phosphate (c). (click to enlarge)

Pyrosequencing technology employs a sequencing-by-synthesis approach. A double-stranded DNA product is synthesized from a single-stranded DNA template by the addition of complementary nucleotides (see Figure 1). Nucleotide incorporation into the DNA template results in the generation of visible light via an enzymatic cascade involving pyrophosphate and the firefly enzyme luciferase.

The light signal is proportional to the number of incorporated nucleotides; its quantitative measurement is captured digitally in a Pyrogram. System software deduces sequence information from the Pyrogram and assigns quality values to the data on the basis of a series of internal controls and standards.

Pyrosequencing generates direct sequence information on the position of interest and up to 50 surrounding bases. Longer reads have been performed, including de novo sequencing (see Figure 2).⁵ The technology can thus be used for the identification and typing of infectious pathogens, where longer stretches of sequence information are needed to distinguish closely related species.⁶

In SNP analysis, establishing the sequence context is crucial for validating genotyping results. Furthermore, accuracy is essential in SNP or genetic analysis, since even a small error rate has a dramatic impact on the utility of the genetic data and since such errors can be amplified when analyses involve large patient samples. The accuracy of Pyrosequencing technology has been evaluated in numerous studies and has been determined to be greater than 99%.⁷

SNP Analysis. Pyrosequencing performs SNP analysis by means of a simple, standardized procedure that takes approximately 1 minute per base. The procedure begins with converting a polymerase chain reaction product into a single-stranded DNA template and annealing a sequencing primer. Samples are held in a standard 96- or 384-well microplate. In one process, one or more samples can be analyzed for several different SNPs. Alternatively, a single SNP can be scored in numerous different samples. The instruments auto-matically dispense the necessary sequencing reagents and enable real-time detection of sequencing events in each well.

The analysis of specific DNA material is used increasingly in clinical microbiology where speed is often critical in the diagnosis and correct treatment of infectious disease, and also in forensics for the rapid identification of a DNA sample important to a legal investigation. Other areas in which genetic analysis will continue to play a significant role include the identification of bacteria or viruses, the determination of the antibiotic resistance status of bacteria (often indicated by a specific mutation in a known region of a gene), and the expeditious tracking of contamination sources. Some techniques take days to identify or characterize bacteria and viruses, but a Pyrosequencing system has been used to process parallel samples in a few hours and generate easily interpretable results.^6,8,9

Figure 2. An example of de novo sequencing by dispensing nucleotides in the order CTGA. (click to enlarge)

Allele Frequency Determination. Traditional techniques for determining the frequency of alleles in target groups or entire populations normally require the independent investigation of thousands of individual samples to provide statistically significant data. As an alternative, allele frequencies can be determined in pooled DNA samples involving hundreds or thousands of individuals per pool. This provides dramatic savings in resources, including reagents, labor, and valuable genomic material.

In a number of independent association and linkage studies and analyses of mutations in mixed populations, allele frequencies of less than 2% have been detected via Pyrosequencing technology using genomic DNA as the starting material.^10,11 The system provides automatic quantification of allele frequencies in the pooled DNA samples, including automatic statistical calculations. Other important applications of this approach include the detection of tumor cells in patient samples and the identification of subtypes in bacterial and viral mixed populations.

Pharmacogenomic Applications

Pyrosequencing technology has a number of phar-macogenomic research uses, including genotyping to identify infectious pathogens and their drug resistance status;^{6,8,9,12–14} to assess disease susceptibility, prognosis, or progression;^{10–12,15,16} and, through the analysis of drug-metabolizing genes, transporters, and the like, to predict drug efficacy and side effects.^17,18 All of the research-use applications provide information that can be used to select or guide therapeutic treatments (see Table I).

In addition to identifying and classifying infecting microbes such as bacteria or viruses, genotypic characterizations could be used to determine genetic factors that affect pathogen virulence, host susceptibility, and antibiotic resistance, and thus would influence the choice of therapeutic regimen. For example, a Pyro-sequencing assay has been developed for research use that is capable of typing the 12 most common types of human papillomavirus (HPV) by sequencing fewer than 25 bases in a conserved region of the HPV genome.⁹

One common application of SNP testing in clinical diagnostic laboratories is the identification of genetic variants in the genes encoding co-agulation factors V and II (pro-thrombin). These variants play a role in the diagnosis of hereditary thrombophilia. The Factor V Leiden mutation is present in 20–40% of venous thrombosis cases and 3% of the general population. It increases the risk of venous thrombosis by seven times in heterozygotes and 80-fold in homozygotes.

Variations in the cytochrome P450 (CYP) genes are known to have a significant effect on the activity of these drug-metabolizing enzymes. CYP2D6 enzyme has been studied extensively because it is responsible for the metabolism of tricyclic anti-depressants, antipsychotics, anti-arrythmics, and beta blockers. Some 7–10% of Caucasians have variant CYP2D6 genes that produce a poor-metabolizer (PM) phenotype and lead to greater risk of drug accumulation and toxicity from the drugs metabolized by these enzymes. Despite the high degree of polymorphism in the gene, four alleles—CYP2D6 *3, *4, *5, and *6—account for 93–97% of the PM phenotypes among Caucasians. Pyrosequencing assays for research use have been developed for the analysis of CYP2D6 alleles *2, *3, *4, *6, *7, and *8; these involve two multiplex reactions. Oth-er CYP gene assays capable of distinguishing the majority of poor metabolizers' genotypes for other drug metabolizing enzymes (CYP2C19 and CYP2C9, for example) have been designed as single-multiplex pyrosequencing reactions for research use.

Figure 3. Pyrosequencing pyrograms of exon 8 in three angiotensin I–converting enzyme genotypes. The A/T SNP (A-240T) position is outlined by the box. The A/T heterozygous sample is at the top, followed by the two homozygous samples, A/A and T/T. The y-axis represents arbitrary luminescence units, and the sequencing primer is displayed below the pyrograms. (click to enlarge)

The angiotensin I–converting enzyme gene represents an essential component of the renin-angiotensin-aldosterone system (RAAS), which plays a crucial role in hormonal mechanisms that regulate blood pressure and electrolyte–blood volume ho-meostasis.¹⁹ SNPs located in RAAS genes are being investigated as potential genetic signatures that may be predictive of clinical drug response. A research-use SNP assay using Pyrosequencing technology can distinguish gene variations in human angiotensin I–converting enzyme (see Figure 3).

Genotypic and phenotypic information is appearing increasingly in federal regulatory submissions. As of February 2002, 44 investigational new drug (IND) and new drug applications (NDA) included such data. Most of these submissions involve drug pharmacokinetic or metabolic information. An intriguing opportunity lies in targeting drugs at specific patient populations, with the drug labels naming the tested or indicated patient group. To achieve this, would a test able to identify the targeted patient group.

The first pharmacogenomic assays to be introduced will likely be tests developed and validated for use by clinical testing laboratories. Authorized clinical molecular laboratories are permitted to perform such so-called home brew assays. In addition, assays can be developed as analyte-specific reagents manufactured for use in home brew assays. Ultimately, approved validated kits could be developed under regulatory guidelines as IVD tests. These would re-quire external assessments of va-lidity, as well as performance and manufacturing quality controls. Test volume and clinical utility will drive the implementation of these tests in clinical laboratories. Regulatory and cost-benefit analyses will figure prominently in the selection, development, and introduction of pharmacogenomic tests.

Conclusion

Within a few years, the value of mapping the human genome may be realized in terms of improved healthcare and better drugs. Genotyping an individual or a biological specimen in order to determine appropriate treatment is likely to become routine clinical procedure. Further, as use of genetic tests for drug response becomes more broadly accepted, drugs targeted toward genetically defined patient groups may begin to appear. Validated assays to identify these specific patients would be developed concomitantly. These assays may include SNP- or other sequence-based tests.

Pharmacogenomics seeks to optimize the selection of drug therapy, increase drug safety and efficacy, reduce the time and overall cost required to manage disease, increase cure rates, and minimize morbidity and mortality. Although it now has little to do with the daily practice of primary-care physicians, pharmacogenomics can be expected to play an integral role in the future of medicine and healthcare delivery.

References

1. N Risch and K Merikangas, "The Future of Genetic Studies of Complex Human Diseases," Science 273 (1996): 1516–1517.

2. L Kruglyak, "Prospects for Whole Genome Linkage Disequilibrium Mapping of Common Disease Genes," Nature Genetics 22 (1999): 139–144.

3. J Lazarou, BH Pomeranz, PN Corey, "Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies," Journal of the American Medical Association 279 (1998): 1200–1205.

4. Frontline Strategic Management Consulting, Inc., "Pharmacogenomics, A Strategic Business Analysis" (2001).

5. B Gharizadeh et al., "Long-Read Pyrosequencing Using Pure 2'-Deoxyadenosine-5'-O-(1-thiotriphosphate) Sp-Isomer," Analytical Biochemistry 301, no. 1 (2002): 82–90.

6. HJ Monstein, NB Shohreh, and J Jonasson, "Rapid Molecular Identification and Subtyping of Helicobacter pylori by Pyrosequencing of the 16S rDNA Variable V1 and V3 Regions," Federation of European Microbiological Societies Microbiology Letters 199 (2001): 103–107.

7. M Ronaghi, "Pyrosequencing Sheds Light on DNA Sequencing" Genome Research 11 (2001): 3–11.

8. H Unnerstad et al., "Pyrosequencing As a Method for Grouping of Listeria monocytogenes Strains on the Basis of Single Nucleotide Polymorphisms in the inlB Gene," Applied Environmental Microbiology 67, no. 11 (2001): 5339–5342.

9. B Gharizadeh et al., "Typing of Human Papillomavirus by Pyrosequencing," Laboratory Investigation 81, no. 5 (2001): 673–679.

10. JD Gruber, PB Colligan, and JK Wolford, "Estimation of Single Nucleotide Polymorphism Allele Frequency in DNA Pools by Using Pyrosequencing," Human Genetics 110, no. 5 (2002): 395–401.

11. J Wasson et al., "Assessing Allele Frequencies of Single Nucleotide Polymorphisms in DNA Pools by Pyrosequencing Technology," BioTechniques 32, no. 5 (2002): 1144–1148.

12. Technical and Application Notes, www. pyrosequencing.com, Pyrosequencing AB, Uppsala, Sweden.

13. M Nygren et al., "Quantification of HIV-1 Using Multiple Quantitative Polymerase Chain Reaction Standards and Bioluminometric Detection," Analytical Biochemistry 288, no. 1 (2001): 28–38.

14. D O'Meara et al., "Monitoring Resistance to Human Immunodeficiency Virus Type Protease Inhibitors by Pyrosequencing," Journal of Clinical Microbiology 39, no. 2 (2001): 464–473.

15. CG Garcia et al., "Mutation Detection by Pyrosequencing: Sequencing of Exons 5–8 of the p53 Tumor Suppressor Gene," Gene 253 (2000): 249–257.

16. LA Hefler et al., "Polymorphisms of the Angiotensin Gene, the Endothelial Nitric Oxide Synthase Gene, and the Interleukin-1beta Gene Promoter in Women with Idiopathic Recurrent Miscarriage," Molecular Human Reproduction 8, no. 1 (2002): 95–100.

17. RA Kittles et al., "Cyp17 Promoter Variant Associated with Prostate Cancer Aggressiveness in African Americans," Cancer Epidemiology Biomarkers Prevention 10, no. 9 (2001): 943–947.

18. H Andréasson et al., "Mitochondrial Sequence Analysis for Forensic Identification Using Pyrosequencing Technology," BioTechniques 32, no. 1 (2002): 124–133.

19. TL Goodfriend, ME Elliot, and KJ Catt, "Angiotensin Receptors and Their Antagonists," New England Journal of Medicine 334 (1995): 1649–1654.

Kellie Watson, PhD, is director of business development at Pyro-sequencing AB (Uppsala, Sweden). She can be reached at kellie.watson@pyrosequencing.com. Jerry Willliamson is president of Pyrosequencing Inc. (Westborough, MA), a subsidiary of Pyrosequencing AB. He can be reached at jerry.williamson@pyrosequencing.com.

Copyright ©2002 IVD Technology