Division of Hematology, Department of Internal Medicine, Nippon Medical School, Tokyo, Japan

Correspondence to: K Inokuchi, Division of Hematology, Department of Internal Medicine, Nippon Medical School, 1-1-5 Sendagi, Bunkyo-ku, Tokyo 113, Japan; Fax: 81-3-5814-6934

Chronic myelogenous leukemia (CML) is characterized by the Philadelphia (Ph) chromosome and bcr/abl gene rearrangement which occurs in pluripotent hematopoietic progenitor cells expressing the c-kit receptor tyrosine kinase (KIT). To elucidate the biological properties of KIT in CML leukemogenesis, we performed analysis of alterations of the c-kit gene and functional analysis of altered KIT proteins. Gene alterations in the c-kit juxtamembrane domain of 80 CML cases were analyzed by reverse transcriptase and polymerase chain reaction-single strand conformation polymorphism (RT-PCR-SSCP). One case had an abnormality at codon 564 (AATAAG, AsnLys), and six cases had the same base abnormality at codon 541 (ATGCTG, MetLeu) in the juxtamembrane domain. Because the change from Met to Leu at codon 541 was a conservative one which was also observed in the normal population and normal tissues of CML patients, it probably represents a polymorphic variation. Although samples of hair roots and leukemic cells from the chronic phase of one CML patient showed no abnormality, an abnormality at codon 541 (ATGCTG, MetLeu) was found only at blastic crisis (BC) of this case. In the case with the abnormality at codon 564, the mutation was detected only in a sample of leukemic cells collected at BC. To examine the biological consequence and biological significance of these abnormalities, murine KIT^L540 and KIT^K563 expression vectors were introduced into interleukin-3 (IL-3)-dependent murine Ba/F3 cells to study their state of tyrosine phosphorylation and their growth rate. Ba/F3 cells expressing KIT^WT, KIT^L540 and KIT^K563 showed dose-dependent tyrosine phosphorylation after treatment with increasing concentrations of recombinant mouse stem cell factor (rmSCF). The cells expressing KIT^L540 and KIT^K563 were found to have greater tyrosine phosphorylation than cells expressing KIT^WT at 0.1 and 1.0 ng/ml of rmSCF. The Ba/F3 cells expressing KIT^K563 proliferated in response to 0.1 and 1.0 ng/ml of rmSCF as well as IL-3. The Ba/F3 cells expressing KIT^L540showed a relatively higher proliferative response to 0.1 ng/ml of rmSCF than the response of cells expressing KIT^WT. These mutations and in vitro functional analyses raise the possibility that the KIT abnormalities influence the white blood cell counts (P < 0.05) and survival (P < 0.04) of CML patients.

Leukemia (2002) 16, 170-177. DOI: 10.1038/sj/leu/2402341

c-kit; CML; Ba/F3; WBC

Introduction

Philadelphia (Ph) chromosome is the cytogenetic hallmark of chronic myelogenous leukemia (CML) and is found in up to 95% of CML patients.¹ The demonstration of bcr/abl mRNA is accepted as a reliable diagnostic marker for CML, and in some cases this evidence is more reliable than the Ph chromosome.² The clinical signs and hematological findings probably depend partly on the presence of P210^BCR/ABL, which plays a central role in the pathogenesis of the chronic phase of CML.³ According to the breakpoint site of the bcr gene in most CML patients, there are two bcr/abl mRNAs for P210^BCR/ABL, one with and one without exon b3 (b3-a2 type and b2-a2 type).⁴ In a smaller number of CML patients, there are two other types of bcr/abl mRNAs based on the breakpoint positions of the bcr gene, ie m-bcr and -bcr for P190^BCR/ABL and P230^BCR/ABL, respectively.^5,6 Extensive studies have been performed on the subtypes of the bcr/abl gene and their relation to the prognosis and clinical features.^7,8 The established findings regarding the influence of the subtype of the bcr/abl gene (b3-a2 or b2-a2) on the clinical characteristics have been similar for each subtype, except for a higher platelet count in patients with the b3-a2 type.^4,9 P190^BCR/ABL in CML may be associated with monocytosis, and P230^BCR/ABL may be associated with the chronic neutrophilic leukemia variant and marked thrombocytosis.^6,10 Other molecular factors which might control the clinical features and hematological characteristics remain unclear.

Recently, the stem cell factor (SCF) c-kit signal system (KIT/SCF) has been shown to play a crucial role in hematopoiesis.¹¹ The c-kit receptor tyrosine kinase (KIT) is expressed on progenitor stem cells as well as mast cells. SCF synergizes in vitro with other cytokines to increase the number and size of colonies of hematopoietic progenitors.¹² Thus, we speculated that the KIT/SCF system possibly controls the hematological characteristics of CML.

The present study was designed to investigate the mutations of the c-kit gene and the relationship between the in vitro function of mutant c-kit and the biological features of CML patients.

Materials and methods

Patients

We studied 116 bone marrow (BM) or peripheral blood (PB) samples obtained from 80 patients with CML in various clinical phases: 65 in chronic phase (CP), seven in accelerated phase (AP), and 44 in blastic crisis (BC). The diagnosis of CML was made on the basis of clinical features, hematological data and Ph chromosome. In 34 patients, both cytogenetic and molecular analyses were performed in more than two phases, ie CP and BC (30 patients), CP and AP (two patients), and CP, AP and BC (two patients). Sixty-eight normal BM or PB samples were obtained to study mutation and polymorphism of the c-kit gene. These samples were obtained with the patients' and normal volunteers' informed consent.

Cells

Ba/F3,¹³ a murine IL-3-dependent pro-B lymphoid cell line, was cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS) and 10% WEHI-3 cell conditioned medium.

Extraction of RNA and DNA

The total RNA of BM or PB leukocytes was extracted with an RNAzol kit (Biotex Laboratories, Houston, TX, USA), which was based on a technique described previously.¹⁴ The total cellular DNA was extracted from BM and PB cells by protease K digestion, phenol-chloroform extraction and ethanol precipitation.

RT-PCR of c-kit mRNA

RT-PCR was performed as described elsewhere.¹⁵ The sense primers were: kit560-1, 5'-CTGTTCACTCCTTTGCTGAT-3' (residues 1582-1601); kit560-2, 5'-TTCGTAATCGT AGCTGGCAT-3' (residues 1605-1624); kit816-1, 5'-ATCATGGAGGATGACGAGTTG-3' (residues 2287-2306); and kit816-2, 5'-CTAGACTTAGAAGACTTGCT-3' (residues 2310-2329). The antisense primers were: kit560-3, 5'-CATGTGATTACCAAGGTAA-3' (residues 1955-1974); kit560-4, 5'-GCTCCAAGTAGATTCACAAT-3' (residues 1978-1997); kit816-3, 5'-ATTTCAGCAGGTGCGTGTTC-3' (residues 2698-2717); and kit816-4, 5'-TTTTTA GGGGATCTGCATCC-3' (residues 2742-2761). Complementary DNA was synthesized from 500 ng of total cellular RNA extracted from cells using 100 ng of primer kit560-4 for analysis of the sequence in the juxtamembrane domain or kit816-4 for analysis of the phosphotransferase domain. Briefly, RT reaction mixture contained 32 U of avian myeloblastosis virus (AMV) RT (Takara Biochemicals, Shiga, Japan) in 25 l of a solution containing 200 mol/l each of all four dNTPs, 80 U of RNase inhibitor, 50 mmol/l Tris-HCl (pH 8.3), 75 mmol/l KCl, 10 mmol/l dithiothreitol and 3 mmol/l MgCl₂. The reaction was allowed to proceed for 60 min at 37°C, and the reaction mixture was used as the PCR substrate. A 35-cycle PCR reaction was performed in a DNA Thermal Cycler with slight modification of our original protocol.⁴ Briefly, 25 l of the RT reaction solution was mixed with a mixture containing 250 mol/l of each of all four dNTPs, 100 ng of 5'-primer ST1, 10 mmol/l Tris-HCl (pH 8.3) , 50 mmol/l KCl, 3 U of Taq DNA polymerase (Takara Biochemicals) and 100 ng of primers kit560-1 and kit560-4 or primers kit816-1 and kit816-4. The reaction conditions and cycle number were the same as those described above. A second PCR was performed for direct sequencing. Primers kit560-2 and kit560-3 or primers kit816-2 and kit816-3 were used to amplify the first PCR products, whereas primers SC3 and ASC3 were used for the 3'-fragment.

Single-stranded conformation polymorphism (SSCP) gel analysis

Detection of c-kit mutations was performed by SSCP gel analysis. The 5'-ends of primers (100 pmol) were labeled with -³²P-ATP (3000 Ci/mmol) and polynucleotide (5 U, Boehringer-Mannheim; Mannheim, Germany) in 10 l of 50 mM Tris-HCl, pH 8.3, 10 mM MgCl₂ and 5 mM DTT at 37°C for 30 min. The PCR mixture contained 10 pmol of each of the labeled primers, 2 nmol each of the four deoxynucleotides, 0.1 g of sample cDNA or DNA, and 0.25 U of Taq polymerase in 10 l of the buffer provided in the GeneAmp kit. The PCR products were mixed with 10 volumes of a loading buffer containing 95% formamide, 20 mM EDTA, 0.05% bromphenol blue and 0.05% xylene cyanol, denatured at 94°C for 5 min, quenched on ice and applied (1 l/lane) to a 10% polyacrylamide gel containing 90 mM Tris-borate, pH 8.3, 4 mM EDTA, and 10% glycerol. Electrophoresis was performed at 40 W for 3 h at 18° with cooling using a water jacket. The gel was dried on a filter paper and exposed to X-ray film at -80° for 1-24 h with an intensifying screen.

Sequence analysis of the c-kit gene

After the RT-PCR products were separated on 2% agarose gels and stained with ethidium bromide, the amplified fragment was excised from the gel, electroeluted, purified with phenol and precipitated with ethanol. The fragments were subcloned into the EcoRV site of the pGEM-5Zf(+/-) cloning vector.¹⁶ The transfected cells were plated on to Luria-Beriani (LB)-ampicillin agar plates containing 5-bromo-4-chloro-3-indolyl--D-galactoside (X-Gal), isotransferred to fresh LB-ampicillin agar plates containing X-Gal and isopropylthio--D-galactoside, and cultured overnight for secondary selection. White colonies were transferred into 150 ml of LB medium containing 50 mg/ml ampicillin and cultured at 37°C for 4 h. The cultures were sedimented by centrifugation, resuspended in 20 ml of water and heated at 98°C for 10 min. After centrifugation, the supernatants were amplified by PCR using the T7 or SP6 primer. Three to five clones of the three independent PCR products were sequenced using a Model 377 ABI sequencer with dye terminators (Perkin Elmer, Warrington, UK). All sequences were confirmed in both orientations. A mutation was defined as when three or more clones showed the same abnormality of the base sequence.

Site-directed mutagenesis and transfection

To examine for functional abnormality of c-kit mutations in activation of c-kit tyrosine kinase activity, site-directed mutagenesis was performed using murine c-kit cDNA as described by Furitsu et al.¹⁷ The genes encoding murine wild-type c-kit and c-kit c-DNA with a mutation at codon 814 (GACGTC) in the XbaI site of the expression vector, pEF-BOS, were used as wild and mutant controls.¹⁸ These two c-kit expression vectors were kindly provided by Prof Y Kanakura (Osaka University, Osaka, Japan). To generate genes containing L-540, two oligonucleotide sets of 5'-GCATTATTGTGCTGATTCTGACCTACAAAT-3' and 3'-CAGAA TCAGCACAATAATGCACATCATGCC-5' were synthesized, annealed to the template and extended. To generate genes containing K-563, two oligonucleotide sets of 5'-GAGGAGA TAAGTGGAAACAATTATGTTTAC-3' and 3'-ATTGTTTCC ACTTATCTCCTCAACAACCTTCC A-5' were synthesized, annealed to the template and extended, according to the instructions accompanying the ExSite PCR-Based Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA, USA). Fifteen temperature-cycles of PCR were performed to extend and incorporate mutation primers by Pfu Turbo DNA polymerase, resulting in formation of mutated circular strands. Methylated parental DNA template was digested with 10 U of DpI. The mutated double-stranded plasmid was transformed into XL1-Blue E. coli cells to expand the plasmid after repairing the nicks with T4 DNA ligase. The c-kit cDNA and expression vectors were sequenced to confirm the mutations and absence of artifact abnormalities.

The expression vectors containing wild-type and mutated c-kit cDNA (10 g) were transfected by electroporation into 5 ´ 10⁷ Ba/F3 cells with pSV2Neo (0.2 g), which encodes and expresses the neomycin-resistant protein. The Ba/F3 cells were cultured in RPMI 1640 containing 10% FCS 600 g/ml G418 (Calbiochem-Novabiochem, San Diego, CA, USA) and 10% WEHI-3 cell-conditioned medium. After selection with G418 and limiting dilution of the cells, expression of KIT was confirmed by flow cytometry.

Flow cytometry

To detect cell surface expression of KIT, cells were incubated with fluorescein isothiocyanate (FITC)-conjugated anti-KIT (2B8 clone; PharMingen, San Diego, CA, USA) MoAb at 4°C for 30 min and analyzed on a FACScan (Becton Dickinson, Los Angeles, CA, USA).¹⁵

Immune complex kinase assay

The procedures of immunoprecipitation, gel electrophoresis and immunoblotting were performed according to the methods described previously.¹⁹ The cells were washed with PBS and lysed in 500 l of RIPA buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA, 1% NP-40, 0.5% deoxycholate, and 0.1% SDS with protease inhibitors (1 mM PMSF, 50 g/ml of antipain, 5 g/ml of aprotinin, and 2 g/ml of leupeptin)). Cell debris was removed by centrifugation. The supernatant was precleared by incubation with protein-G Sepharose (Pharmacia, Uppsala, Sweden) on ice for 30 min. The precleared lysates were incubated with anti-mouse c-kit MoAb, 2B8 (PharMingen) for 1 h on ice and with Protein-G Sepharose for 30 min on ice to collect antigen-antibody complexes. The immunoprecipitates were washed three times with NP-40 buffer (50 mM Tris, pH 8.0, 1.0% NP-40, 50 mM NaCl) containing protease inhibitors. The proteins released from the immunoprecipitates by Laemli's sample buffer were subsequently analyzed by electrophoresis on 5-25% SDS-PAGE. The proteins were electrophoretically transferred from the gel on to a nitrocellulose membrane. After the transfer, the filter was blocked by incubation in 1% BSA/Tween-PBS (1.37 M NaCl, 27 mM KCl, 81 mM Na₂HPO₄, 15 mM KH₂PO₄, 1% Tween-20, 1% BSA). Immunoblotting was performed with an antiphosphotyrosine MoAb (Py20; PharMingen) or anti-c-kit antibody, and detected with HRP-conjugated goat anti-mouse IgG (Santa Cruz Biotechnology, Santa Cruz, CA, USA). All immunoprecipitates were detected with HPR-conjugated goat anti-mouse IgG and visualized by autoradiography.

Semiquantification of the photographic signals was performed using an MCID image analysis system (Imaging Research, St Catherines, Ontario, Canada). The entire width of the lane was analyzed with appropriate background subtraction. All bands in one photograph were analyzed together. The relative intensity was defined as the ratio of the KIT phosphorylation signals to KIT^V814 signals at 15 min incubation.

Cell proliferation assay

Proliferation of cells was quantified by ³H-thymidine incorporation. Briefly, the cells were washed twice with IMDM medium, and triplicate aliquots of cells (5 ´ 10⁴) suspended in 200 l of Cosmedium-001 (Cosmo Bio Co., Tokyo, Japan) were cultured in 96-well microtiter plates for 72 h at 37°C in the presence of various concentrations of rmIL-3 or rmSCF, provided by Kirin Brewery Company, Ltd. At 72 h after initiation of the culture, 0.5 Ci ³H-thymidine (specific activity, 5 Ci/mmol; Amersham, Arlington Heights, IL, USA) was added to each well. Five hours later, the cells were harvested with a semi-automatic cell harvester (Pharmacia), and the incorporation of ³H-thymidine was measured with a liquid scintillation counter.

Determination of the type of bcr/abl mRNA

Reverse transcriptase-polymerase chain reaction analysis (RT-PCR) for determination of the type of bcr/abl mRNA was carried out as described previously.^4,16 Briefly, complementary DNA (cDNA) was prepared from 250 ng of total RNA using an antisense ABL cDNA primer. Primers bcr-1 and abl-1 were used for PCR for p210 mRNA, whereas primers e-1 and abl-1 were used for p190 mRNA. The PCR products were electrophoresed through a 2.5% agarose gel, stained with ethidium bromide, Southern-transferred and hybridized.

Statistical analysis

Statistical analysis was performed using the Statview (Brain Power, Calabashes, CA, USA) software package for the Macintosh personal computer. Comparisons of groups were analyzed using Fisher's exact test for 2 ´ 2 tables. Values of P < 0.05 were considered significant.

Results

c-kit mutations in CML patients

Eighty CML cases were screened for mutations in the coding region of the c-kit gene by RT-PCR-SSCP. Expression of the c-kit gene was detected in all 80 CML patients by RT-PCR. Seven cases showed aberrant bands on RT-PCR-SSCP gels. Sequencing of the RT-PCR fragments which showed aberrant bands on the RT-PCR-SSCP gels revealed one of the seven cases had an abnormality at codon 564 (AATAAG, AsnLys) in the juxtamembrane domain, while six cases had an abnormality at codon 541 (ATGCTG, MetLeu) (Figure 1). These two base abnormalities were observed in cDNA clones generated from three independent PCR products. Sixty-eight normal BM aspirate samples were obtained from hematologically healthy volunteers after obtaining informed consent and used as normal controls. One of the 68 healthy normal BM samples had the same migration pattern by SSCP and the same alteration as that of our CML patients at codon 541 in which ATG was changed to CTG, resulting in a change in the encoded amino acid from Met to Leu (Table 1). The same codon 541 abnormality was detected in RNA samples from skin and hair roots of cases 3, 4, 5 and 6 (Table 2). Because the amino acid substitution was a conservative one which was also observed in the normal population and normal tissues of the CML patients, it probably represents a polymorphic variation. Intriguingly, however, RNA samples from the hair roots and leukemic cells at CP of case 2 were demonstrated to have no abnormality, whereas this case showed the same abnormality at codon 541 (ATGCTG, MetLeu) in the leukemic cells at BC. Thus, we think the alteration of codon 541, in which ATG was changed to CTG, occurred during leukemic progression of case 2. In case 1, the abnormality at codon 564 was detected only in an RNA sample from leukemic cells at BC (Figure 1).

c-kit receptor tyrosine kinase

In order to determine the functional role of c-kit abnormalities in ligand-independent activation of c-kit products, expression vectors containing normal or mutated murine c-kit genes were transfected into a murine IL-3-dependent cell line, Ba/F3 cells.

We used site-directed mutagenesis to construct two mutant murine c-kitR expression vectors: c-kit^K563 coding for substitution of Lys for Asn at codon 563, and c-kit^L540 coding for substitution of Leu for Met at codon 540, which correspond to Lys-564 and Leu-541 of abnormal human c-kitR, respectively. These murine c-kit-expression vectors were co-transfected into the Ba/F3 murine IL-3-dependent cell line with pSV₂Neo, which contains the neomycin resistance gene, by electroporation. As a negative control, the pEF-BOS vector, without the c-kit gene, was transfected into Ba/F3 cells. After selection in a G418-containing medium for 2 weeks, surface expression of KIT on the transfected cells was examined with DX2, an anti-mouse c-kit MoAb. Flow cytometric analysis showed that, although transfection of Ba/F3 cells with the pEF-BOS vector alone resulted in no expression of KIT, Ba/F3 cells transfected with pEF-BOS-KIT^WT, pEF-BOS-KIT^V814, pEF-BOS-KIT^K563 or pEF-BOS-KIT^L540 showed abundant surface expression of KIT^WT, KIT^V814, KIT^K563 or KIT^L540 on their surface, respectively (Figure 2a).

State of tyrosine phosphorylation of KIT^K563 or KIT^L540

To examine the state of KIT-tyrosyl phosphorylation in the transfected Ba/F3 cells, the cells were deprived of serum and growth factors for 12 h and then stimulated with 0.1, 1 or 100 ng/ml of rmSCF for 15 min. KIT was then immunoprecipitated and assayed by immunoblotting with either antiphosphotyrosine MoAb or anti-c-kit MoAb. As shown in Figure 2b, when wild-type or mutated c-kit genes were transfected into Ba/F3, the c-kit gene products were synthesized in the cells as 145- and 125-kDa proteins, respectively. Immunoblotting with an antiphosphotyrosine MoAb showed that increased phosphotyrosine was observed in KIT^V814 regardless of rmSCF stimulation. In contrast, KIT^WT, KIT^L540 and KIT^K563 were dose-dependently phosphorylated on tyrosine after treatment with increasing concentrations of rmSCF. KIT^L540 and KIT^K563 were found to be more phosphorylated on tyrosine than KIT^WT at 0.1 and 1.0 ng/ml of rmSCF (Figure 2).

Repeated semi-quantification of immunoblots (Figure 3a) and the time course for the induction using a narrower range of concentrations between 0 and 1.0 ng/ml of rmSCF (Figure 3b) successfully showed dose-dependent activation of KIT^L540 and KIT^K563 as well as KIT^WT, although KIT^V540 and KIT^K563 showed relatively higher efficiency than KIT^WT between 0.1 and 1.0 ng/ml of rmSCF (Figure 3).

KIT^L540 and KIT^K563 modulation in IL-3-dependent Ba/F3 growth

To determine if KIT^L540 and KIT^K563 could modulate IL-3- or SCF-dependent growth, Ba/F3 cells expressing KIT^WT, KIT^W814, KIT^L540 or KIT^K563 were cultured in the presence of 0 to 100 ng/ml of rmIL-3 or 0 to 1000 ng/ml of rmSCF for 72 h, followed by measurement of cell proliferation using ³H-thymidine uptake assay (Figure 4). It was demonstrated that rmIL-3 induced dose-dependent proliferation whereas rmSCF induced no proliferation in the parental Ba/F3 or pEF-BOS-transfected Ba/F3 cells. In addition to the proliferative response to rmIL-3, Ba/F3 cells expressing KIT^WT, KIT^K563 and KIT^L540 induced dose-dependent proliferation in response to rmSCF over the range of 0.1 to 1000 ng/ml, indicating functional expression of KIT^WT, KIT^K563 and KIT^L540. KIT^K563 induced a higher proliferative response to 0, 0.1 and 1 ng/ml of rmSCF than KIT^WT. The proliferative response to 0.1 ng/ml of rmSCF by Ba/F3 cells expressing KIT^L540 was relatively higher than the response by cells expressing KIT^WT. In contrast, Ba/F3 cells expressing KIT^V814 proliferated even in the absence of both rmIL-3 and rmSCF.

Clinical features of CML patients with c-kit abnormality

Table 2 shows the relationships between the clinical data and the molecular characteristics of the seven CML cases with a c-kit abnormality. Three kinds of bcr/abl junctions, ie the b2-a2 and b3-a2 types of major bcr breakpoint, and the minor bcr/abl type of minor bcr breakpoint, are shown in Table 2. One of the seven CML cases with c-kit gene alterations had the b2-a2 type, while four had the b3-a2 type. Intriguingly, two CML patients with a c-kit abnormality had the rare, minor bcr/abl type. Case 6 had extramedullary onset, and case 7 had received 5-FU medication for several years for esophageal cancer. As shown in Table 2, the platelet counts of all seven cases at diagnosis of CP were within or below the normal range, regardless of the type of bcr/abl junction. The platelet counts of CML with 564^Lys-, 541^Leu-KIT were marginally lower (P < 0.07, Table 3) than the counts of CML with wild-type KIT (^WT-KIT). The WBC count at diagnosis of six mutated cases was markedly higher than the other 73 cases of CML with KIT^WT (P < 0.05, Table 3). The WBC count was relatively low in case 2 with a normal c-kit gene in CP. The mean survival duration of the seven CML patients with 564^Lys- or 541^Leu-KIT was 45.9 months, while that of the CML patients without any KIT abnormality was 69.6 months (Table 3). 564^Lys-KIT and 541^Leu-KIT may be prognostic factors (P < 0.04), as shown in Table 3.

Discussion

The interstitial cells of Cajal (ICCs) and hematopoietic stem cells express both KIT and CD34. Gastrointestinal stromal tumors (GISTs) may originate from the ICCs.²⁰ CML may originate from hematopoietic stem cells that are double-positive for KIT and CD34. High expression of c-kit has been demonstrated on CD34⁺ cells from chronic-phase CML patients.²⁰ Here, we successfully demonstrated that the juxtamembrane domain mutant occurs in some CML cells which are double-positive for KIT and CD34 as well as in ICCs. The juxtamembrane domain mutant is found in GISTs.²¹ The juxtamembrane domain mutant is also found in mast cell leukemia.²² The tyrosine kinase domain mutant occurs only in human and murine mast cell leukemia, but the mechanism of tumorigenesis differs between the juxtamembrane domain mutant and the tyrosine kinase domain mutant. The former, the juxtamembrane domain mutant, is constitutively dimerized and activated without binding SCF, whereas the latter, the tyrosine kinase domain mutant, is constitutively activated without forming dimers.²³ In GISTs, mutations are located within an 11-amino acid stretch (Lys-550 to Val-560).²¹ The tyrosine kinase and proliferation of Ba/F3 cells expressing these mutated KIT proteins were constitutively activated without SCF.²¹ Another report shows two major types of deletion mutations involving codons 550-565 and codons 566-580, stretches which were especially common in GISTs.²² In murine mastocytoma cells, the mutation is deletion of a 7-amino acid stretch (Thr-573 to His-579).²⁴ Point mutations/polymorphism at codon 564 or 541 is within or near these stretches involved in GIST mutations and deletion mutations of murine mastocytoma, although these mutations resulted in these mutant KITs not being constitutively activated. These in vitro biological differences may be partly due to differences in the mutation type, such as deletions of many nucleotides or one nucleotide mutation.

The 541^Leu-KIT mutation may be one of polymorphism, since it is found even in normal healthy people. In our CML patients, however, the frequency of the 541^Leu-KIT polymorphism was relatively more common compared with in normal individuals (Table 1). This higher frequency in CML was partly due to a newly occurring mutation at BC, such as in case 2.

The Ba/F3-cell proliferation-inducing activity of KIT^L540 was the same as that of KIT^WT. Tyrosine kinase activation of KIT^L540 was slightly higher than that of KIT^WT in medium containing 0.1 ng/ml of SCF. The tyrosine kinase activation and Ba/F3-cell proliferation activities of KIT^K563 were relatively higher than those of wild-type KIT in medium containing between 0.1 and 1.0 ng/ml of SCF. Based on these in vitro data on tyrosine kinase activation and proliferation activities, KIT^L540 and KIT^K563 do not have exactly the same function as KIT^WT. Unlike KIT^WT, KIT^L540 and KIT^K563 may cause very low levels of spontaneous tyrosine kinase activation and cell proliferation even without SCF, as shown in Figures 3 and 4. These findings may indicate that 541^Leu-KIT does not fall within the category of polymorphism. This is because the KIT abnormality of these cases was more common (8.8%) than in the normal healthy cases (1.5%; P < 0.05).

We speculate that these tiny differences in in vitro function may influence the clinical phenotype of CML. Recently, an internal tandem duplication (ITD) of the juxtamembrane (JM) domain-coding sequence of the FLT3 gene in acute myeloid leukemia patients was found to lead to leukocytosis and shorter survival.²⁵ The KIT and FLT3 proteins are members of the class III receptor tyrosine kinase family.²⁶ The ITD mutant of FLT3 is constitutively dimerized and autophosphorylated on tyrosine residues.²⁷ The KIT^L540 and KIT^K563 abnormalities of JM may be more easily dimerized and autophosphorylated than KIT^WT between 0 and 1.0 ng/ml SCF. This phenomenon may result in a tiny change in the structure of the KIT molecule which causes CML cells to undergo greater proliferation and culminates in shorter survival of CML patients.

Overall, the present analysis found that these c-kit mutations do not greatly affect the pathogenesis of CML, although the presence of these mutations resulted in clinical heterogeneity of the blood and a poorer prognosis for CML patients bearing a mutation vs CML patients with normal c-kit. The poor prognosis is probably due to the higher proliferation-inducing activities of 564^Lys-, 541^Leu-KIT than ^WT-KIT between 0 and 1.0 ng/ml SCF. The serum SCF concentration in our CML patients was between 0.5 and 1.5 ng/ml (preliminary data). Thus, CML blasts exposed to SCF concentrations between 0.5 and 1.5 ng/ml possibly undergo greater phosphorylation and proliferation.

Although our findings do not fully explain the mechanisms by which 564^Lys-, 541^Leu-KIT lead to clinical heterogeneity of CML, a small functional difference from KIT^WT is thought to be one cause of differences in the prognosis and the phenotype of leukocytosis in CML.

Acknowledgements

We thank Dr Y Kanakura for providing full-length murine c-kit expression vectors and Kirin Brewery Company, Ltd for providing rmIL-3 and rmSCF. This work was supported by Grants-in-Aid from the Ministry of Education, Science and Culture of Japan and the Takahashi Foundation.

1 Groffen J, Stephenson JR, Heisterkamp N, de Klein A, Bartram CR, Grosveld G. Philadelphia chromosomal breakpoints are clustered within a limited region, bcr, on chromosome 22. Cell 1984; 36: 93-99. MEDLINE

2 Kantarjian HM, Smith TL, McCredie KB, Keating MJ, Walters RS, Talpaz M, Hester JP, Bligham G, Gehan E, Freireich EJ. Chronic myelogenous leukemia: a multivariate analysis of the associations of patient characteristics and therapy with survival. Blood 1985; 66: 1326-1335. MEDLINE

3 Shtivelman E, Lifshiftz B, Gale RP, Canaani E. Fused transcripts of abl and bcr genes in chronic myelogenous leukemia. Nature 1985; 315: 550-554. MEDLINE

4 Inokuchi K, Inoue T, Tojo A, Futaki M, Miyake K, Yamada T, Tanabe Y, Ohki I, Dan K, Ozawa K, Asano S, Nomura T. A possible correlation between the type of bcr-abl hybrid messenger RNA and platelet count in Philadelphia-positive chronic myelogenous leukemia. Blood 1991; 78: 3125-3127. MEDLINE

5 Yamaguchi H, Inokuchi K, Shinohara T, Dan K. Extramedullary presentation of chronic myelogenous leukemia with p190 BCR/ABL transcripts. Cancer Genet Cytogenet 1998; 102: 74-77. MEDLINE

6 Yamagata T, Mitani K, Kanda Y, Yazaki Y, Hirai H. Elevated platelet count features the variant type of BCR/ABL junction in chronic myelogenous leukemia. Br J Haematol 1996; 94: 370-372. MEDLINE

7 Mills KI, Birnie GD. Does the breakpoint within the major breakpoint cluster region (M-bcr) influence the duration of the chronic phase in chronic myeloid leukemia? An analytical comparison of current literature. Blood 1991; 78: 1155-1161. MEDLINE

8 Shtalrid M, Talpaz M, Kurzrock R, Kantarjian H, Trujillo J, Gutterman J, Galina Y, Mark B. Analysis of breakpoints within the bcr gene and their correlation with clinical course of Philadelphia-positive chronic myelogenous leukemia. Blood 1988; 72: 485-490. MEDLINE

9 Faderl S, Talpaz M, Estrov Z, O'Brien S, Kurzrock R, Kantarjian HM. The biology of chronic myeloid leukemia. N Engl J Med 1999; 341: 164-172. MEDLINE

10 Melo JV, Myint H, Galton DAG, Goldman JM. P190 BCR-ABL chronic myeloid leukemia: the missing link with chronic myelomonocytic leukemia? Leukemia 1994; 8: 208-211. MEDLINE

11 Ogawa M, Matsuzaki Y, Nishikawa S, Hayashi S-I, Kunisada T, Sudo T, Kina H, Nakauchi H, Nishikawa S-I. Expression and function of c-kit in hematopoietic progenitor cells. J Exp Med 1991; 174: 63-71. MEDLINE

12 Bernstein ID, Andrew RG, Zsebo KM. Recombinant human stem cell factor enhances formation of colonies by CD34⁺ and CD34⁺lin^- cells, and the generation of colony-forming cell progeny from CD34⁺lin^- cells cultured with interleukin-3, granulocyte colony-stimulating factor, or granulocyte-macrophage colony-stimulating factor. Blood 1991; 77: 2316-2321. MEDLINE

13 Palacios R, Steinmetz M. IL3-dependent mouse clone that express B-220 surface antigen, contain Ig genes in germ-line configration, and generate B lymphocytes in vivo. Cell 1985; 41: 727-734. MEDLINE

14 Tanosaki S, Inokuchi K, Shimada T, Dan K. Relation between microsatellite instability and N-ras mutation and duration of disease free survival in patients with acute leukemia. Cancer 1998; 83: 475-481. Article MEDLINE

15 Abo J, Inokuchi K, Dan K, Nomura T. p53 and N-ras mutations in two new leukemia cell lines established from a patient with multilineage CD7-positive acute leukemia. Blood 1993; 82: 2829-2836. MEDLINE

16 Inokuchi K, Fukaki M, Hanawa H, Tanosaki S, Yamaguchi H, Iwakiri R, Nomura T, Dan K. Heterogenous expression of bcr-abl fusion mRNA in a patient with Philadelphia-chromosome-positive acute lymphoblastic leukemia. Br J Haematol 1997; 97: 837-840. MEDLINE

17 Furitsu T, Tsujimura T, Yono T, Ikeda H, Kitayama H, Koshimizu U, Sugahara H, Butterfield JH, Ashman LK, Kanayama Y, Matsuzawa Y, Kitamura Y, Kanakura Y. Identification of mutations in the coding sequence of the proto-oncogene c-kit in a human mast cell leukemia cell line causing ligand-dependent activation of c-kit. J Clin Invest 1993; 92: 1736-1744. MEDLINE

18 Mizushima S, Nagata S. pEF-BOS, a powerful mammalian expression vector. Nucleic Acids Res 1990; 18: 5322. MEDLINE

19 Inokuchi K, Miyake K, Takahashi H, Dan K, Nomura T. DCC protein expression in hematopoietic cell populations and its relation to leukemogenesis. J Clin Invest 1996; 97: 852-857. MEDLINE

20 Martin-Henao GA, Quiroga R, Sureda A, Garcia S. CD7 expression on CD34+ cells from chronic myeloid leukemia in chronic phase. Am J Hematol 1999; 61: 178-186. MEDLINE

21 Hirota S, Isozaki K, Moriyama Y, Hashimoto K, Nishida T, Ishiguro S, Kawano K, Hanada M, Kurata A, Takeda M, Tunio GM, Matsuzawa Y, Kanakura Y, Shinomura Y, Kitamura Y. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 1998; 279: 577-580. Article MEDLINE

22 Tsujimura T, Furitsu T, Morimoto M, Isozaki K, Nomura S, Matsuzawa Y, Kitamura Y, Kanakura Y. Ligand-independent activation of c-kit receptor tyrosine kinase in a murine mastocytoma cell line P-815 generated by a point mutation. Blood 1994; 83: 2619-2626. MEDLINE

23 Sakurai S, Fukasawa T, Chong JM, Tanaka A, Fukayama M. C-kit gene abnormalities in gastrointestinal stromal tumors (tumors of interstitial cells of Cajal). Jpn J Cancer Res 1999; 90: 1321-1328. MEDLINE

24 Tsujimura T, Morimoto M, Hashimoto K, Moriyama Y, Kitayama H, Matsuzawa Y, Kitamura Y, Kanakura Y. Constitutive activation of c-kit in FMA3 murine mastocytoma cells caused by deletion of seven amino acids at the juxtamembrane domain. Blood 1996; 87: 273-283. MEDLINE

25 Kiyoi H, Naoe T, Nakano Y, Yokota S, Minami S, Miyawaki S, Asou N, Kuriyama K, Jinnai I, Shimazaki C, Akiyama H, Saito K, Oh H, Moyoji T, Omoto E, Saito H, Ohno R, Ueda R. Prognostic implication of FLT3 and N-RAS gene mutations in acute myeloid leukemia. Blood 1999; 93: 3074-3080. MEDLINE

26 Matthews W, Jordan CT, Wiegand GW, Pardoll D, Lemischka IR. A receptor tyrosine kinase specific to hematopoietic stem and progenitor cell-enriched populations. Cell 1991; 65: 1143-1152. MEDLINE

27 Hayakawa F, Towatari M, Kiyoi H, Tanimoto M, Kitamura T, Saito H, Naoe T. Tandem-duplicated Flt3 constitutively activates STAT5 and MAP kinase and introduces autonomous cell growth in IL-3-dependent cell lines. Oncogene 2000; 19: 624-631. MEDLINE

Figure 1 RT-PCR-SSCP analysis and sequencing of the cloned RT-PCR products of the c-kit gene of the CML cases detected abnormalities. (a) and (b) show the RT-PCR-SSCP and sequence of the RT-PCR products of case 1. (c) and (d) show the RT-PCR-SSCP and sequence of the RT-PCR products of case 2. (a) Lane 1: RT-PCR product of RNA sample from chronic phase; lane 2: RT-PCR product of blastic crisis; lane 3: RT-PCR product of hair root; lane 4: RT-PCR product of skin. The arrow indicates an abnormal band. (b) Sequence of the cloned RT-PCR product of blastic crisis sample. The arrow indicates a one-base change (codon 564, AATAAG). (c) Lane 1: RT-PCR product of RNA sample from chronic phase; lane 2: RT-PCR product of blastic crisis; lane 3: RT-PCR product of hair root. The arrow indicates an abnormal band. (d) Sequence of the cloned RT-PCR product of blastic crisis sample. The arrow indicates a one-base change (codon 541, ATGCTG).

Figure 2 Flow cytometric analysis and tyrosine phosphorylation of Ba/F3^Vector, BaF3- KIT^WT, BaF3- KIT^V814, BaF3- KIT^K563 and BaF3- KIT^L540. (a) Flow cytometric analysis of the surface binding of a monoclonal anti-c-kit antibody to Ba/F3 cells transfected with pEF-BOS vector alone (Vector), pEF-BOS-KIT^WT, pEF-BOS-KIT^V814, pEF-BOS-KIT^K563 or pEF-BOS-KIT^L540. Cells were incubated with either FITC-conjugated negative-control antibody (----) or 2B8 (¾¾), washed and analyzed on a FACScan. (b) Tyrosine phosphorylation of KIT in Ba/F3 cells expressing KIT^WT, KIT^V814, KIT^K563 and KIT^L540. KIT was immunoprecipitated with an anti-c-kit MoAb (2B3) from lysates of the indicated cells before and after stimulation with rmSCF (0, 0.1, 1.0 or 100 ng/ml). The immunoprecipitates were separated by SDS-PAGE and then immunoblotted with antiphosphotyrosine (anti-P-Tyr) MoAb (second to fifth rows). The immunoprecipitates of Ba/F3 cells were cultured with 0 ng/ml of SCF, divided into two aliquots, separated by SDS-PAGE, and then immunoblotted with anti-c-kit (2B3) (first row). All immunoprecipitates were detected with HPR-conjugated goat anti-mouse IgG and visualized by autoradiography.

Figure 3 Semiquantification of tyrosine-phosphorylation signals of KIT in Ba/F3 cells expressing KIT^WT, KIT^V814, KIT^K563 and KIT^L540.

Figure 4 Proliferation of Ba/F3 cells in response to various concentrations of rmSCF (left panel) and rmIL3 (right panel). Quadruplicate aliquots of cells expressing KIT^WT, KIT^V814, KIT^K563 or KIT^L540 were cultured with each factor, and cell proliferation was measured by ³H-thymidine incorporation assay. The results are shown as the mean ± s.d. for three separate experiments. Wild, V814, L563 and L540 mean Ba/F3 cells expressing KIT^WT, KIT^V814, KIT^K563 and KIT^L540, respectively. Vector means Ba/F3 cells transfected with the pEF-BOS vector. Significantly higher values were obtained for Ba/F3 cells expressing KIT^K563 or KIT^L540 compared with those expressing KIT^WT. *P < 0.05; **P < 0.01.

Table 1 Frequency of c-kit abnormality in CML patients and hematologically healthy volunteers

Table 2 Summary of CML patients with a c-kit abnormality

Table 3 Summary of total CML patients of the two groups