Introduction

Nucleophosmin (NPM) is a phosphoprotein that is mainly expressed in the nucleolus. It shuttles between the nucleus and cytoplasm and is involved in many cellular activities. Mutations in the exon 12 of the NPM1 gene have been recently detected in 50–60% of acute myelogenous leukemia (AML) patients with a normal karyotype. AML patients with mutated NPM1 differed from those without NPM1 mutations in terms of clinical features, response to remission induction chemotherapy, and prognosis [1–5]. Therefore, NPM1 study is important for the diagnostic classification, research of pathomechanisms, detection of minimal residual diseases, prognosis, and choice of treatment target. In this study, in order to detect NPM1 gene mutations, DNA from the bone marrow (BM) of untreated adult AML patients from the Chinese mainland were examined by using the TaqMan® MGB probe-based real-time quantitative polymerase chain reaction (RQ-PCR) and sequence analysis. Simultaneously, the detection of internal tandem duplications (ITD) in the FLT3 gene, chromosome examination, and immunophenotyping were also performed.

Materials and methods

Patient samples

This study involved untreated 220 adult AML patients. Diagnosis was established on the basis of the widely used WHO criteria. Informed consent was obtained from all the patients prior to their enrollment in the study. The study design adhered to the principles of the Helsinki Declaration and was approved by the ethics committees of our hospital.

Analysis of NPM1 mutations

Mononuclear cells were isolated from the BM and peripheral blood samples by using standard Ficoll-Hypaque density gradient centrifugation. DNA and RNA were extracted using DNAzol and TRIzol kits (Invitrogen, Carlsbad, CA, USA), respectively, according to the manufacturers’ instructions. Complementary DNA (cDNA) was synthesized as described previously [6]. Mutations in the exon 12 of the NPM1 gene (types A, B, and D) were detected using RQ-PCR [7]. The primers for the albumin (ALB) gene, which was the internal control, TaqMan® probes, and mutated regions of the NPM1 gene were designed using Primer Express 2.0 software (Applied Biosystems, Foster City, CA, USA). The primers for the ABL gene, which was the internal control for cDNA template, was synthesized as described previously [8]. All the primers and probes used in the study are listed in Table 1. The PCR mixture contained 1× TaqMan® Universal PCR Master mix, 900-nM primers, 250-nM fluorescent probes, and 150–500 ng DNA. PCR was performed by using an ABI PRISM® 7500 Sequence Detection System by employing the following protocol: 50°C for 2 min, 95°C for 10 min, and 50 cycles at 95°C for 15 s and 60°C for 1 min (when cDNA was used as the template, the conditions were 62°C for NPM1 mutation A and 59°C for NPM1 mutation B).

Table 1 Primers and probe sequences for detecting NPM1 mutations by real-time quantitative polymerase chain reaction
Full size table

Preparation of plasmid standards

The ALB gene and NPM1 mutation A gene were amplified using the DNA isolated from the patients with the NPM1 type A mutation. The PCR products were purified and subcloned into the pMD18-T vector (Takara, Dalian, China). Transformation, screening, and sequencing were then performed to obtain ALB and NPM1 mutation A plasmid standards. The copy number was calculated based on the optical density (OD) value, and the standards were subjected to a tenfold dilution (101–106 copies) in order to plot a standard curve. For each measurement, the threshold of amplification for the curve was set at 0.08, and it included a positive control (samples from the patients with the NPM1 type A mutation), blank control (without template), and negative control (samples from patients without NPM1 mutation). ABL was used as the internal control when cDNA was used as the template, and the ABL plasmid standard was prepared as described previously [8].

Calculation of the NPM1 mutation level

The copy numbers of ALB and NPM1 mutants were calculated using the C t value and standard curve. Our previous study revealed similar efficacies for the amplification of ALB and NPM1 mutation A genes (the slope of standard curve was −3.45 and −3.42, respectively). In order to decrease the differences in plasmid quantification, only one ALB standard curve was plotted (Fig. 1). The NPM1 mutant copy number in 100 ALB copies was used as the NPM1 mutant gene content. If the ALB copy number was ≥3 × 104, the samples were considered for quantitative detection. ABL was used as the internal control gene when cDNA was used as the template. The results were analyzed based on the guidelines proposed by the European Study Group for the RQ-PCR detection of minimal residual diseases [9].

Fig. 1
figure 1

Standard curve of plasmids containing the ALB gene obtained by real-time quantitative polymerase chain reaction. The standard curve shows a linear correlation between the C t value and dilution of the ALB plasmid (correlation coefficients, >0.99)

Full size image

NPM1 gene sequencing

PCR was performed using DNA from the samples as the template and primers (sense: 5′-GGTCTCTGTTCTTTCTGTTGATTTCC-3′ and antisense: 5′-CAACACATTCTTGGCAATAGAACCT-3′). The 50-μl PCR mixture contained 25 μl of 2× Universal PCR Master mix (TianGen Biotech, Beijing, China), 900 nM primers, and 300 ng DNA. The PCR protocol was as follows: pre-denaturing at 95°C for 5 min, followed by 30 cycles at 95°C for 40 s, 55°C for 40 s, and 72°C for 1 min, and a final extension at 72°C for 10 min. The purified PCR products were subjected to sequence analysis with an ABI3700 DNA sequencer. The newly screened mutants were subcloned into pMD18-T vectors, followed by transformation, screening, and further sequencing.

Screening of the ITD mutations in the FLT3 gene

The presence of an FLT3/ITD was determined by amplifying a region spanning exons 14 and 15 by using the primers 14F (5′-CAA TTT AGG TAT GAA AGC C-3′) and 15R (5′-GTA CCT TTC AGC ATT TTG AC-3′) followed by 8% polyacrylamide gel electrophoresis as described previously [10].

Immunophenotyping studies

Freshly obtained BM and peripheral blood samples were analyzed using well-established methods [11] involving the four-color immunofluorescence technique for the simultaneous staining of surface and cytoplasmic antigens.

Cytogenetics

Cytogenetic G-banding analysis was performed using standard methods [12]. The definition of a cytogenetic clone and descriptions of karyotypes followed the International System for Human Cytogenetic Nomenclature.

Response to remission induction treatment

Among the 220 patients analyzed, 43 had acute promyelocytic leukemia (APL). APL has been considered as a separate disease entity among AML subtypes, and the treatment of all-trans retinoic acid has dramatically improved its clinical outcome. In addition, data of 65 patients with AML, excluding those with APL, were not analyzed due to their treatment with independent regimens or unavailable data. We therefore analyzed the response to treatment of 112 patients with AML who were treated with the protocols of daunomycin plus cytarabine or homoharringtonine plus cytarabine for induction therapy.

Statistical analysis

Statistical analysis was performed using SPSS software version 11.0 (Chicago, IL, USA). Differences between two groups were evaluated using one-way analysis of variance. Chi-square or a two-sided Fisher exact test was used to calculate the significance of association between NPM1 mutation and other discrete variables. Values of p below 0.05 were considered to be significant.

Results

NPM1 mutation detection sensitivity

Plasmid standards of ALB and NPM1 mutation A were prepared. Plasmid DNA concentration was determined by absorbance measurement; six serial plasmid dilutions (106, 105, 104, 103, 102, and 101 copies) were amplified by RQ-PCR to construct a standard curve for the absolute quantitative assessment of copy number. The results for amplified the ALB plasmids revealed that all correlation coefficients were above 0.99 (Fig. 1); the average C t values for the amplified NPM1 mutation A with copy numbers of 106–101 copies were 19.42, 22.71, 26.40, 29.24, 33.26, and 35.39, respectively, and the coefficients of variation were 1.3%, 2.9%, 0.8%, 1.1%, 1.6%, and 1.8%, respectively; the sensitivity of detection was ten copies (Fig. 2). Eight DNA samples of NPM1-mutation-positive patients were prepared with a tenfold dilution series (100, 10−1, 10−2, 10−3, 10−4, and 10−5). In all cases, the C t value was in the linear range for all dilutions. In addition, the mutations were detected with a sensitivity of 10−4 or 10−5, and the correlation coefficient was above 0.99 (Table 2). High correlation coefficients allowed accurate assessment of the quantity of NPM1 in unknown samples. All amplification plots were analyzed by positioning the threshold at 0.08.

Fig. 2
figure 2

Representative amplification plot of the six plasmid dilutions that were tested. The plasmid contains NPM1 mutation A. The sensitivity for detection was ten copies. The coefficient of variation (CV) of the C t value at each dilution was <3%

Full size image
Table 2 Features of the amplification curves obtained with a tenfold dilution series for patients with NPM1 mutations
Full size table

Detection of NPM1 mutations with RQ-PCR

NPM1 gene mutations in 220 BM DNA samples of untreated AML patients were detected by RQ-PCR with the TaqMan® MGB probe (Table 3). The detection rate of NPM1 mutations in AML patients (220 cases) was 16.4%, while that of M2 (107 cases), M3 (43 cases), and M4/M5 (52 cases) were 14.2%, 2.3%, and 30.8%, respectively. However, only one case each of NPM1 mutation was detected in the M0 (four cases), M1 (seven cases), M6 (five cases), and M7 (two cases) groups. In 36 NPM1-mutation-positive specimens (Table 4), the average NPM1 mutant copies/ALB copy was 19.7% (2.3–66.6%); background was amplified in eight cases whose C t values were above 42.

Table 3 Clinical data, FAB subtype, genotypic changes, and outcome in AML patients
Full size table
Table 4 Clinical features of NPM1-mutation-positive patients
Full size table

Eight NPM1-mutation-positive patients were followed -up after treatment; among them, five continued to test negative for the NPM1 mutation (Fig. 3) and three patients remained negative even after BM transplant. In one case, the NPM1 mutation copies/ALB copies (%) value was 0.013% on remission, implying a decrease of 2 log as compared with the baseline. In the three cases that were not alleviated even after inductive treatment, the NPM1 mutation copies/ALB copies (%) value was 8.8%, 7.3%, and 2.4%, thus indicating no obvious decline as compared with the baseline. Simultaneously, NPM1 mutation A and NPM1 mutation B had been detected in 56 cDNA specimens, and ten NPM1-positive patients exhibited higher levels of NPM1 mutations (346.7–4,075.9 NPM1 copies/100 ABL copies; average, 1,272.7). However, among 46 NPM1-mutation-negative patients, all NPM1 mutation B cDNA amplifications were negative, 34 NPM1 mutation A amplification cases were negative, while nonspecific amplification curves were observed in 12 other cases (threshold, 0.08; C t value, 30.3–34.5; average, 31.9). Because the number of NPM1 mutations at the DNA level represented the number of leukemia cells and the preparation of DNA samples was easier than that of cDNA, we used DNA samples for NPM1 mutation detection in this study.

Fig. 3
figure 3

NPM1 mutated copies/ALB copies (%) at diagnosis, after induction therapy, and during follow-up in AML patients with NPM1 mutations. NR no hematological remission, Rel. hematological relapse, SCT allogene haemopoietic stem cell transplantation

Full size image

NPM1 sequencing results

Sequence analysis confirmed that all these mutations were heterozygous for a 4-bp insertion at the 959 site. Type A mutations accounted for 80.6% (29/36) of the total mutations, type B for 11.1% (4/36), type D for 0.0%, and CTCG and CAAG (BJ-01 and BJ-02, respectively) insertions were the two new mutations detected in the other two cases. Apart from the 35 cases in which the NPM1 mutation was detected by PCR, only one case with the DD-3 mutation was detected by sequencing the other 55 normal karyotype specimens. This implies that the primer used for amplifying NPM1 types A, B, and D mutations can also be used for screening BJ-01 and BJ-02. However, some rare NPM1 mutations, like DD-3, require sequencing or other methods for detection.

Characteristics of NPM1-mutation-positive patients

NPM1 mutations were found in patients with AML of all FAB subtypes (Table 3). Patients with AML M4/M5 subtype had a higher NPM1 mutation rate than those with M2 or M3 subtype (p < 0.05, p < 0.001, respectively). Cytogenetic data were available for 135 patients. The detection rates of NPM1-positive mutations in specimens exhibiting in the normal and abnormal karyotypes were 38.6% (22/57) and 5.1% (4/78), respectively, which was significantly different (p < 0.001). Moreover, NPM1 mutations have been detected in three patients with t(8;21), t(15;17) or inv(16) (Table 3). NPM1 mutations were significantly associated with higher initial white blood cell counts in the peripheral blood and age, but not sex, blast in the bone marrow, hemoglobin and platelet counts in the peripheral blood.

Among the 36 NPM1-mutation-positive patients, 19 were males and 17 were females; of them, 38.9% (14/36) of the patients coherently exhibited FLT3/ITD mutations, and 66.7% (22/33) did not express CD34 on cells (the case was regarded as CD34 negative on cells when the percentage of express CD34 on blast cells is lower than 20%), as shown in Table 4. Of the 112 patients with AML, 84(75.0%) achieved complete response (CR) after induction chemotherapy. The CR rate was higher in the patients with NPM1 mutation (23 of 27, 85.2%) than without them (61 of 85, 71.8%), but no statistical significant difference was observed (p = 0.161; Table 3).

Discussion

NPM is a phosphoprotein with extensive expression that is present in abundance. It is involved in many cellular activities, including ribosomal biosynthesis and processing and the maintenance of genome stability. Abnormal NPM expression and/or its localization in cells may be involved in tumorigenesis by different mechanisms. Recent researches have indicated that mutation could occur in the exon 12 of the NPM1 gene. Currently, more than 40 different types of NPM1 mutations have been detected [13], and types A, B, and D are the most common. All these three types of mutations follow a similar pattern wherein a 4-bp sequence is inserted at the 959 site, causing a frame shift mutation that leads to a change in the amino acid sequence downstream of the mutation site.

By using RQ-PCR with a TaqMan® MGB probe, 220 BM DNA samples of untreated AML patients were screened for three types (A, B, and D) of NPM1 gene mutations. The results demonstrated that the detection rate of NPM1 mutations was 16.4% in 220 untreated AML patients and that in M2 (107 cases), M3 (43 cases), and M4/M5 (52 cases) cases was 14.2%, 2.3%, and 30.8%, respectively. Only one case each of NPM1 mutation was detected in the M0 (four cases), M1 (seven cases), M6 (five cases), and M7 (two cases) groups. The detection rate of our study was slightly lower (16.4%, 36/220) than that of studies conducted in China Taiwan (19.1%, 33/173) [14]. This might be because the primers we used only amplified the A, B, and D mutations and more M3 subtype in 220 AML patients which probably affects the frequency of NPM1 mutations. The primers were thought to amplify 90–95% of the NPM1 mutations (type A, approximately 75%; types B and D, approximately 15–20%); however, a small number of mutations were left out. Using PCR, we detected 35 cases with NPM1 mutations, while there was one case with the DD-3 mutation that was detected by sequencing the other 55 normal karyotype samples. This implies that the primers used for amplifying the A, B, and D types of mutations could also be used for screening BJ-01 and BJ-02 [our results confirmed that the type A mutation accounted for 80.6% (29/36); type B, for 11.1% (4/36); type BJ-01, for 2.8% (1/36); and type BJ-02, for 2.8% (1/36)]. However, some rare NPM1 mutations, like DD-3, need sequencing or other methods for their detection.

Thiede et al. [15] studied 1,485 AML patients and found that AML patients that tested positive for NPM1 mutations often had a normal karyotype, lacked hematopoietic stem cell/progenitor cell surface markers such as CD34, and also exhibited a high frequency of FLT3/ITD mutations; these observations were in accordance with our results. Based on these observations, the AML patients were divided into four groups [15, 16]: simple NPM1 mutation group, simple FLT3/ITD mutation group, double NPM1 and FLT3 mutation group, and wild-type NPM1 or FLT3 mutation group. They found that in the simple NPM1 mutation group, the overall survival and disease-free survival rates were much higher, while the cumulative recurrence rates were lower. Therefore, the NPM1 mutation was considered to comprise adult AML patients with common gene abnormality. In the absence of FLT3/ITD mutations, NPM1 mutations may be one of the criteria to determine a patient’s response to treatment. Suzuki et al. [3] also analyzed the prognostic value of NPM1 mutations and found that the CR rate was significantly higher in the patients with NPM1 mutation (42 of 49, 85.7%) than in those without (97 of 141, 68.8%; p = 0.025). Our results showed that the CR rate was higher in the patients with NPM1 mutation (23 of 27, 85.2%) than without them (61 of 85, 71.8%), but no statistical significant difference was observed (p = 0.161) probably because of the small sample size.

In Italy, Gorello et al. [7] analyzed mutations in the NPM1 gene in 13 AML patients by RQ-PCR and found that these mutations could be used to monitor and quantify the minimal residual diseases. Our results showed that the average number of the NPM1 mutation copies/ALB copies in 35 untreated patients was 19.7% (2.3–66.6%). Of the eight patients screened for NPM1 mutations at follow-up after treatment, five patients in remission continued to test negative for NPM1 mutations within 2–14 months of follow-up. The average number of NPM1 mutation copies/ALB copies (%) in one case with the ITD mutations in the FLT3 gene was 0.013% at remission, which implied a decrease of 2 log as compared with the baseline, rose, and reached 15.01% at clinical relapse, while three cases showed no alleviation even after inductive treatment. The number of NPM1 mutation copies/ALB copies (%) was 8.8%, 7.3%, and 2.4%, with no obvious decline as compared with the baseline.

At present, NPM1 gene mutations are one of the most extensive and specific gene mutations in a normal karyotype. In AML patients, therefore, the detection of NPM1 mutations might be the only means to monitor AML alleviation at the molecular level, minimal residual leukemia, and recurrence [17].

We also found that RQ-PCR was a reliable and sensitive method for detecting NPM1 mutations. It not only detects NPM1 gene mutations in AML patients and monitor minimal residual diseases by quantitative analysis of NPM1 mutations but also plays a role in predicting the response to treatment, in the assessment of prognosis, and research related to the pathomechanisms underlying AML.