Polymerase chain reaction now key tool in genomic discovery

Genomics is rapidly escalating past simply identifying the genetic makeup of organisms to include examining how variations in genotype affect physiological function. Genomic analysis applications include mutation discovery, single nucleotide polymorphism (SNP) genotyping, DNA mapping, genetic screening, and population studies.

One of the key tools engaged in the pursuit of genomic discovery is polymerase chain reaction (PCR). Post-PCR analysis techniques used for genotyping and mutation detection include single stranded conformation polymorphism (SSCP), temperature gradient capillary electrophoresis (TGCE), and restriction fragment length polymorphism (RFLP). These techniques typically require costly laboratory equipment.

A more sensitive technique, probe-based genotyping can be completed on any real-time PCR instrument, but requires expensive specially labelled probes.

What's needed is a low-cost-per-sample method that also enables high-throughput sample processing. High-resolution melt (HRM) analysis is an alternative to probe-based genotyping assays that overcomes the cost, time, and labor-intensity challenges. HRM does not require special instrumentation as it can be performed on existing real-time PCR systems. This simple, yet highly sensitive and accurate tool is quickly becoming a mainstream part of the genetic analysis workflow.

DNA melt-curve analysis - applying temperature to melt and characterise the resulting curve profiles of double-stranded (ds) DNA samples - has proven useful for scanning for sequence variations, primarily to confirm the specificity of primers by ensuring no primer-dimers are present in quantitative PCR assays (Brisson et al. 2002). The goal of this method has been to prevent nonspecific amplification and improve data accuracy.

Recent advances in dye chemistry, instrument sensitivity, and data acquisition rates have led to the next generation of this technique - HRM analysis. A closed-tube, post-PCR analysis method that requires no post-PCR handling, HRM analysis generates DNA melt-curve profiles sufficiently specific and sensitive for mutation scanning, methylation analysis, and genotyping.

HRM analysis is performed to discriminate nucleotide sequence differences between samples. HRM analysis also enables mixed DNA fragments to be distinguished from each other - important for SNP genotyping of wild-type, heterozygous, and homozygous mutant individuals.

Three basic tools are used in HRM analysis:

- Specialised PCR reagents and dsDNA binding dye - third-generation saturating, low-toxicity dyes including EvaGreen (in Bio-Rad's SsoFast EvaGreen supermix) can be used in high concentrations to yield strong melt curves, but do not interfere with amplification during PCR (Wittwer et al. 2003).

- Real-time PCR instrumentation - system should offer sensitive detection for accurate quantitation, target discrimination and precise thermal control (such as Bio-Rad's CFX96TM or CFX384 TM real-time PCR systems).

- HRM analysis-compatible software - generates and analyses melt-curve profiles, and clusters samples with similar properties.

Because of its simplicity and abbreviated workflow, HRM analysis offers a cost-effective, yet accurate alternative to probe-based genotyping assays such as SSCP, RFLP, and DNA sequencing (White and Potts 2006)

After DNA is amplified in a real-time PCR instrument using a saturating dye-based master mix, the PCR product is melted using high data acquisition rates to generate melt curves. Software developed specifically for HRM analysis (such as Precision Melt Analysis software from Bio-Rad) enables analysis of the melt curves for genotyping and mutation scanning. Melt curves are analysed in the software using three basic steps:

- Normalisation - all samples are normalised along the fluorescence axis such that their average relative fluorescence value at the pre-melt signal is set to 100 per cent and post-melt signal is set to 0 per cent. This serves as a visual aid to interpret the data.

- Difference plotting - to magnify the differences in the melting curves between samples, each melt curve is subtracted from a user-defined reference melt curve and the fluorescence differences between samples are plotted. Similarly curved shapes will be clustered automatically into groups representing different genotypes/sequences.

- Temperature shifting (optional) - makes it easier to distinguish heterozygous from wild-type homozygous samples. Curves can be shifted along the high-end of the temperature axis to meet at the same specific temperature so that curve shapes are more accurately compared.

HRM Analysis Case Studies

One of the most challenging issues facing researchers seeking to identify the genetic factors involved in breast cancer is the polymorphic nature of genes commonly involved in expression of the disease: BRCA1 and BRCA2. Kim de Leeneer is a PhD student at the Center for Medical Genetics, Ghent, Belgium (CMGG) researching the genetics of breast cancer - specifically the BRCA1 and BRCA2 genes. De Leeneer conducted her first HRM experiments in January 2007, screening 212 positive control samples for breast cancer. Until these experiments, the primary tools used in CMGG to screen for the genetic inheritance of breast cancer were denaturing gradient gel electrophoresis (DGGE) and direct sequencing of both large exons 11 of BRCA1 and BRCA2. Traditional sequencing experiments were conducted in parallel to compare results and verify accuracy of the new technique. All controls were recognised, so de Leeneer and colleagues began converting traditional assays to HRM analysis.

Initially, all HRM analysis results were confirmed by sequencing; results demonstrated 100 per cent sensitivity and 98.7 per cent specificity of HRM analysis, with very few false positives. Because they have developed confidence in HRM analysis and the diagnostic aspects of their work require a high degree of throughput, they are now processing sample assays in single replicates. Only aberrant melting curves get sequenced to confirm the presence of a genetic variant.

As a prescreening tool, HRM analysis makes it easy to identify samples with genetic variants - and with a significant reduction in cost and time required over traditional methods. In de Leeneer's lab, HRM has reduced by about one-third the costs and workloads compared to DGGE and direct sequencing. De Leeneer believes HRM can possibly replace current prescreening techniques for other disorders and in other laboratories.

Alessandro Martino is a PhD student in the Department of Biology at Pisa University. The main focus of his studies is on SNP-based pharmacogenetics of multiple myeloma (MM). Currently, he is studying the rat gene expression that triggers repair responses in heart perfusion following heart failure. Martino's experiments mainly centre on membrane transporters, cytokines, and other pathways that potentially modulate drug response and survival rates after chemotherapy. Studies involve analysing SNP mutations - primarily class I and II, but also some class IV - in MM patient blood samples.

Until recently, the primary techniques used for SNP analysis were dual-labelled hydrolysis probe assays, but Martino has tested HRM analysis with the aim to replace probe-based screening methods. In initial HRM experiments, researchers used the same primer pairs and reagents as with dual-labelled hydrolysis probe assays. These early HRM experiments were run in parallel with probe-based assays, and Martino observed good correlation between the two. These studies demonstrated that melting temperature of the amplicon and primer pair specificity are influenced by the primer pairs used, so Martino began to develop primer pairs for HRM studies.

Replacing a self-made reagent mix with a commercial supermix further optimised experiments. The new reagent enabled amplification of targets that were previously problematic. HRM analysis also improved their results with allelic discrimination experiments over probe-based assays. Because allelic groups cluster using HRM-based methods, they can be identified via melting curves generated by the software.

Viresh Patel, PhD, is a Senior Product Manager for Amplification Reagents at Bio-Rad Laboratories, Hercules, CA, USA. www.bio-rad.com

REFERENCES:

Brisson M et al (2002). Identification of nonspecific products using melt-curve analysis in the iCycler iQ detection system. Bio-Rad bulletin 2684;

Erali M et al (2008). High resolution melting applications for clinical laboratory medicine. Exper and Mol Pathol 85, 50-58;

White H and Potts G (2006). Mutation scanning by high resolution melt analysis;

Evaluation of Rotor-Gene 6000 (Corbett Life Science), HR-1, and 384-well;

LightScanner (Idaho Technology). Natl Gen Ref Lab - Wessex;

Wittwer CT et al (2003). High resolution genotyping by amplicon melting analysis using LCGreen. Clin Chem 49, 853-860.

Recent Issues