A Strategy to Setup Codominant Microsatellite Analysis for High-Resolution-Melting-Curve-Analysis (HRM)
© Mader et al. 2008
Received: 03 July 2008
Accepted: 03 November 2008
Published: 03 November 2008
Skip to main content
© Mader et al. 2008
Received: 03 July 2008
Accepted: 03 November 2008
Published: 03 November 2008
High resolution melting curve analysis (HRM) is a technique that measures exactly the decreasing fluorescence of intercalating dye in the process of dissociation of double stranded DNA. The measurement is immediately following PCR in a one-step, closed-tube method. The shape of the melting curve depends on the GC content, length and sequence of the amplicon. Hence it is a powerful, fast and cheap method to detect Single Nucleotide Polymorphisms (SNPs) and other mutations.
Here we present a strategy to set up microsatellite analysis for HRM including the correct assignment of heterozygous samples by comparative analysis and artificial mixtures of samples. The approach is demonstrated on two Simple Sequence Repeat (SSR) loci of different complexity in the genus Origanum. Following this strategy all alleles of our sample sets could be classified correctly.
HRM can be used in microsatellite analysis and other codominant marker systems implementing a protocol of comparative melting curve assignment with artificial mixtures of samples to overcome difficulties in correctly assigning heterozygous samples. The method is faster, more sensitive and cheaper than standard protocols for microsatellite analysis.
High resolution melting curve analysis (HRM) has been introduced several years ago, extending the possibilities of the analysis of DNA melting curves, a standard diagnostic feature in qPCR , towards a sensitive method in detecting mutations. The main field of application is the reliable and fast determination of SNPs , where the sensitivity of the method has already been widely demonstrated . It is also used for assessment of DNA methylation . HRM also proved applicable for larger sequence aberrations such as Internal Tandem Duplications (ITDs) . It was used for the detection of unknown mutations in a region of low complexity  and recently also for analysis of microsatellites .
HRM can be seen as an offspring from qPCR technology, where fluorescent dyes are used to detect the quantity of double stranded DNA during the PCR. At the end of the PCR, the fluorescence signal level is high, due to the high amount of dsDNA with intercalated fluorescence dye accumulated in the process. Primer dimers and other non-specific products can subsequently be detected in most qPCR instruments by melting the dsDNA and a stepwise measurement of the decreasing fluorescence. The distance between Tm values allows the discrimination of the targeted amplicon and non-specific products . Technological progress facilitated a decrease of the size of temperature steps between fluorescence measurement points in the melting process, generating more data and thus drawing a high resolution curve of the level of fluorescence. The HRM curves obtained are highly characteristic for each amplicon and depend on GC content, amplicon length and sequence .
Variation in microsatellite loci is often visualized by electrophoresis on polyacrylamide gels, which is principally sensitive only to amplicon-length. Gel-electrophoresis is not only time-consuming, but may also lead to problems in interpretation due to stutter bands. The nowadays common method of microsatellite analysis with multiplex PCR and capillary electrophoresis affords the use of labeled primers, which have to be optmized for multiplex PCR. The problem of stutter peaks also occurs and the method is still quite costly offside routine analysis with well known organisms and markers. Above this it requires at least one additional step following PCR.
Here we describe a strategy to develop a co-dominant microsatellite marker system in diploid organisms using HRM with the advantages of visualization immediately following PCR, low costs and higher sensitivity than electrophoretic detection systems.
Microsatellite primers for Origanum vulgare, which have also cross-amplified in other species of the genus Origanum  are used to demonstrate the strategy.
PCR and HRM were performed on a RotorGene 6500 (Corbett Research Pty Ltd, Sydney, Australia) with a HRM-module, the results were analysed using the RotorGene 6000 series software, Version 1.7.65. To obtain interrun comparability standard samples were used in every run.
DNA was extracted from dried leaves of Origanum using a modified CTAB extraction protocol . The samples represent a subset from a study on natural hybridizations in the section Majorana of the genus Origanum (data in preparation). The individuals tested were taken from populations in southern Turkey, sample set 1 consists of 33 individuals of Origanum onites (from nearby Manavgat, province of Antalya), sample set 2 comprises 27 individuals of Origanum majorana (same location). Sample set 3 is a mixed set of 22 individuals from a total of 11 populations from Southern Turkey (Origanum onites and Origanum majorana).
The 10 μl PCR reaction volume contained 5.6 μl H2O dest., 0.4 Units Taq HOT FIREPol® polymerase, 1 μl Buffer B (Solis BioDyne, Tartu, Estonia), 1.4 μl MgCl2 (25 mM), 0.2 μl DMSO (14.08 M), 0.1 μl dNTP-mix (10 mM), 0.1 μl of each Primer (10 µM) and 0.6 μl fluorescent dye BEBO 36 μM (TATAA Biocenter ). 1 μl DNA solution was added to each reaction, containing between 0.25 and 0.8 ng/μl DNA. The mixed samples also contained 1 μl DNA solution with the two compounds at a ratio of 1:1 or 1:2. All reactions were done in duplicate.
The PCR cycling started with an initial phase of 15 min (for the Taq HOT FIREPol® polymerase) at 95°C, then 40 cycles of 10 s at 95°C, 20 s at 60°C and a 20 s elongation step at 72°C. High resolution melting was carried out immediately following PCR from 70°C to 90°C at steps of 0.05°C, each step with a 1 s hold.
For the locus SSR214 the forward primer is 5'-TGTTTGGTGGAAACCGATCC-3', the reverse primer is 5'-AGACGACGAGCTCCAATAACG-3', and the repeat motif is GAT. For the locus SSR244 the forward primer is 5'-TCAAGGGTAGAGCTGCTGCAG-3', the reverse primer is 5'-GCTTTACGGAGGAAGAATGGG-3', and the repeat motif is (GAT)3GAA(GAT)4. 
For verification 20 samples were selected for sequencing. Their PCR products were checked on 1.4% agarose gels before being purified with EXO1 and SAP (Fermentas GmbH, St. Leon-Rot, Germany) according to the manufacturer's instructions. Sequencing of both strands was performed using BigDye Terminators (Applied Biosystems, Brunn am Gebirge, Austria) and primers from the original amplifications. The sequences were generated with an ABI 3130x automated sequencer (Applied Biosystems, Brunn am Gebirge, Austria) and edited using Chromas Vers. 2.24 (Technelyseum, Tewantin, Australia).
Observed and expected heterozygosities as well as the fixation index (an index ranging between -1 and +1, values close to zero indicate random mating), were calculated with the software Microsatellite analyzer MSA 4.05 .
In order to demonstrate the usability of HRM for analysing microsatellites, two microsatellites were chosen from a set of recently published sequences , one with low complexity (three alleles), and the other more complex with seven alleles in the sample sets.
With this approach the characteristics of the two loci in the sample sets of Origanum sp. were elaborated.
Results of microsatellite allele analysis
Sample set 1
Sample set 2
Sample set 3
The experiment demonstrated that the strategy to setup codominant microsatellite analysis worked perfectly well. With known and thoroughly tested microsatellite markers of grapevine HRM was recently implemented by MacKay et al. In species or taxa where no such markers exist yet, like Origanum, the lack of certified reference DNA and the uncertainty of alleles present makes an elaborated development strategy for melting curve analysis necessary. To overcome the difficulties with correctly assigning heterozygous samples, the artificial mixes proposed proved to be accurate. The bottleneck in the setup of such a comparative analysis of dissociation curves is the number of different alleles in a given subset of samples. Therefore the sample size to setup this approach has to be increased when the number of heterozygous genotypes is high in order to correctly identify alleles, since the presence of at least some homozygous individuals is necessary to follow this approach. Considering this fact together with the comparative inexpensiveness and ease of use, HRM can be classified as a method with a mid-range capacity. With the ongoing advances in PCR-technology and miniaturization , HRM has the capacity to develop into a high throughput method.
Concerning the number of alleles in a population, there is no limit for analysis with HRM as long as a sufficient number of homozygous controls is available to mix with. The only case, which is not entirely resolvable with mixes is the occurrence of two rare alleles in a heterozygous genotype never occurring in a homozygous form. Sequencing such genotypes as well as unexpected genotypes can fill this gap.
As recently shown  interrun comparability of melting profiles is also manageable without reference samples in each run, which of course expands the number of samples that can be tested in one run. Furthermore a database utility of melting plots is already envisaged by MacKay et al .
To further increase sample throughput, a low number of alleles would allow sample pooling and could enable even higher throughputs with today's standard equipment. For correct classification of pooled sample melting curves it would be necessary to run mixed samples of known genotypes as calibrators or to re-run samples with ambiguous curves individually. A slightly modified strategy could be used to setup the analysis of organisms with higher ploidy levels. Here, quantitative variation between heterozygous and homozygous samples ('curve convergings') could be used to assign the correct allele composition.
HRM also reveals additional mutations in putative microsatellite loci, which can be seen as a great advantage in comparison to electrophoretic analysis, since the alleles of imperfect microsatellites often can not be correctly classified by their length, due to electrophoretic homoplasy of additional SNPs . Additional mutations may also occur quite frequently in the flanking regions of microsatellites. In this aspect melting curve analysis has a great potential in the study of loci involving both SSR and SNPs , expanding the observed polymorphisms and thus lead to deeper insights into population structures .
With the implementation of a systematic comparison with artificial mixes of samples – as shown herein – HRM becomes a useful tool for the fast, reliable and cheap analysis of codominant markers. The high sensitivity is a major advantage over electrophoretic analysis when unknown mutations occur. Hence HRM has the potential to substitute other techniques in microsatellite analysis.
This work has been carried out with financial support from the FWF Austrian Science Fund (project no. P18027-B06), which is gratefully acknowledged.
Furthermore we thank Christian Schlötterer for helpful discussions.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.