No abstract

(3,4), and two SH2-containing adapter proteins, Grb2 and Shc, have been implicated in its activation. Specifically, these proteins have been shown to bind directly to tyrosine-phosphorylated receptors (5-7) or SH2 docking proteins (such as the insulin receptor substrate 1) (8). Grb2, a 25-kDa protein with two SH3 domains flanking one SH2 domain, shuttles the Ras guanine nucleotide exchange factor, Sosl, to activated receptors (or to insulin receptor substrate 1) (5, 7-11) so that Sosl can activate Ras by catalyzing the exchange of GDP for GTP (5, 9-11). Shc, another widely expressed protein that contains an N-terminal phosphotyrosine binding (PTB) domain (12-16) and a C-terminal SH2 domain (17), can associate, in its tyrosine-phosphorylated form, with Grb2-Sosl complexes and may increase Grb2-Sosl interactions after growth factor stimulation (9,18,19 MATERIALS AND METHODS Reagents. COS-cell-derived murine IL-3 and GM-CSF were produced as described (29). Rabbit antiserum to the Shcassociated p145 protein was generated by immunizing rabbits with a 15-mer obtained from amino acid sequencing (VPAE-GVSSLNEMINP) and crosslinked to keyhole limpet hemocyanin with glutaraldehyde (30). The glutathione S-transferase (GST) fusion proteins, consisting of the 27-kDa N-terminal and of GST linked to the Grb2 N-terminal (amino acid residues 5-56) or C-terminal (residues 163-215) SH3 domains of Grb2, were expressed in Escherichia coli in pGEX-2T plasmids (Pharmacia/LKB) and the fusion proteins were recovered from clarified lysates with glutathione (GSH)-agarose beads as described (31 Abbreviations: GM-CSF, granulocyte-macrophage colony-stimulating factor; GSH, glutathione; GST, glutathione S-transferase; IL-3, interleukin 3; NRS, normal rabbit serum; 5-ptase, inositol polyphosphate 5-phosphatase; Ptd1ns(3,4,5)P3, phosphatidylinositol 3,4,5-trisphosphate; Ins(1,3,4,5)P4, inositol 1,3,4,5-tetrakisphosphate; SH2, src homology 2 domain; PTB, phosphotyrosine binding; p145, the Shc-associated 145-kDa protein.

The molecular characterization of leukemia has demonstrated that genetic alterations in the leukemic clone frequently fall into 2 classes, those affecting transcription factors (e.g., AML1-ETO) and mutations affecting genes involved in signal transduction (e.g., activating mutations of FLT3 and KIT). This finding has favored a model of leukemogenesis in which the collaboration of these 2 classes of genetic alterations is necessary for the malignant transformation of hematopoietic progenitor cells. The model is supported by experimental data indicating that AML1-ETO and FLT3 length mutation (FLT3-LM), 2 of the most frequent genetic alterations in AML, are both insufficient on their own to cause leukemia in animal models. Here we report that AML1-ETO collaborates with FLT3-LM in inducing acute leukemia in a murine BM transplantation model. Moreover, in a series of 135 patients with AML1-ETO-positive AML, the most frequently identified class of additional mutations affected genes involved in signal transduction pathways including FLT3-LM or mutations of KIT and NRAS. These data support the concept of oncogenic cooperation between AML1-ETO and a class of activating mutations, recurrently found in patients with t(8;21), and provide a rationale for therapies targeting signal transduction pathways in AML1-ETO-positive leukemias. IntroductionThe cloning of recurring chromosomal translocations and, increasingly, the molecular characterization of point mutations in patients with acute leukemia have substantially contributed to the understanding of the pathogenesis of this disease. In acute myeloid leukemia (AML), chromosomal translocations most frequently target transcription factors involved in the regulation of normal hematopoietic differentiation, whereas point mutations often affect genes involved in signal transduction pathways associated with cell proliferation (1-3). The systematic analyses of genetic alterations in patients with AML have demonstrated that genetic lesions of more than 1 transcriptional regulator, such as AML1-ETO (RUNX1-MTG8), HOX fusion genes, or PML-RARA, rarely occur in the leukemic clone. Similarly, patients with concurrent mutations of FLT3, KIT, or NRAS are rare. However, there are numerous examples in which fusion genes are identified together with activating mutations of receptor tyrosine kinases, exemplified by PML-RARA and the FLT3 length mutation (FLT3-LM), which occur together in up to 35% of all patients with t(15;17)-positive AML (4).

HOX genes, notably members of the HOXA cluster, and HOX cofactors have increasingly been linked to human leukemia. Intriguingly, HOXD13, a member of the HOXD cluster not normally expressed in hematopoietic cells, was recently identified as a partner of NUP98 in a t(2;11) translocation associated with t-AML/MDS. We have now tested directly the leukemogenic potential of the NUP98-HOXD13 t (