Tandem duplication of the FLT3 gene is found in acute lymphoblastic leukaemia as well as acute myeloid leukaemia but not in myelodysplastic syndrome or juvenile chronic myelogenous leukaemia in children
Department of Paediatrics, Faculty of Medicine, University of Tokyo,
Dr Yasuhide Hayashi Department of Paediatrics, Faculty of Medicine, University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-8655, Japan.
We examined mRNA expression and internal tandem duplication of the Fms-like tyrosine kinase 3 (FLT3) gene in haematological malignancies by reverse transcriptase-polymerase chain reaction (RT-PCR) and genomic PCR followed by sequencing. By RT-PCR, expression of FLT3 was detected in 45/74 (61%) leukaemia cell lines and the frequency of expression of FLT3 was significantly higher in undifferentiated type (B-precursor acute lymphoblastic leukaemia; ALL) than in differentiated type cell lines (B-ALL) (P = 0.0076). Using the genomic PCR method, 194 fresh samples including 87 acute myeloid leukaemias, 60 ALLs, 32 myelodysplastic syndromes (MDSs) and 15 juvenile chronic myelogenous leukaemias (JCMLs) were examined. Tandem duplication was found in 12 (13.8%) AMLs and two (3.3%) ALLs. Sequence analyses of the 14 samples with the duplication revealed that eight showed a simple tandem duplication and six a tandem duplication with insertion. Most of these tandem duplications occurred within exon 11, and two duplications occurred from exon 11 to intron 11 and exon 12. No tandem duplications of FLT3 gene were detected in MDS or JCML. The frequency of tandem duplication of FLT3 gene in childhood AML was lower than that in adult AML so far reported. All of the 12 AML patients with the duplication died within 47 months after diagnosis, whereas two ALL patients with the duplication have survived 44 and 72 months, respectively. These two ALL patients expressed both lymphoid and myeloid antigens and were considered to have biphenotypic leukaemia. These results suggest that tandem duplication is involved in ALL in addition to AML, but not in childhood MDS or JCML, and that childhood AML patients with the tandem duplication have a poor prognosis.
In recent years an internal tandem duplication within JM/TK-1 domains as a somatic mutation of the FLT3 gene has been reported in approximately 17–20% of adult AML (Nakao et al, 1996; Yokota et al, 1997), 3% of adult myelodysplastic syndrome (MDS), 15% of adult AML with myelodysplasia (Horiike et al, 1997; Yokota et al, 1997) and 20.3% of adult acute promyelocytic leukaemia (APL) (Kiyoi et al, 1997). The duplicated sequences were recognized in exon 11, intron 11 and exon 12, and their location and length was different in every case. Notably, the altered FLT3 gene was always transcribed in frame, and coded mutant FLT3 with a long JM domain (Nakao et al, 1996).
In this study we found a higher frequency of mRNA expression in B-precursor ALL cell lines among leukaemia cell lines. We also found tandem duplication of the FLT3 gene in ALL in addition to AML, but not in MDS or juvenile chronic myelogenous leukaemia (JCML). AML patients with the tandem duplication were considered to have a poor prognosis.
MATERIALS AND METHODS
Leukaemia cell lines
74 leukaemia cell lines were examined in this study. They included 25 B-precursor ALL cell lines (SCMC-L20, SCMC-L1, SCMC-L10, UTP-L51, UTP-L5, LC4-1, UTP-L20, NALL-1, KOPN-K, OM 9;22, UTP-2, NALM-26, NALM-17, NALM-20, NALM-24, BV-173, P30/OHK, A4/FUK, THP-8, REH, NALM-16, NALM-6, LAZ-221, KOPN-8, HPB-NULL) (Kawamura et al, 1995; Ohnishi et al, 1996), 11 B-ALL cell lines (BALM-9, BALM-13, BALM-14, BJAB, DAUDI, NAMALLA, BAJI, RAMOS, BALM-1, BALM-6, BAL-KH), 16 T-ALL cell lines (CCRF-HSB-2, ALL-SIL, MOLT-16, RPMI-8402, PEER, DND-41, MOLT-14, KOPT-KI, HPB-MLT, THP-6, L-SMK, L-SMY, KE-37, HPB-ALL, JURKAT, HAL-01) (Ohnishi et al, 1995), 17 AML cell lines (UTP-L3, P39/TSU, UTP-M2, UTP-M1, KG-1, CS-R, YNH-1, ML-1, SN-1, THP-1, P31/FUJ, SCC-3, J-111, MOBS-1, CTS, MOLM-13, CTV-1) and five CML cell lines (MOLM-1, MOLM-7, TS 9;22, SS 9;22, K 562). Some of these cell lines were established in our laboratory (Hayashi et al, 1993), while others were obtained from Fujisaki Cell Bank, Okayama, or from the Japanese Cell Resource Bank (JCRB). All these cell lines were cultured in RPMI-1640 medium containing 9% fetal bovine serum.
Fresh leukaemia specimens
194 fresh childhood leukaemia specimens including 87 AMLs, 60 ALLs, 32 MDSs and 15 JCMLs were examined in this study. Bone marrow (BM) and/or peripheral blood (PB) cells were obtained from these leukaemia patients at diagnosis and/or at relapse. ALL patients were mainly treated with Tokyo Children's Cancer Study Group (TCCSG) L13 or L14 protocol, and AML patients were mainly treated with the Japanese childhood AML protocol of the Ministry of Welfare. Specimens were immediately frozen in liquid nitrogen or stored at −80°C until processed at Saitama Children's Medical Centre, at the University of Tokyo Hospital and Hamamatsu University School of Medicine. Normal peripheral blood lymphocytes were obtained from 10 healthy volunteers and corresponding Epstein-Barr Virus (EBV)-induced B cell lines as control. Morphologic diagnoses of these patients were defined according to the French–American–British (FAB) classification (Bennett et al, 1981, 1985).
Total RNA was extracted from cell lines using the acid guanidine thiocyanate–phenol chloroform (AGPC) method (Chomczynski & Sacchi, 1987). Randomly primed cDNA was reverse-transcribed (Kong et al, 1997; Ida et al, 1997) from 4 μg of total RNA using a cDNA synthesis kit (Pharmacia Biotech) in a 33 μl mixture solution as described (Ida et al, 1997). 1 μl of the cDNA conversion mixture was amplified by PCR in a total volume of 50 μl with 10 mm Tris-HCl (pH 8.3), 50 mm KCl, 1.5 mm MgCl2, 0.001% (wt/vol) gelatin, 200 μm of each deoxyribonucleotide triphosphate (dNTP), 2.5 units Taq polymerase (Applied Biosystems) and 40 pmol of each primer. PCR amplification was performed in a DNA thermal cycler (Perkin-Elmer Cetus) under the conditions of preheating of the mixture at 94°C for 10 min, followed by 35 cycles of denaturation for 1 min at 94°C, annealing for 1 min at 59°C, and extension for 1 min 30 s at 72°C. A final extension proceeded for 7 min at 72°C. β-actin was amplified on the same cDNA as a control for the presence of amplifiable RNA (305 bp fragment). The β-actin gene was amplified at 94°C for 1 min, 55°C for 1 min and 72°C for 2 min for 30 cycles in a solution containing β-actin specific oligonucleotide primers. The sequences of the oligonucleotide primers for RT-PCR were R5 (sense), 5′-TGTCGAGCAGTACTCTAAACA-3′ and R6 (antisense), 5′-ATCCTAGTACCTTCCCAAACTC-3′. The sequences of the primers for β-actin were sense 5′-CTTCTACAATGAGCT GCGTG-3′ and antisense 5′-TCATGAGGTAGTCAGTCAGG-3′. 8 μl of the amplified PCR products were electrophoresed on a 3% agarose gel, stained with ethidium bromide and photographed with a polaroid camera under UV light.
Genomic PCR method
High molecular weight DNA was extracted from fresh samples using the proteinase-K phenol–chloroform method (Ohnishi et al, 1996). 50–100 ng of genomic DNA was amplified in a total volume of 50 μl of reaction mixture as described above except that 6% dimethyl sulphoxide was added to the reaction (Ohnishi et al, 1995). Genomic PCR was performed with primers of 11F and 11R combination (Nakao et al, 1996) which located at exon 11 and the amplification process consisted of 35 or 40 cycles of 30 s at 94°C for denaturation, 45 s at 49–51°C for annealing, 1 min at 72°C for extension as an amplification step and 7 min at 72°C for the final extension step. Because the size of intron 11 reached up to 90 bp, exons 11–12 were also amplified. For sequence analysis, genomic PCR was also performed using the primer pair 11F and 12R combination (Nakao et al, 1996) under the conditions described above whenever unexpected products were detected. The genomic structure of the FLT3 gene (Agnes et al, 1994) and primers for genomic PCR are described in Fig 11B. The sequences of the primers used are as follows: 11F (sense), 5′-CAATTTAGGTATGAAAGCC-3′; 11R (antisense), 5′-CAAACTCTAAATTTTCTCT-3′; 12R (antisense), 5′-GTACCTTTCAGCATTTTGAC-3′.
Cloning and sequencing of PCR products
The RT-PCR and genomic PCR products were cloned into TA cloning vector, then subjected to sequencing (Kong et al, 1997; Taki et al, 1997) with primers T7 and M13-Reverse using Dye Terminator Cycle Sequencing Kit (Applied Biosystems, Urayasu, Japan) on an Applied Biosystems DNA sequencer (model ABI 310).
Southern blotting analysis
Southern blotting was performed in two ALL patients with the tandem duplication according to the previous method (Ohnishi et al, 1995). Briefly, the 5.7 kb BamHI-HindIII fragment containing the coding sequence of J region of immunoglobulin heavy chain (JH) gene and 3.7 kb HindIII fragment containing T-cell receptor (TCR) Cβ1 gene were labelled with [α-32P]dCTP by the random priming method and used as the probe for Southern blot analysis. 10 μg DNA were digested with EcoRI, HindIII and BamHI, electrophoresed in TAE buffer, transferred to a nylon membrane filter, and fixed to the filter by UV light. The filters were hybridized with 32P-labelled probes overnight in a solution. After high-stringency washing, filters were exposed to X-ray film with an intensifying screen at −80°C for 2–4 d.
Expression and internal tandem duplication of FLT3 gene in leukaemia cell lines
We examined expression of FLT3 in 74 leukaemia cell lines by RT-PCR with primers of R5 and R6 which completely covered the TM and JM domain of FLT3 (Rosnet et al, 1993) (Fig 1A). Results are summarized in 1Table I, with representative results shown in Fig 22A. The expected FLT3 mRNA fragment of 366 bp was observed in all of the normal control samples (data not shown). Among the 74 leukaemia cell lines the FLT3 transcript was observed in 45 (61%) cell lines, which consisted of 21/25 (84%) B-precursor ALL, 4/11 (36%) B-ALL, 7/16 (44%) T-ALL, 10/17 (59%) AML and 3/5 (60%) CML, respectively. In respect to the incidence of FLT3 expression, there seemed to be no apparent difference among ALL (61%), AML (59%) and CML (60%) cell lines (Table I). However, in ALL cell lines the frequency of FLT3 expression tended to be significantly higher in undifferentiated type ALL (B-precursor ALL) than in differentiated type B-ALL cell line (P = 0.0076).
Table 1. Table I. Incidence of FLT3 expression in leukaemia cell lines determined by RT-PCR.
* Peripheral blood lymphocyte obtained from healthy adult volunteers.† EB-virus induced B-cell lines established from PB.
Only one acute monocytic leukaemia (AMOL) cell line (MOLM-13 or MOLM-14 derived from the same patient) of the 74 leukaemia cell lines showed a simple tandem duplication (21 bp) within exon 11 in addition to a normal transcript (Fig 3).
Internal tandem duplication of the FLT3 gene in fresh childhood leukaemia, MDS and JCML specimens
We investigated the tandem duplication of FLT3 gene in 194 fresh childhood leukaemia specimens by genomic PCR. The tandem duplication was found in 2/60 (3.3%) ALL and 12 (13.8%, one M1, five M2, three M4, three M5) of 87 AML samples. The frequency was relatively high in M4 (27.3%) subtypes (Table II, Fig 2B). Among the 14 patients with tandem duplication, sequence analysis revealed a simple duplication within exon 11 in seven patients (nos. 2–4, 6, 8, 10 and 14), a duplication with insertion within exon 11 in five patients (nos. 1, 5, 9, 11 and 13), and duplications derived from exon 11 to intron 11 and exon 12 in two patients (nos. 7 and 12). In most patients where the numbers of duplicated sequences were multiples of three, no insertion of nucleotides occurred. The insertion of these nucleotides made the total number of increased nucleotides multiples of three (Table III). We also deduced the amino acid sequences from the DNA sequences (Fig 3). Although the duplications varied in size and location, they always involved either Y591 or Y597 except for one patient, but the TK-1 domain was not significantly affected. All duplicated fragments contained one to four Tyr residues except for patient 6.
Table 2. Table II. Incidence of tandem duplication of FLT3 gene in childhood acute leukaemia, MDS and JCML determined by genomic PCR.
Table 3. Table III. Characteristics of tandem duplication of FLT3 gene.
We also examined the FLT3 genes in both the MDS and JCML by genomic PCR (Nakao et al, 1996), but did not find any tandem duplication.
Clinical characteristics of the leukaemia patients with tandem duplication of FLT3 gene
Of the 14 patients showing tandem duplication of FLT3 gene, one was diagnosed as AML-M1, five as M2, three as M4, three as M5 and two as ALL (Table IV). The 14 patients with the duplication consisted of 10 males and four females, aged from 3 to 14 (median 12) years. Among the 14 patients the duplication was found in 12 at diagnosis. In the remaining two patients (nos. 5 and 8) the duplication was found at relapse, not at diagnosis. All of the AML patients with the duplication died, whereas two ALL patients have been alive for 72 and 44 months. The survival duration of the 12 AML patients was from 0 to 47 months (median 13 months). One of the two ALL patients with the duplication was a 14-year-old boy whose leukaemic cells at diagnosis expressed CD 5, 7, 11b and 33, but not CD 10, 19 or HLA-DR. Although neither rearrangements of TCR Cβ gene nor JH gene was observed in his leukaemic cells, he was diagnosed with FAB-L2, not AML-M0, according to the AML-M0 criteria of FAB classification (presence of CD5 antigen) (Bennett et al, 1991). This patient had a normal karyotype and failed to achieve complete remission (CR) by ALL-oriented chemotherapy, and received peripheral blood stem cell transplantation (PBSCT) 9 months after diagnosis, but relapsed 12 months after PBSCT. Remarkably, myeloid antigens (CD11b and CD33) were not expressed at relapse. The patient received bone marrow transplantation (BMT) from his mother with the same HLA typing 6 months after relapse and has been in CR ever since. The other ALL patient was a 12-year-old boy whose leukaemic cells expressed CD 13, 19, 34 and HLA-DR, but not CD10. Rearrangements of the TCR Cβ and Ig JH genes were found in his leukaemic cells (data not shown). He has been in continuous CR for 44 months after diagnosis. Clinical characteristics of these 14 patients are listed in IV. Clinical data on 14Table IV.
Table IV. Clinical data on 14. acute leukaemia patients with tandem duplication of FLT3 gene.
* No duplication at diagnosis. † The blasts expressed CD 5, 7, 11b and 33, but not 10, 19 or 13. ‡ The blasts expressed CD 13, 19, 34 and HLA-DR, but not CD 10. TCR Cβ and JH genes were rearranged. § +, alive.
In the present study we investigated mRNA expression and internal tandem duplication of FLT3 by RT-PCR in 74 leukaemia cell lines. A relatively high percentage of the expression was shown in B-precursor ALL, CML and AML cell lines (84%, 60% and 59% respectively). This finding was comparable with the previous frequency of expression of FLT3 in myelogenous and B-precursor cell lines by Northern blot analysis (Meierhoff et al, 1995; Drexler, 1996). Expression of FLT3 was rarely found in T-cell lines (0/11) and B cell lines (3/11) by Northern blot analysis (Meierhoff et al, 1995). However, in our RT-PCR study the frequency of expression of FLT3 mRNA was higher both in T-ALL cell lines (44%) and B-ALL cell lines (36%) than previously reported. This may be due to the difference in sensitivity between RT-PCR and Northern blotting methods. In B-lineage ALL cell lines the frequency of FLT3 expression was significantly higher in B-precursor ALL cell lines than B-ALL cell lines, being compatible with the previous report that FLT3 gene was expressed more frequently in undifferentiated than differentiated cells (Meierhoff et al, 1995).
Among 74 leukaemia cell lines the tandem duplication was found only in one AML cell line (MOLM-13), derived from a 9-year-old boy with FAB-M5a phenotype. This is not in accordance with the p53 gene which is frequently altered in cell lines, but not in fresh leukaemia (Kawamura et al, 1995), suggesting that the FLT3 tandem duplication is not associated with in vitro growth.
We also investigated the tandem duplication of FLT3 gene by genomic PCR method in 194 fresh samples from children with various haematological malignancies, such as AML, ALL, MDS and JCML. Recently, the tandem duplication within JM/TK-1 domains was reported in 17–20% of patients with adult AML (Nakao et al, 1996; Yokota et al, 1997). However, in our present study the tandem duplication was found in 12 AML (13.8%) including three M4 (27.3%). Thus, the frequency of the duplication in childhood AML seemed to be lower than that in adult AML. Interestingly, tandem duplication of FLT3, which has not so far been reported in ALL, was also found in two ALL (3.3%). Remarkably, both these ALL patients had myeloid antigens, suggesting the duplication of FLT3 gene occurs mainly in myeloid lineage cells and partly in lymphoid lineage cells with the potential of myeloid differentiation.
All of the abnormal products were in-frame and the deduced amino acid sequence of TK-1 domain remained unchanged. Although no direct evidence is obtained so far, these amino acid sequences could play a crucial role in altering the function of FLT3 kinase and thus reinforcing proliferation of the leukaemia cells (Yokota et al, 1997).
In this study, tandem duplication of the FLT3 gene was detected at relapse, not at diagnosis, in two AML patients, suggesting that the duplication is associated with late-stage leukaemia or progression (Horiike et al, 1997).
No tandem duplications of FLT3 gene were detected in childhood MDS or JCML in this study. Horiike et al (1997) have reported that the duplications were found in 2/58 (3%) adult MDS patients and these two transformed to overt leukaemia within a few months. The reason for the difference between Horiike's study and ours is unknown. One explanation is that adult MDS may be aetiologically different from childhood MDS, that is adult MDS often has complex chromosomal abnormalities whereas childhood MDS usually has only a simple abnormality (7 monosomy, etc.). Further studies are necessary to clarify this difference.
All the 12 AML patients with FLT3 tandem duplication died within 47 months after diagnosis. Most of these patients were resistant to initial chemotherapy, and could not be induced to enter CR. These results suggest that childhood AML patients with the duplication have a poor prognosis, and that this clinical outcome is compatible with that of adult AML (Horiike et al, 1997; Kiyoi et al, 1997). A large prospective study is necessary to confirm this.
We thank Dr Y. Matsuo, Okayama Hayashibara Biochemical Laboratories Inc., Fujisaki Cell Centre, for the generous gift of cell lines. We express appreciation to S. Sohma for her dedicated technical assistance. Special thanks are also addressed to clinicians at various hospitals in Japan for providing samples and clinical data.
This work was supported by a Grant-in-Aid for Cancer Research from the Ministry of Health and Welfare of Japan, a Grant-in-Aid for Scientific Research on Priority Areas and Grant-in-Aid for Scientific Research (B) and (C) from the Ministry of Education, Science, Sports and Culture of Japan.