AML1-ETO and C-KIT mutation/overexpression in t(8;21) leukemia: implication in stepwise leukemogenesis and response to Gleevec

Abstract

To explore the genetic abnormalities that cooperate with AML1-ETO (AE) fusion gene to cause acute myeloid leukemia (AML) with t(8;21), we screened a number of candidate genes and identified 11 types of mutations in C-KIT gene (mC-KIT), including 6 previously undescribed ones among 26 of 54 (48.1%) cases with t(8;21). To address a possible chronological order between AE and mC-KIT, we showed that, among patients with AE and mC-KIT, most leukemic cells at disease presentation harbored both genetic alteration, whereas in three such cases investigated during complete remission, only AE, but not mC-KIT, could be detected by allele-specific PCR. Therefore, mC-KIT should be a subsequent event on the basis of t(8;21). Furthermore, induced expression of AE in U937-A/E cells significantly up-regulated mRNA and protein levels of C-KIT. This may lead to an alternative way of C-KIT activation and may explain the significantly higher C-KIT expression in 81.3% of patients with t(8;21) than in patients with other leukemias. These data strongly suggest that t(8;21) AML follows a stepwise model in leukemogenesis, i.e., AE represents the first, fundamental genetic hit to initiate the disease, whereas activation of the C-KIT pathway may be a second but also crucial hit for the development of a full-blown leukemia. Additionally, Gleevec suppressed the C-KIT activity and induced proliferation inhibition and apoptosis in cells bearing C-KIT N822K mutation or overexpression, but not in cells with D816 mC-KIT. Gleevec also exerted a synergic effect in apoptosis induction with cytarabine, thus providing a potential therapeutic for t(8;21) leukemia.

Fig. 1.

Coexistence of t(8;21) and mC-KITs in leukemic cells of de novo patients and persistence of AE but absence of mC-KIT in patients in CR. (A and B) Dual color FISH was performed, and 1,000 cells including interphase and metaphase were analyzed. The fusion signals (arrow) is seen in 96% in case 1 and 87.4% in case 2 of leukemic cells. Green signal, AML1 gene; orange signal, ETO gene. (C and D) Analysis of the peak height ratio of wild-type/mutant alleles in chromatogram of genomic DNA from leukemic cells of the two cases reveals that ≈85% in case 1 and 70% in case 2 of leukemic cells carried a mutant C-KIT allele. (E) The allele-specific RT-PCR is performed to detect the mutant alleles in three patients with t(8;21) leukemia at de novo and in remission. The primers used have an extra 2′-O,4′-C-methylene bridge at the ribose ring of nucleic acids at the 3′ end. The assays for Kasumi-1 (Kas) cells, which have a N822K (T → A) type mC-KIT, clearly show that these primers are sensitive and allele specific. At de novo, mC-KITs and AE are positive in the three patients, whereas the mC-KIT becomes undetectable and AE remains positive in CR.

Fig. 2.

Reduction of AE in CR and expression of C-KIT in t(8;21) leukemia. (A) AE expression in patients in CR decreased significantly compared to that in disease presentation. (B) Median of C-KIT log(CN) in t(8;21) and non-t(8;21) leukemias. (C and D) Immunohistochemical analysis of CD117 expression on bone marrow specimens of leukemia patients. (C) CD117-positive cells and their staining intensity are higher in patients with t(8;21) and mC-KIT than in those with t(8;21) and wtC-KIT, whereas the percentage of CD117-positive cells and their staining intensity are higher in patients with t(8;21) and wtC-KIT than in patients without t(8;21). (D) Accordingly, the immunoreactivity score (IRS) and staining intensity is higher in t(8;21) leukemia than in other hemopathies. (E) The C-KIT PTK activity is higher in t(8;21) leukemia than in other subtypes of leukemias.

Fig. 3.

Induction of C-KIT expression by AE in vitro. (A) In U937-A/E cells treated with PA, the expression of AE at mRNA level increases gradually, whereas TGF-β1 is down-regulated at an early stage. The expression of C-KIT is up-regulated subtly at early stage (6–24 h) and dramatically at late stage (48 h). (B) At the protein level, the induced expression of AE down-regulates TGF-β1 and up-regulates phosphorylated (pC-KIT) and unphosphorylated C-KIT significantly (Upper). However, activation of C-KIT by SCF (μg/liter) do not modify AE and TGF-β1 expression (Lower). PA, ponasterone A (in μM); Gle, Gleevec (in μM). (C) The C-KIT PTK activity, which is reflected by absorbance at 450 nm on a spectrophotometer, is enhanced significantly 12 h after PA treatment.

Fig. 4.

Effects of Gleevec on t(8;21) leukemic cells. (A) Gleevec induces apoptosis of Kasumi-1 cells in a dose- and time-dependent manner. Gle, Gleevec, given in μM; time is given in hours. (B) Gleevec suppresses C-KIT PTK activity of the Kasumi-1 cell. (C–F) Effects of Gleevec and/or cytarabine (Ara-C, in μM) on primary leukemic cells from patients with t(8;21) leukemia. These effects are detected by annexin V assay (C–E) and analysis of morphological changes (F).