Introduction

Isocitrate dehydrogenase (IDH) is a component of the citric acid cycle which catalyzes oxidative decarboxylation of isocitrate. Two isocitrate dehydrogenases have been reported which utilizes NADP(+) as the electron acceptor one of which is mitochondrial (IDH2) and the other predominantly cytosolic (IDH1). IDH utilizes isocitrate as a substrate to convert NADP+ into NADPH, which is necessary for nucleotide and fatty acid biosynthesis and cellular antioxidant activity. It has been shown that mutations in IDH2 result in loss of function towards its ability to use isocitrate as a substrate. However, mutated form of IDH2 is more effective in converting alpha ketoglutarate into an oncogenic metabolite 2-hydroxy glutarate (2-HG) [1–3]. A recent study has shown high levels of 2-HG in IDH mutants as compared to wild-type IDH. No significant difference was found in the levels of 2-HG between IDH1 and IDH2 mutated patients [4]. Arginine-based mutations in IDH1 and IDH2 genes have been reported in AML by whole genome sequencing [5, 6]. The available literature shows that 15–33 % of AML patients harbor in their leukemic cells somatic mutations in IDH genes [7–9] with IDH2 mutations occurring more frequently than IDH1 (19 and 6–14 %, respectively)[9, 10]. Mutations in IDH2 gene are located in exon 4 at two positions, i.e. codon R140Q and codon R172K.

The prognostic impact of IDH2 mutations in AML is controversial. However, recent findings showed that IDH2-mutated AML patients with normal karyotype (AML-NK) tend to show better overall survival as compared to patients with other IDH AML-NK mutations [11, 12]. Given the prevalence and prognostic relevance of IDH2 mutations in AML, there is a need to develop simpler techniques for identifying these alterations in routine diagnostic screening of AML. Currently available methods are direct sequencing and PCR-based high-resolution melting (HRM) analysis [13]. Here, we developed an allele-specific PCR (ASO-PCR) technique for the detection of IDH2 R140Q and R172K point mutations.

Patients and methods

One hundred and twenty AML patients (males 71, females 49, median age 69.5 years, range 60–90 years) consecutively diagnosed at the Department of Biopathology, Policlinico Tor Vergata, University of Rome, were included in the study. It has previously been reported that frequency of IDH2 mutations increases with the age [7]; therefore, only elderly AML patients were selected in order to increase the potential number of mutated cases. Written informed consent was obtained from all patients according to the declaration of Helsinki and the study was approved by the IRB of Policlinico Tor Vergata.

DNA Extraction and ASO-PCR

DNA was extracted from Ficoll-Hypaque-isolated mononuclear cells obtained from patient bone marrow samples using Nucleospin DNA extraction kit (Macherey-Nagel, GmbH & Co., Germany). Specific forward primers containing a mismatch to improve specificity and a common reverse primer were designed to specifically amplify IDH2 (R140Q) and IDH2 (R172K) mutations on exon 4. ABL was used as an internal control (using ABL-ENF as forward and ABL-ENR as a reverse primer). Primer sequences and PCR product size obtained after amplification are shown in Table 1. One hundred nanograms of DNA was amplified in a total volume of 20 μl using HotStarTaq-Plus Master-Mix Kit (QIAGEN, Germany). Ten picomoles of each IDH2 specific and common primer and 5 pmol of forward and reverse ABL primers were included. The PCR conditions included preheating of the mixture at 95 °C for 5 min followed by 35 cycles for 30 s at 95 °C, annealing for 1 min at 62 °C for IDH2 (R140Q), and 1 min at 60 °C for IDH2 (R172K); elongation for 45 s at 72 °C with a final extension of 5 min was carried out at 72 °C. PCR products were visualized by electrophoresis on a 2 % (w/v) agarose gel by loading 5 μl of PCR product and then the remaining PCR product was purified using a QIAquick PCR purification kit (Qiagen, Chatsworth, CA) for sequencing analysis.

Table 1 Primer sequences used in the study
Full size table

Direct DNA sequencing

All patient samples showing IDH2 mutations by ASO-PCR were subject to direct sequencing using the primers IDH2F and IDH2R [9] (Table 1). In brief, all purified amplicons were directly sequenced using the BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) in conjunction with GeneAmp 9700 PCR Systems (Applied Biosystems). Each 10 μl of the sequencing reaction contained the following: 1 μl of BigDye v3.1 (Applied Biosystems), 2 μl of BigDye® Terminator v1.1/3.1 Sequencing Buffer (5×), 1 μl of PCR primer (5 μM), approximately 3 ng/200 bp of purified PCR product, and enough Gibco distilled water (Invitrogen, Grand Island, NY) to bring the total volume to 10 μl. Thermocycling parameters for PCR product sequencing were as follows: 1 min at 96 °C, 10 s at 96 °C, 5 s at 50 °C, and 4 min at 60 °C for 40 cycles. Sequencing reaction products were purified using the Centri·Sep columns (Applied Biosystems) according to the manufacturer’s recommendations. Samples were diluted with 16 μl of ABI HiDi Formamide (Applied Biosystems) and resolved on an ABI 3130 automated sequencer (Applied Biosystems).

Results and discussion

ASO-PCR allowed us to identify IDH2 mutations in 22 out of total 120 (18.3 %) AML cases, a frequency similar to that published by Paschka et al. in a larger study [8].

Of these, 17 harbored IDH2: R140Q (14.16 %) and 5 IDH2: R172K (4.16 %) mutations. Direct DNA sequencing in both directions was performed in all 120 (wild-type and mutated) cases to confirm the results obtained by ASO-PCR. Following Sanger sequencing, no other substitutions were detected in exon 4 of IDH2 except for IDH2 140 or IDH2 172 positions. All mutated patients of IDH2 140 revealed the substitution CGG → CAG and IDH2 172 showed the substitution AGG → AAG. The DNA sequencing results of IDH2: R140Q and IDH2: R172K in two representative patients are shown in Fig. 1

Fig. 1
figure 1

(A1) ASO-PCR amplification band of 376 bp containing the mutation of IDH2 (R140Q) is visualized in lane 1. Lanes 2–5 show the amplification band of about 686 bp containing the IDH2 wild-type sample showing only amplification of the ABL gene used as internal control. Lane 6 is the blank (without DNA). (A2) Sequence result showing the IDH2 gene substitution CGG → CAG. (B1) Amplification band of 277 bp containing the mutation of IDH2 (R172K) is visualized in positive patients sample in lane 1. Lane 2–6 shows the IDH2 wild-type amplification band. Lane 7 is the blank (without DNA). (B2) Sequence result showing the substitution AGG → AAG

Full size image

Clinical and biologic features of AML patients at diagnosis according to IDH2 mutations are shown in Table 2. We observed that (1) IDH2 R140Q mutation is more frequent (13.6 %) than IDH2 R172K (4 %); (2) out of 15 NPM1 type A mutated patients, five (33 %) were also positive for IDH2 R140Q; (3) with regard to FLT3-ITD positive patients (n = 13), two patients had IDH2 R140Q mutation. We did not find any IDH2 R172K-positive patient carrying either NPM1 type A or FLT3-ITD mutation.

Table 2 Patient’s detail
Full size table

To assess the sensitivity of the ASO-PCR assay, we performed serial dilution experiments using DNA of IDH2 wild-type and mutated patients. The median blasts at diagnosis in IDH2-mutated patients were 80 %. R140Q was weakly detected at 10−3 but amplification was clearly evident up to 10−2 dilution, whereas in case of R172K amplification was clearly visible until the 10−3 dilution (Fig. 2a, b). Direct sequencing of serially diluted IDH mutated patient’s DNA did not reveal any positivity from 10−1 to 10−5 dilutions while undiluted DNA showed double peaks confirming positivity for R140Q mutation. In case of R172K mutation, the double peak remains detectable until 10−1 dilution (Fig. 3a, b)

Fig. 2
figure 2

ASO_PCR: serial dilution of patient DNAs containing IDH2 (R140Q) (a) or IDH2 (R172K) (b) mutation. Lane 1 shows negative control; lane 2, undiluted positive control. Lanes 3–7 mutated samples with dilutions of 10−1 to 10−5. Lane 8 shows the blank (without DNA)

Full size image
Fig. 3
figure 3

Sanger sequencing in serially diluted DNA samples. a IDH2 (R140Q): double peak is visible only in 4 ng/μl DNA; b IDH2 (R172K): mutation is detectable till 10−1 dilution

Full size image

Following the discovery of IDH mutations in 2009, more than 40 additional studies on this alteration and its prognostic significance in AML have been reported in literature [14, 15]. In recent past, several studies have reported various strategies for IDH2 mutation detection yet no study has published ASO-PCR-based detection of IDH2 mutations. This report in terms of its simplicity is equivalent to the one designed by Baxter et al. [16] for JAK2 mutational screening in myeloproliferative syndromes, which has become extremely popular and widespread due to its easy execution and reliability. Patel et al. [13] recently utilized HRM curve analysis for IDH1 and IDH2 mutational screening in 146 AML patients and identified 12 patients with IDH2 R140Q and 4 patients IDH2 R172K mutations. In conclusion, the ASO-PCR method described here represents a specific and sensitive method for the screening of IDH2 (R140Q and R172K) mutations in AML patients at diagnosis. Our method is convenient, easily applicable, and rapid. Finally, this ASO-PCR-based detection method for IDH2 mutations would be useful in countries with limited resources and no access to either DNA sequencer or DHPLC instruments.