K-ras mutations and N-ras mutations in childhood acute leukemias with or without mixed-lineage leukemia gene rearrangements
It is believed that Ras mutations drive the proliferation of leukemic cells. The objective of this study was to investigate the association of Ras mutations with childhood acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) with special reference to the presence or absence of mixed-lineage leukemia gene (MLL) rearrangements.
Bone marrow samples from 313 children with B-precursor ALL and 130 children with de novo AML were studied at diagnosis. Southern blot analysis was used to detect MLL rearrangements, and reverse transcriptase-polymerase chain reaction (RT-PCR) analysis was used to detect common MLL fusion transcripts. Complementary DNA panhandle PCR was used to identify the infrequent or unknown MLL partner genes. DNA PCR or RT-PCR followed by direct sequencing was performed to detect mutations at codons 12, 13, and 61 of the N-Ras and K-Ras genes.
Twenty of 313 patients with B-precursor ALL and 17 of 130 patients with de novo AML had MLL rearrangements. N-Ras mutations were detected in 2 of 20 patients with MLL-positive ALL and in 27 of 293 patients with MLL-negative ALL (P = 1.000). N-Ras mutations were detected in 2 of 17 patients with MLL-positive AML and in 14 of 113 patients with MLL-negative AML (P = 1.000). K-Ras mutations were present in 8 of 20 patients with MLL-positive ALL compared with 32 of 293 patients with MLL-negative ALL (P = 0.001). K-Ras mutations were detected in 3 of 17 patients with MLL-positive AML compared with 5 of 113 patients with MLL-negative AML (P = 0.069).
Ras mutations were detected in 20.8% of patients with childhood B-precursor ALL and in 17.7% of patients with childhood AML. MLL-positive B-precursor ALL was associated closely with Ras mutations (50%), especially with K-Ras mutations (40%), whereas MLL-positive AML was not associated with Ras mutations. Cancer 2006. © 2006 American Cancer Society.
The activating mutations of the Ras gene or other abnormalities in RAS signaling pathways lead to the uncontrolled growth factor-independent proliferation of hematopoietic progenitors.1, 2 RAS proteins regulate cellular proliferation by cycling between active guanine triphosphate (GTP)-bound and inactive guanine diphosphate-bound states.3, 4 Many hematopoietic growth factors transduce signals from cell membrane to nucleus through RAS members.5 Oncogenic mutations in N-Ras and K-Ras genes have been observed with variable prevalence in hematopoietic malignancies,1 including acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). All point mutations of the Ras gene occurred at different nucleotides of codons 12, 13, or 61. Ras mutations have been studied more extensively in adult AML and less often in childhood AML.6–14 There have been several reports on Ras mutation in childhood ALL; however, those studies used different methodologies, which led to inconsistency in the reported prevalence and impact of Ras mutations on childhood acute leukemias.15–18
Specific chromosomal translocations often are associated with specific types of acute leukemia, except for translocations that involve 11q23, which have been described in both AML and ALL. The gene at 11q23 that is disrupted by chromosomal rearrangement has been cloned and designated as the mixed-lineage leukemia (MLL) gene. The MLL gene rearrangement serves as a particularly illustrative example of the clinical, biologic, and epidemiologic implications of contemporary molecular oncology investigations. A “two-hit” model of leukemogenesis of AML recently has been proposed.19 Very recently, Ono et al. further demonstrated direct evidence of a multistep leukemogenesis mediated by MLL fusion proteins and activated FLT3.20 The MLL gene rearrangement is one of the gene mutations that blocks the differentiation of leukemic cells; conversely, Ras gene mutations drive the proliferation of leukemic cells. There has been only one report that addressed the coexistence of the MLL rearrangement and Ras mutations; however, Mahgoub et al.21 found that Ras mutations were absent in 25 pediatric patients with acute leukemia who had MLL gene translocations, except for 2 patients who had infant monoblastic variants. Determining frequency of the spectrum of Ras mutations according to the status of MLL rearrangement in childhood acute leukemias will require further studies in larger numbers of patients.
Because direct sequencing is the most unambiguous method of detecting Ras mutations, in the current study, we used direct sequencing for each polymerase chain reaction (PCR) product that contained codons 12, 13, and 61 of the N-Ras and K-Ras genes. Our objective was to determine the frequencies of N-Ras and K-Ras mutations in pediatric acute leukemias, with special reference to those with MLL rearrangements, to gain a better understanding of the collaboration of both genetic aberrations.
MATERIALS AND METHODS
Patients and Samples
Three hundred thirteen children (age 18 yrs and younger) with B-precursor ALL and 130 children (age 18 yrs and younger) with de novo AML diagnosed at Mackay Memorial Hospital and Chang Gung Children's Hospital at Linkou were studied at diagnosis. Bone marrow (BM) aspiration was done at diagnosis. Informed consent was obtained from the guardians of the patients. The study was approved by the Institutional Review Board of Mackay Memorial Hospital. BM samples were enriched by Ficoll-Hypaque (1.077 g/mL; Amersham Bioscience, Uppsala, Sweden) density-gradient centrifugation and cryopreserved in 10% dimethyl sulfoxide and 20% fetal bovine serum at − 70 °C or in liquid nitrogen until they were tested. At diagnosis, BM samples underwent Romanowsky system staining, cytochemical staining, immunophenotyping, and cytogenetics. The morphologic subtypes of AML were classified according to the French–American–British Cooperative Group.22–24 Reverse transcriptase (RT)-PCR assay for the detection of fusion transcripts of PML-RARα, AML1-ETO, and CBFβ-MYH11 were performed as described previously,25 For ALL, a multiplex RT-PCR assay was performed for the detection of 4 major fusion transcripts, including TEL-AML1, E2A-PBX1, MLL-AF4, and BCR-ABL, as described previously.26
DNA and RNA extraction
Genomic DNA was extracted from frozen BM cells by using a DNA extraction kit (Puregene Gentra System, Minneapolis, MN) according to the manufacturer's instructions. RNA was extracted and reverse transcribed to complementary DNA (cDNA), as described previously.27
Determination of MLL rearrangements
Southern blot analysis was performed to detect MLL rearrangements.28, 29 RT-PCR was used to detect common MLL fusion transcripts, including MLL-AF4, MLL-AF6, MLL-AF9, MLL-AF10, MLL-ENL, MLL-ELL and partial tandem duplication of MLL (MLL-PTD).30–32 cDNA panhandle PCR technology was used to identify the infrequent or unknown MLL partner genes in patents with MLL rearrangements according to the method described by Megonigal et al.29, 33
Detection of N-Ras mutations
DNA and/or RNA samples were analyzed for point mutations at codons 12, 13, and 61 of the N-Ras gene by using DNA PCR amplification of exon 1 and 2.34 The DNA-PCR assay was performed on a mixture of 25 μL containing 100 ng DNA with 0.625 U Taq DNA polymerase (Invitrogen, Carlsbad, CA), 1 × PCR buffer, 1.5 mM MgCl2, 200 μM deoxyribonucleotide triphosphate (dNTP), and 1 μM each forward primer (5′-GAC TGA GTA CAA ACT GGT GG-3′; NRas-F1) and reverse primer (5′-TGC ATA ACT GAA TGT ATA CCC-3′; NRas-R1) for exon 1, which amplifies codons 12 and 13; and the forward primer (5′-CAA GTG GTT ATA GAT GGT GAA ACC-3′; NRas-F2) and reverse primer (5′-AAG ATC ATC CTT TCA GAG AAA ATA AT-3′; NRas-R2) for exon 2, which amplifies codon 61. The cDNA PCR procedure was similar to that used for DNA PCR, except that only 1 primer set was used (forward primer: 5′-CTG TGG TCC TAA ATC TGT CC-3′; reverse primer: 5′-CAG TGC AGC TTG AAA GTG G-3′). PCR was performed in a 9600 thermal cycler (Applied Biosystems, Foster City, CA) with the following schedule: An initial preheating at 94 °C for 5 minutes was followed by denaturation at 94 °C for 1 minute, 60 °C for 1 minute, and 72 °C for 1 minute for total 40 cycles, and a final extension at 72 °C for 10 minutes. The PCR products were analyzed by electrophoresis on 2% Nusieve agarose gels (BioWhittaker Molecular Applications, Rockland, ME), stained with ethidium bromide, and visualized under an ultraviolet lamp. The PCR bands were cut out from the gel, purified, and directly sequenced in both directions with the BigDye Terminator Cycle Sequencing Ready Reaction kit, which contained AmpliTaq DNA polymerase FS (Applied Biosystems) on an automated ABI PRISM 3730 DNA sequencer (Applied Biosystems) according to the manufacturer's instructions.
Detection of K-Ras mutations
For the detection of K-Ras mutations, the DNA and cDNA PCR assays were performed similar to those described for N-Ras but with different primers, as follows: for the cDNA PCR assay, the K-Ras-c forward primer was 5′-CAT TTC GGA CTG GGA GCG AG-3′, and the reverse primer was 5′-CTA TAA TGG TGA ATA TCT TCA AAT GAT TTA GT-3′. For the DNA PCR assay, the exon 1 K-Ras-E1 forward primer was 5′-GGT GAG TTT GTA TTA AAA GGT ACT GGT G-3′, and the reverse primer was 5′-CCT CTA TTG TTG GAT CAT ATT CGT CC-3′. The exon 2 K-Ras-E2 forward primer was 5′-GGA TTC CTA CAG GAA GCA AGT AGT AA-3′, and the reverse primer was the same K-Ras-c reverse primer that was used for the cDNA PCR assay. Eighty-six samples (58 ALL samples and 28 AML samples) were analyzed with the DNA PCR assay, 345 samples were analyzed with the cDNA PCR assay, and 12 samples were analyzed by using both methods, which yielded identical results.
The Fisher exact test (2-sided) was used to compare the frequencies of N-Ras mutations or K-Ras mutations between different groups of leukemias. For all analyses, the P values were 2-tailed and P values < 0.05 were considered statistically significant.
Ras Mutations in MLL-Positive Acute Leukemias
Of the 313 patients who had B-precursor ALL, 20 patients were positive for MLL rearrangements. The MLL fusion transcripts included 10 patients with MLL-AF4, 7 patients with MLL-ENL, 2 patients with MLL-AF9, and 1 patient with MLL-AF10. All of those patients had the pro-B subtype (CD10 negative). Of the 130 patients who had de novo AML, 17 patients with MLL rearrangements, including 4 patients with MLL-AF9; 6 patients with MLL-AF10; 2 patients with MLL-ENL; 1 patient each with MLL-AF1, MLL-AF4, MLL-ELL, MLL-SEPT6,29 and MLL-PTD. Ten of 20 children who had pro-B MLL-positive ALL harbored Ras mutations, and 8 of those children had K-Ras mutations, including 7 mutations at codon 12 (4 Gly12Asp, 2 Gly12Val, and 1 Gly12Ala) and 1 mutation at codon 13 (Gly13Asp); and 2 patients had N-Ras mutations of Gly12Asp. Five of 17 patients who had MLL-positive AML had Ras mutations, including 2 patients with N-Ras mutations (1 each of Gly12Ala and Gln61Pro), and 3 patients had K-Ras mutations (1 each of Gly12Ala, Gly12Ser, and Gly13Asp). The age, leukemia subtype, MLL fusion transcripts, and patterns of Ras mutations in children with MLL-positive acute leukemias are shown in Table 1.
Table 1. Clinical Hematologic Features and Ras Mutations in Children with Acute Leukemias and Mixed-Lineage Leukemia Gene Rearrangements
Ras Mutations in MLL-Negative ALL
Twenty-seven of 293 patients without MLL rearrangements harbored N-Ras mutations. Mutations at codon 12 were detected in 10 MLL-negative patients, including 7 mutations of Gly12Asp, and 1 mutation each of Gly12Cys, Gly12Ser, and Gly12Ala. Mutations at codon 13 were detected in 7 patients, and all were of Gly13Asp. Among those 7 patients, 1 patient had mutation of both Gly12Asp and Gly13Asp. Mutations at codon 61 of the N-Ras gene were present in 11 patients, including 5 mutations of Gln61Leu, 3 mutations of Gln61Lys, 2 mutations of Gln61Arg, and 1 mutation of Gln61His. Thirty-two of 293 patients without MLL rearrangements harbored K-Ras mutations. Mutations at codon 12 were detected in 14 patients, including 6 mutations of Gly12Asp, 3 mutations of Gly12Val, 2 mutations of Gly12Ser, and 1 mutation each of Gly12Arg, Gly12Cys, and Gly12Ala. Mutations at codon 13 were detected in 15 patients: All 15 of those patients had mutations of Gly13Asp, and 4 of them had concurrent N-Ras mutations of Gly12Asp (1 patient), Gln61Lys (1 patient), and Gln61Arg (2 patients). Mutation at codon 61 (Gln61Pro) was detected in 3 patients.
The frequency of N-Ras mutations for the subtypes of common fusion transcripts was 1 of 60 patients (2%) with t(12;21)/TEL-AML1, 0 of 16 patients with t(1;19)/E2A-PBX1, and 1 of 13 patients (8%) with t(9;22)/BCR-ABL. The frequency of K-Ras mutations was 2 of 60 patients (3%) with t(12;21)/TEL-AML1, 2 of 16 patients (13%) with t(1;19)/E2A-PBX1, and 0 of 13 patients with (9;22)/BCR-ABL.
Ras Mutations in MLL-Negative AML
Fourteen of 113 patients without MLL rearrangements harbored N-Ras mutations: Mutations at codon 12 were detected in 5 patients (2 mutations of Gly12Asp, 2 mutations of Gly12Cys, and 1 mutation of Gly12Ser), mutations at codon 13 were detected in 4 patients (2 mutations of Gly13Cys, 1 mutation of Gly13Arg, and 1 mutation of Gly13Asp), and mutations at codon 61 were detected in 5 patients (2 mutations of Gln61His, 2 mutations of Gln61Leu, and 1 mutation of Gln61Arg). Five of 113 patients without MLL rearrangements harbored K-Ras mutations. Mutations at codon 12 were detected in 4 patients (2 mutations of Gly12Cys and 1 mutation each of Gly12Ala and Gly12Asp). Mutation at codon 13 was present in 1 patient (Gly13Asp) who also had a Gln61Arg mutation of the N-Ras gene.
Of the 11 patients who had inv(16)/CBFβ-MYH11, 3 patients (27%) had N-Ras mutations. All 21 patients who had t(8;21)/AML1-ETO lacked N-Ras mutations. The frequency of K-Ras mutations was 5% (1 of 21 patients) in those who had t(8;21)/AML1-ETO. K-Ras mutations were not observed in patients who had inv(16)/CBFβ-MYH11. Taken together, 4 of 32 patients (12.5%) who had core-binding factor AML had Ras mutations. None of the 14 patients who had t(15;17)/PML-RARα had Ras mutations.
Comparison of the Frequency of Ras Mutations in Childhood Acute Leukemia With and Without MLL Rearrangements
Ras mutations were detected in 20.8% of patients with childhood B-precursor ALL and in 17.7% of patients with childhood AML (Table 2). The frequency of Ras mutations in patients with MLL-positive childhood acute leukemia was compared the frequency of Ras mutations in patients with MLL-negative childhood acute leukemia. The frequency of Ras mutations in MLL-positive B-precursor ALL was significantly higher than that of MLL-negative ALL (50% vs. 18.8%; P = 0.003), whereas there was no difference in the frequency of Ras mutations between MLL-positive AML and MLL-negative AML (P = 0.182). No difference was observed in the frequency of N-Ras mutations between patients with MLL-positive ALL and patients with MLL-negative ALL (P = 1.000). MLL-positive ALL was associated with a significantly higher frequency of K-Ras mutations compared with MLL-negative B-precursor ALL (P = 0.001). The frequency of N-Ras mutations in children with de novo AML did not differ between those with MLL-positive disease (2 of 17 patients) and those with MLL-negative disease (14 of 113 patients; P = 1.000). There was no significant difference in the frequency of K-Ras mutations between patients with MLL-positive AML and patients with MLL-negative AML (P = 0.069).
Table 2. Frequencies of Ras Mutations in Childhood B-Precursor Acute Lymphoblastic Leukemia and Acute Myeloid Leukemia With and Without Mixed-Lineage Leukemia Gene Rearrangements
|N-Ras|| || ||1.000|| || ||1.000|
| Positive||2||27|| ||2||14|| |
| Negative||18||266|| ||15||99|| |
|K-Ras|| || ||0.001|| || ||0.069|
| Positive||8||32|| ||3||5|| |
| Negative||12||261|| ||14||108|| |
|N-Ras and/or K-Ras|| || ||0.003|| || ||0.182|
| Positive||10||55|| ||5||18|| |
| Negative||10||238|| ||12||95|| |
Activated Ras mutations confer proliferative and survival signals. Mutations in the N-Ras gene are frequent genetic aberrations in adult AML.6–11 However, there have been only a few studies on childhood AML.12–14 With different mutation-detection techniques used and heterogeneous patient populations studied, the reported incidence of N-Ras mutations in patients with childhood AML at presentation vary considerably, from 6% to 33%. The previous studies that reported a higher incidence of N-Ras mutations mostly detected those mutations by using hybridization with oligonucleotide probes that were specific for each possible mutation, whereas the incidence of N-Ras mutations detected by direct sequencing was much lower. Because direct sequencing is the most unambiguous method for detecting Ras mutations, we used direct sequencing for each PCR product. The sensitivity of the method we used for analyzing Ras mutations was at a detection level of 2% of mutant DNA,34 which compares favorably with a sensitivity of 10% for oligonucleotide hybridization.6 Moreover, K-Ras mutations, which were studied less in previous reports, were examined completely in the current study, thus allowing more reliable estimates of the frequency of Ras mutations.
We found that 17.7% of patients who had de novo childhood AML had Ras mutations, with 12.3% N-Ras mutations and 6.2% K-Ras mutations. We failed to find a difference in the frequency of N-Ras mutations between patients who had AML with and without MLL rearrangements. The frequency of K-Ras mutations was greater in patients with MLL-positive AML than in patients with MLL-negative AML, but the difference did not reach statistical significance (P = 0.069). Only five patients with MLL-positive AML harbored Ras mutations, a number that clearly was too small to allow a meaningful statistical analysis of the clinical and prognostic impact of Ras mutations. The frequency of Ras mutations in patients with core-binding factor AML in the current study was 12.5%, which was lower than the incidence of 29.6% (8 of 27 patients) described in a very recent pediatric series35 but comparable to the frequency of 1 in 37 patients (which included both children and adults) with t(8;21) and Ras mutation.36 The discrepancy of those results probably was attributable to the different methodology used for the mutational analysis and the small patient numbers of core-binding factor AML in those studies.
N-Ras mutations were detected in 9.3% of patients with B-precursor ALL, and K-Ras mutations were detected in 12.8% of patients with B-precursor ALL. Taken together, the frequency of Ras mutations in the current series was 20.8% in children with B-precursor ALL, which comprised of 86% of patients with childhood ALL in Taiwan.26 Lübbert et al. reported that 6 of 100 children (6%) with ALL carried N-Ras mutations,15 but none had K-Ras mutations. Yokota et al. observed that 14 of 125 Japanese children (11%) with ALL had N-Ras mutations that were not correlated with outcome.16 Wiemels et al. reported that 20% of patients with childhood B-lineage ALL had Ras mutations, but data for the respective N-Ras or K-Ras mutations were not available.17 Recently, the Children's Oncology Group reported that 15.1% of children with ALL (10.5% with N-Ras mutations and 4.7% with K-Ras mutations) had Ras mutations and that the presence of Ras mutations was not associated with unique clinical presentation and did not predict outcome.18 We specifically addressed the association of Ras mutations with MLL gene rearrangement status. We observed a high frequency of K-Ras mutations (40%) in patients with B-precursor ALL who had MLL rearrangements, suggesting that K-Ras mutations play a role in the leukemogenesis of childhood MLL-positive ALL. Our results were contradictory to the findings of Mahgoub et al.,21 who used single-strand conformation polymorphism PCR and specific restriction enzyme method and did not find any Ras mutations among 15 patients with MLL-positive ALL, but they did detect 2 K-Ras mutations in 10 patients with AML. Because patients who have childhood ALL with MLL rearrangements have a much worse prognosis compared with patients who have childhood de novo AML, differences in the frequency of K-Ras mutations in patients with MLL-positive ALL and MLL-positive AML (8 of 20 patients vs. 3 of 17 patients, respectively) suggest that K-Ras mutations may play a role that contributes in the poor outcome of patients with MLL-positive childhood ALL. However, the numbers of patients in the current study were relatively small, and further study in a larger series of patients with MLL rearrangements will be required to draw firm conclusions.
The role of Class I mutations in the “two-hit theory,” including receptor tyrosine kinase and Ras pathway mutations, in driving the proliferation of leukemic cells in AML has been recognized more commonly,19 whereas the role of Ras mutations in driving the proliferation of lymphoblasts in ALL is less clear. Our observation of a strong association of Ras mutations in MLL-positive ALL also supports the “two-hit theory” of leukemogenesis. For patients with MLL rearrangements who do not have Ras mutations, other genetic alterations may play a role in leukemogenesis. FLT3 activation mutations, which also signal through the Ras pathway, often were mutated in patients with ALL who had MLL rearrangements.37, 38 The mutations of PTPN11, which is another gene of the Ras pathway, and the mutual exclusion of Ras mutations were found in patients with non-TEL/AML1 B-lineage ALL.39 Further investigation of cooperating mutations that involve genes related to the Ras pathway will provide a better understanding of the implication of collaborating genetic aberrations in patients with acute leukemias who have MLL gene rearrangements but do not harbor Ras mutations.