Abnormality of c-kit oncoprotein in certain patients with chronic myelogenous leukemia--potential clinical significance.

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 (AAT --> AAG, Asn --> Lys), and six cases had the same base abnormality at codon 541 (ATG --> CTG, Met --> Leu) 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 (ATG --> CTG, Met --> Leu) 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(L540)showed 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.

A cytoplasmic serine protein kinase binds and may regulate the Fanconi anemia protein FANCA.

Fanconi anemia (FA) is an autosomal recessive disease with congenital anomalies, bone marrow failure, and susceptibility to leukemia. Patient cells show chromosome instability and hypersensitivity to DNA cross-linking agents. At least 8 complementation groups (A-G) have been identified and 6 FA genes (for subtypes A, C, D2, E, F, and G) have been cloned. Increasing evidence indicates that a protein complex assembly of multiple FA proteins, including FANCA and FANCG, plays a crucial role in the FA pathway. Previously, it was reported that FANCA was phosphorylated in lymphoblasts from normal controls, whereas the phosphorylation was defective in those derived from patients with FA of multiple complementation groups. The present study examined phosphorylation of FANCA ectopically expressed in FANCA(-) cells. Several patient-derived mutations abrogated in vivo phosphorylation of FANCA in this system, suggesting that FANCA phosphorylation is associated with its function. In vitro phosphorylation studies indicated that a physiologic protein kinase for FANCA (FANCA-PK) forms a complex with the substrate. Furthermore, at least a part of FANCA-PK as well as phosphorylated FANCA were included in the FANCA/FANCG complex. Thus, FANCA-PK appears to be another component of the FA protein complex and may regulate function of FANCA. FANCA-PK was characterized as a cytoplasmic serine kinase sensitive to wortmannin. Identification of the protein kinase is expected to elucidate regulatory mechanisms that control the FA pathway.

Chromosomal breakage syndromes and the BRCA1 genome surveillance complex.

Chromosomal instability can occur when the DNA damage response and repair process fails, resulting in syndromes characterized by growth abnormalities, hematopoietic defects, mutagen sensitivity, and cancer predisposition. Mutations in ATM, NBS1, MRE11, BLM, WRN, and FANCD2 are responsible for ataxia telangiectasia (AT), Nijmegen breakage syndrome, AT-like disorder, Bloom and Werner syndrome, and Fanconi anemia group D2, respectively. This diverse group of disorders is thought to be linked through protein interactions with the breast cancer tumor susceptibility gene product, BRCA1. BRCA1 forms a multi-subunit protein complex referred to as the BRCA1-associated genome surveillance complex (BASC), which includes DNA damage repair proteins such as MSH2-MSH6 and MLH1, as well as ATM, NBS1, MRE11, and BLM. Although still controversial, this finding suggests similarities in the pathogenesis of the human chromosome breakage syndromes and a complementary role for each protein in DNA structure surveillance or damage repair.

Fanconi anemia protein, FANCG, is a phosphoprotein and is upregulated with FANCA after TNF-alpha treatment.

Fanconi anemia (FA) is a genetic syndrome characterized by bone marrow failure, birth defects, and a predisposition to malignancy. At this time, six FA genes have been identified, and several gene products have been found to interact in a protein complex. FA cells appear to overexpress the proinflammatory cytokine, tumor necrosis factor-alpha (TNF-alpha). We therefore examined the effects of TNF-alpha on the regulation of FA complementation group proteins, FANCG and FANCA. We found that treatment with TNF-alpha induced FANCG protein expression. FANCA was induced concurrently with FANCG, and the FANCA/FANCG complex was increased in the nucleus following TNF-alpha treatment. Inactivation of inhibitory kappa B kinase-2 modulated the expression of FANCG. We also found that both nuclear and cytoplasmic FANCG fractions were phosphorylated. These results show that FANCG is a phosphoprotein and suggest that the cellular accumulation of FA proteins is subject to regulation by TNF-alpha signaling.

The IVS4 + 4 A to T mutation of the fanconi anemia gene FANCC is not associated with a severe phenotype in Japanese patients.

Fanconi anemia (FA) is an autosomal recessive disease characterized by congenital anomalies, aplastic anemia, and a susceptibility to leukemia. There are at least 8 complementation groups (A through H). Extensive analyses of the FA group C gene FANCC in Western countries revealed that 10% to 15% of FA patients have mutations of this gene. The most common mutation is IVS4 + 4 A to T (IVS4), a splice mutation in intron 4, which has been found only in patients of Ashkenazi Jewish ancestry. When we screened 29 Japanese patients (20 unrelated patients and 4 families) using polymerase chain reaction-single strand conformation polymorphism, we found 8 unrelated patients homozygous for IVS4. This is apparently the first non-Ashkenazi-Jewish population for whom this mutation has been detected. The Ashkenazi Jewish patients homozygous for IVS4 have a severe phenotype, in comparison with other FA patients. Our analyses of Japanese patients indicate no significant difference between IVS4 homozygotes and other patients with regard to severity of a clinical phenotype. Thus, ethnic background may have a significant effect on a clinical phenotype in FA patients carrying the same mutation. (Blood. 2000;95:1493-1498)

Engraftment of hematopoietic progenitor cells transduced with the Fanconi anemia group C gene (FANCC).

Fanconi anemia (FA) is an autosomal recessive disorder that leads to aplastic anemia. Mutations in the FANCC gene account for 10-15% of cases. FA cells are abnormally sensitive to DNA-damaging agents such as mitomycin C (MMC). Transfection of normal FANCC into mutant cells corrects this hypersensitivity and improves their viability in vitro. Four FA patients, representing the three major FANCC mutation subgroups, were entered into a clinical trial of gene transduction aimed at correction of the hematopoietic defect. Three patients received three or four cycles of gene transfer, each consisting of one or two infusions of autologous hematopoietic progenitor cells that had been transduced ex vivo with a retroviral vector carrying the normal FANCC gene. Prior to infusion, the FANCC transgene was demonstrated in transduced CD34-enriched progenitor cells. After infusion, FANCC was also present transiently in peripheral blood (PB) and bone marrow (BM) cells. Function of the normal FANCC transgene was suggested by a marked increase in hematopoietic colonies measured by in vitro cultures, including colonies grown in the presence of MMC, after successive gene therapy cycles in all patients. Transient improvement in BM cellularity coincided with this expansion of hematopoietic progenitors. A fourth patient, who received a single infusion of transduced CD34-enriched BM cells, was given radiation therapy for a concurrent gynecologic malignancy. The FANCC transgene was detected in her PB and BM cells only after recovery from radiation-induced aplasia, suggesting that FANCC gene transduction confers a selective engraftment advantage. These experiments highlight both the potential and difficulties in applying gene therapy to FA.

Establishment of a cell line with variant BCR/ABL breakpoint expressing P180BCR/ABL from late-appearing Philadelphia-positive acute biphenotypic leukemia.

In acute leukemia (AL) with a late-appearing Philadelphia (la-Ph) translocation, it is unclear whether these translocations arise from the same molecular event as classical Ph translocations. In order to elucidate the molecular events of la-Ph and subsequent translocations of la-Ph leukemia, we performed molecular analysis on the complex rearrangements, in a cell line, MY, which was established from bone marrow mononuclear cells of a patient with a la-Ph acute biphenotypic leukemia. This la-Ph, expressing an acute lymphoblastic leukemia (ALL)-type BCR/ABL transcript, produces a novel P180BCR/ABL fusion protein reflecting deletion of 174 bases (58 amino acids) encoded by the a2 exon of the ABL gene. An immune complex kinase assay showed that this protein had autophosphorylation activity. Fluorescence in situ hybridization (FISH) in conjunction with G-banding analysis revealed that the initial der(9)t(9;22)(q34;q11) progressed to a der(9)(9pter-->9q34::22q11-->22q13::5q11.2 -->5q15:: 10q23-->10qter) by, first, a three-way translocation among the der(9)t(9;22)(q34;q11), chromosome 5, and the normal chromosome 22, and then a subsequent translocation with chromosome 10. Moreover, both the end-stage leukemic cells of the patient and the MY cell line had another translocation, t(X;12)(p11.2;p13). The 12p breakpoint was located near the ETV6 gene by analysis of pulsed-field gel electrophoresis, but transcription of ETV6 was unaffected. Tumorigenicity analysis indicated that an additional translocation, t(2;3)(p16;q29), may have caused a more malignant clone, because only MY cells with the t(2;3)(p16;q29) were capable of growing subcutaneously in nude mice within 40 days. The molecular events of leukemogenesis and leukemic progression in the present la-Ph AL occurred by accumulation of unique translocations. This cell line, MY, expressing a novel variant P180BCR/ABL protein with a deletion of the a2 exon of the ABL gene, may be useful for elucidating the pathophysiology of this fusion protein and for studying ETV6-related leukemogenesis and t(2;3), as well as the molecular mechanisms of the complex translocations.

Minimal residual disease in acute myelogenous leukemia with PML/RAR alpha or AML1/ETO mRNA and phenotypic analysis of possible T and natural killer cells in bone marrow.

Here we studied minimal residual disease (MRD) of patients with acute myeloid leukemia (AML) who have PML/RAR alpha or AML1/ETO as well as the phenotypic analysis of lymphocyte subsets involved in antitumor immunity. Eight patients in long-term (LT; 3 to 15 years) and 15 patients in short-term (ST; up to 3 years) remission were studied. Using the reverse transcription-polymerase chain reaction (RT) assay, the limit of detection was 10(-5) to 10(-6) for PML/RAR alpha transcript and 10(-4) to 10(-5) for the AML1/ETO transcript. Simultaneously, T lymphocyte subsets and NK cells from the peripheral blood (PB) and bone marrow (BM) were investigated by flow cytometric analysis. Four of the eight patients in LT and 7 of the 15 patients in ST remission were MRD-positive. Although all MRD-positive patients in LT remission are still until now event-free, 3 of the 7 MRD-positive (MRD+) patients in ST remission soon relapsed. The total populations of CD4+, CD8+ and CD56+ [possible T-cell and natural killer (T/NK) populations] in the BM of ST patients and MRD+/LT patients were significantly (p < .01) low. The CD8+ CD28+ population showed the same tendency (p < .01-.02). The T/NK subsets in the BM of MRD-negative (MRD-) LT (MRD-/LT) patients showed similar numbers of cells as normal volunteers. Basically, the total percentage of the CD4+, CD8+ and CD56+ cell populations in the BM was increased and in the following order: MRD-/LT patients, normal volunteers, MRD+/LT patients and MRD+ or -/ST patients. The percentages of the T/NK-cell subsets in the PB were not significantly different among these groups. Thus, the difference of the possible T/NK-cell phenotype in the BM may strongly influence clinical and molecular remission. These results still remain to be confirmed by further studies of the functional anti-tumor immunity of T/NK cells of AML in remission.

Genotype configuration in a case of primary gastric lymphoma with T-cell phenotype.

T-cell malignant lymphoma of the gastrointestinal tract is rare. The genotype of gastric T-cell lymphoma remains unclear. The aim of this study was to elucidate the pathogenesis of a case of primary gastric T-cell lymphoma by using cytogenetics and molecular biology. Gastric biopsy specimens and lymphoma cells in the ascites were examined by immunocytology, cytogenetic analysis, and Southern blot analysis. The histological diagnosis of the gastric lymphoma was diffuse large cell type. T-cell markers were positive in immunocytochemistry of the gastric lymphoma cells and in FACS analysis of lymphoma cells in the ascites. All lymphoma cells in the ascites had complex abnormal karyotypes containing t(8;14)(q24;q32). Southern blot analysis revealed rearrangement of the IgH and C-MYC genes of the lymphoma cells in both the stomach and the ascites, but no comigration of the C-MYC with the JH locus could be detected. The TCR-beta and -gamma genes were in their germ-line configurations. In this patient, although the phenotype was T-cell lymphoma, the karyotype t(8;14)(q24;q32) and genotype had the characteristics of B-cell lymphoma. The unique B-cell genotype configuration and the C-MYC activation suggested that the cellular origin of this rare case of malignant lymphoma with a T-cell phenotype was quite immature lymphocytes.

Heterogenous expression of bcr-abl fusion mRNA in a patient with Philadelphia-chromosome-positive acute lymphoblastic leukaemia.

We performed molecular and cytogenetic analysis on a 56-year-old woman with Philadelphia chromosome (Ph1)-positive acute lymphoblastic leukaemia (ALL) having two types of major and minor bcr-abl (M-bcr-abl, m-bcr-abl/fusion mRNA at diagnosis. In the course of her disease, unexpected heterogenous bcr-abl fusion mRNA was detected by sequential analysis using the reverse transcription and polymerase chain reaction (RT-PCR). The RT-PCR analysis showed both M-bcr-abl (b2-a2 type) and m-bcr-abl at diagnosis. Although b2-a2 type M-bcr-abl disappeared after complete remission (CR) was achieved with intensive induction chemotherapy, expression of both m-bcr-abl and a new type of M-bcr-abl mRNA (b1-a2 type), which may have been produced through splicing out of exon b2, was detected in the early stage of CR. The leukaemic cells contained these heterogenous bcr-abl mRNAs in both the CR and relapsed state, and showed more stable expression of the m-bcr-abl type mRNA than that of the b2-a2 type. These findings of heterogenous bcr-abl mRNA due to alternative or missplicing during the disease course in the present ALL case may provide important evidence of disease progression.

Possible transforming activity of interferon regulatory factor 2 in tumorigenicity assay of NIH3T3 cells transfected with DNA from chronic myelogenous leukemia patients.

Little is known about the transforming gene identified in the genomic DNA of chronic myelogenous leukemia (CML) by the tumorigenicity assay. To detect a new transforming gene of CML, we re-investigated the transforming activity of interferon regulatory factor (IRF)-1 and -2 genes in the tumorigenicity assay of NIH3T3 cells transfected with genomic DNA of leukemic cells from 15 patients with CML (12 patients in the chronic phase, one in the blastic phase and two in both phases). We detected the functionally active IRF-2 gene only in the tumor DNA from two CML patients in the blastic phase. We did not detect integration of the IRF-1 gene in the DNA of any tumors derived from the CML patient samples, and also we detected no expression of human IRF-1 mRNA. Thus, NIH3T3 cells may have been transformed due to integration of the functionally active IRF-2 gene from CML patients in the blastic phase. We surmise that there is a possibility that the IRF-2 gene may be involved in the evolution of the blastic phase of CML.

Dual rearrangement of immunoglobulin and T-cell receptor genes in a case of splenic lymphoma with villous lymphocytes.

We report here a case of "splenic lymphoma with villous lymphocytes" (SLVL) which exhibited both B- and T-cell phenotypes and genotypes. The patient was a 73-year-old man. Physical examination revealed splenomegaly and lymphadenopathy. The white blood cell count was 55.2 x 10(9)/L with 70.5% atypical lymphocytes, having cytoplasmic villi, characteristic of SLVL. The atypical cells infiltrated both the red and white pulps. Immunological analysis of the peripheral leukocytes showed both B- and T-cell phenotypes (CD5,CD11c,CD19,CD20,HLA-DR, SmIgM and lambda positive). DNA analysis revealed a dual rearrangement of the immunoglobulin heavy chain gene and T-cell receptor beta gene. SLVL has been identified as a B-cell leukemia with a relatively benign clinical course. This case had both B- and T-cell pheno- and genotypes with a progressive course. To the best of our knowledge, no case of SLVL with dual genotypes has ever been reported.

Possible correlation of b3-a2-type bcr-abl messenger RNA defined by semiquantitative RT-PCR to platelet and megakaryocyte counts in Philadelphia-positive chronic myelogenous leukemia.

Thirty-five patients with Philadelphia chromosome (Ph1)-positive chronic myelogenous leukemia (CML) were classified on the basis of the fusion pattern of bcr-abl mRNA determined by the reverse-transcriptase-polymerase chain reaction (RT-PCR) method. Semiquantitative assay of the bcr exon 2/abl exon 2 fused mRNA (b2-a2) and bcr exon 3/abl exon 2 fused mRNA (b3-a2) resulted in 21 patients showing b3-a2 type mRNA, seven showing b2-a2 type and seven showing coexpression. Quantification of the autoradiographic signals of amplified products was estimated using an MCID image analysis system. The relative intensity was defined as the ratio of bcr-abl signal to that of beta-actin. The relationship between the semiquantified bcr-abl mRNA and the platelet/megakaryocyte counts was analyzed. A possible correlation was found between the semiquantified b3-a2 type mRNA and the platelet (p < .05, N = 28) and megakaryocyte (p < .05, N = 13) counts of these patients. This finding suggests the possibility that b3-a2 mRNA may affect the thrombopoietic activity in Ph1-positive CML in a dose-response manner.

Megakaryocyte, erythroid and granulocyte-macrophage colony formation in myelodysplastic syndromes.

Bone marrow progenitor cell assays of three cell lineages, i.e., colony-forming unit megakaryocytes (CFU-Meg), burst-forming unit erythrocytes (BFU-E) and colony-forming unit granulocyte-macrophages (CFU-GM), were performed for 21 patients with myelodysplastic syndromes (MDS). Markedly reduced or absent colony formation was found in 67% of the patients for CFU-Meg and all patients except 2 with refractory anemia (RA) for BFU-E. Abnormal CFU-GM colony formation was found in only 5 of 12 patients with RA and RA with ring sideroblasts, in contrast to all of the RA patients with excess of blasts and excess of blasts in transformation. Defective colony formation of all three cell lineages was seen in 63% of the MDS patients. The colony number of CFU-Meg correlated significantly with the numbers of both BFU-E and CFU-GM. These findings indicate that hematopoiesis in MDS patients is disturbed due to a qualitative or quantitative defect at the multipotent stem cell level.

[Two cases of acute leukemia with t(6;9) (p23;q34)].

Two cases of acute leukemia with a t (6;9) (p23;34) chromosome abnormality are reported. The first case was a 34-year-old female who was hospitalized in October 1989. A diagnosis of FAB-M1 was made. Chromosomal analysis of the bone marrow cells showed a 46, XX, t (6;9) (p23;q34). Complete remission was achieved after two courses of BHAC-DMP therapy. In September 1991, at the time of relapse, chromosomal analysis revealed two abnormal clones consisting of a 46, XX, t (6;9) (p23;q34), -12, -17, +der (12) t (12;17) (p11.2;q11.2) with a residual normal clone. She died in February 1992. The second case was a 42-year-old male who was hospitalized in January 1990. He was diagnosed as having RAEB. Chromosomal analysis of the bone marrow cells showed 46, XY, t (6;9) (p23;q34). Three months later, the disease progressed to acute leukemia accompanied by leg ulceration with leukemic cell infiltration. Small-dose ara-C therapy was given, but with no effect. After two subsequent courses of therapy with low-dose etoposide, complete remission was achieved. Four months later, relapse occurred, and the patient died of sepsis in February 1991. In the literature, 31 cases of myeloproliferative disorders with t (6;9) have been reported.

Relationship of the type of bcr-abl hybrid mRNA to clinical course and transforming activity in Philadelphia-positive chronic myelogenous leukemia.

We studied the type of bcr-abl mRNA for 34 patients with chronic myelogenous leukemia and analyzed for correlations among the mRNA type, the clinical outcome and the transforming activity using the tumorigenicity assay. There was no difference in the distribution of the mRNA-types (b2-a2 and b3-a2) between clinical phases. We found no correlation between the two types of bcr-abl mRNA and the chronic phase duration or survival. The DNA from 12 of 20 chronic phase patients and all five blastic phase patients showed transforming activity. Although there was no difference in the positive rate of transforming activity among the two mRNA-type groups, the blastic phase patients showed a tendency to have higher transforming activity.

N-RAS activation in the terminal stage of undifferentiated chronic myeloproliferative disease.

The relationship between activation of the N-RAS gene and the leukemic progression of undifferentiated chronic myeloproliferative disease (UCMPD) was investigated in a 71-year-old male. Hematologically, it was difficult to differentiate the UCMPD from chronic myelogenous leukemia. Chromosomal analysis revealed no Philadelphia chromosome (Ph1-), and DNA analysis revealed no BCR rearrangement (BCR-) either at the beginning or in the terminal stages of the disease. We performed a tumorigenicity assay, using NIH3T3 cells, and molecular analysis, using the polymerase chain reaction (PCR) and direct sequencing. The DNA of leukemic cells at the beginning of the leukemic progression did not show any abnormalities, but at the terminal stage of the disease the DNA showed a point mutation in codon 12 (GGT----GCT) of the N-RAS gene. Interestingly, a codon 13 mutation (GGT----GTT) was also detected by tumorigenicity assay. These observations suggest that the activated N-RAS gene contributes to the hematologic progression of UCMPD.

A possible correlation between the type of bcr-abl hybrid messenger RNA and platelet count in Philadelphia-positive chronic myelogenous leukemia.

The Philadelphia (Ph1) chromosome, in which the hybrid bcr-abl gene is formed, is thought to be the initial event in chronic myelogenous leukemia (CML). The position of the breakpoint within the breakpoint cluster region (bcr) on Ph1 chromosome and the splicing pattern determine the species of the fused bcr-abl messenger RNA (mRNA). We tried to detect the two types of fused mRNAs in 57 chronic-phase cases of Ph1-positive CML using the polymerase chain reaction procedure (RT-PCR). The bcr exon 2/abl exon 2 fused mRNA (b2-a2) was detected in 17 patients, the bcr exon 3/abl exon 2 fused mRNA (b3-a2) was detected in 34 patients, and both types of mRNA were detected in six patients. The platelet counts of patients who expressed b3-a2 mRNA or both types were significantly higher than those of patients who expressed only b2-a2 (841.5 v 373.5 x 10(9)/L; P less than .015), although there was no significant difference in the white blood cell counts or hemoglobin. This finding suggests a possibility that the type of bcr-abl mRNA may affect the thrombopoietic activity in CML.

Transforming genes and chromosome aberrations in therapy-related leukemia and myelodysplastic syndrome.

The presence of activated transforming genes was investigated in four patients with therapy-related leukemia and in three with therapy-related myelodysplastic syndrome. DNA of bone marrow cells from six of the patients exhibited transforming activity in the tumorigenicity assay. Five of the six patients who were positive in the tumorigenicity assay contained activated N-ras oncogenes, and three contained activated K-ras oncogenes. Thus, concurrent activation of N-ras and K-ras oncogenes was observed in two patients. In vitro DNA amplification followed by oligonucleotide dot-blot analysis was used to investigate mutations in codons 12, 13, and 61 of the N-ras and K-ras oncogenes. Two patients exhibited an N-ras mutation, substituting aspartic acid (GAT) for glycine (GGT), and three patients exhibited an N-ras codon 13 mutation, substituting valine (GTT) for glycine. Two patients exhibited K-ras codon 12 mutations, substituting aspartic acid (GAT) or cysteine (TGT) for glycine (GGT), respectively, and one case exhibited a K-ras codon 61 mutation, substituting lysine (AAA) for glutamic acid (CAA). Cytogenetic analysis revealed that loss of chromosome 7 was frequent (four patients: 57%). Our data indicate that activation of N-ras and K-ras genes, as well as loss of heterozygosity for specific alleles on chromosome 7, plays a more important role in the leukemogenesis of both therapy-related leukemia and myelodysplastic syndrome.