Rapid identification of IDH1 and IDH2 mutations in acute myeloid leukaemia using high resolution melting curve analysis
S. M. Noordermeer,E. Tönnissen,Inge Vissers,A. V. D. van der Heijden,L. T. van de Locht,P. Deutz‐Terlouw,E. Marijt,J. Jansen,B. A. van der Reijden
DOI: https://doi.org/10.1111/j.1365-2141.2010.08423.x
2011-02-01
British Journal of Haematology
Abstract:Screening for recurrent mutations in leukaemia is becoming increasingly important because many of them have an impact on disease outcome. Nowadays, mutational detection is mostly based on DNA sequencing of PCR products or QPCR using allele specific probes. With the increasing identification of recurrent mutations in cancer, the development of fast and efficient approaches for mutational screening is needed. An alternative technique for mutation detection is high resolution melting (HRM) analysis (Wittwer, 2009). HRM analysis is a fast ‘single-well-technique’ combining PCR using a fluorescent saturating dye that is intercalated in the DNA, and melt curve analysis afterwards. As mutations cause different melting properties, they are identified when compared to non-mutated samples. Positive samples in HRM analysis can be sequenced subsequently to determine exact nucleotide changes. Recently, mutations in the NADP-dependent isocitrate dehydrogenase genes IDH1 and IDH2 were shown in acute myeloid leukaemia (AML), myeloid dysplastic syndromes and myeloproliferative diseases (Mardis et al, 2009; Abbas et al, 2010; Chou et al, 2010; Green & Beer, 2010; Kosmider et al, 2010; Marcucci et al, 2010; Wagner et al, 2010). We study the application of HRM analysis and direct sequencing to screen for these mutations in a cohort of 168 AML patients, using the real-time PCR platform of Applied Biosystems (Carlsbad, CA, USA). Sequence analysis detected IDH1 or IDH2 mutations in 32 patients (19%) (Tables SI and SII). Using the methods as described in the Data S1, mutations could also be detected by HRM analysis (used primer combinations are indicated in Table SIII). Different primers with and without M13sequences were tested, showing differential results for IDH1 or IDH2 mutations (Fig 1). Using HRM analysis, no mutations were missed. For IDH1 R132, 12 mutated samples (7Æ1%) were found both by HRM and sequence analysis while an additional three positive samples (1Æ8%) were found only by HRM analysis. To elucidate whether the extra mutations found by HRM analysis were false positives, we compared the sensitivity of both techniques. For this, a dilution series of genomic DNA from a mutated sample was measured by HRM analysis followed by direct sequencing. Positivity in HRM analysis was lost at an allele frequency of approximately 9%, whereas the mutation could still be detected at a percentage of 4% by sequencing (Fig S1). As sequencing was more sensitive, this suggests that the extra positive samples found by HRM analysis, that did not contain any SNPs, represent false positive results. Nineteen samples were found positive for the IDH2 R140 mutation (11Æ3%) by sequence analysis. Initial HRM analysis for this mutation identified many false positive results. However, exclusion of samples with low PCR yields (Ct>26Æ8, eight samples) significantly improved the discrimination between mutated and wild type samples. After exclusion, all positive samples found by sequencing were confirmed by HRM analysis while two cases remained false positive. R172 IDH2 mutations were found in two patients (1Æ2%) both by HRM and sequence analysis and two false positive samples were found by HRM (1Æ2%) (Tables SI and SII). One patient showed both an IDH1 R132C and IDH2 R140Q mutation. Allele frequencies of the mutations were measured by pyrosequencing (primer combinations are specified in Table SIV). The allele frequency of the IDH1 R132C mutation was 15Æ5%, while the IDH2 R140Q frequency was even lower (data not shown). Because IDH1 and IDH2 mutations are typically heterozygous and the total allele frequencies of both mutations was below 50%, we cannot conclude whether the mutations represent one and the same or independent clones. One patient exhibited a homozygous IDH2 R140W mutation, identified by both sequence and HRM analysis (Fig S2). SNP array analysis showed that the homozygous mutation was caused by uniparental disomy (UPD) on a large part of chromosome 15, including the IDH2 locus (Fig S3). Of the 32 patients with IDH mutations, we found cooccurrences with other chromosomal aberrations confirming data of other studies (Chou et al, 2010; Kosmider et al, 2010; Marcucci et al, 2010; Paschka et al, 2010; Wagner et al, 2010). Furthermore, none of the IDH1 or IDH2 mutated patients showed overexpression of MECOM (also known as EVI1) (Table SI). Low frequencies of mutated alleles may be missed using sequencing or HRM analysis. To test whether IDH1 mutations occur at frequencies under the detection limit of HRM analysis, we designed an allele specific QPCR for the most frequently occurring IDH1 R132H mutation (for primer