Loss of C/EBP alpha and favorable prognosis of acute myeloid leukemias: a biological paradox.

Abstract

Acute myeloid leukemia (AML) is characterized by the accumulation of immature precursors arrested at different stages of myelopoiesis. Normal myeloid differentiation depends on the coordinated activity of transcription factors that control the commitment and maturation of the hematopoietic stem/progenitor cell pool by directing the expression of lineage-determining genes, including those encoding the receptor for myeloid-specific cytokines, in a timeand stage-specific manner [1]. Although it seems intuitive that genetic or functional inactivation of transcription factors regulating proliferation, survival, and differentiation of myeloid progenitors is requisite for the pathogenesis of AML, a direct link with myeloid leukemogenesis has been formally demonstrated only for a few transcription factors [2]. Among these, the CCAAT enhancer binding protein alpha (C/EBP ) of the basic region leucine zipper transcription regulators recently has been reported to be inactivated in approximately 4% to 15% of patients with AML [2]. C/EBP is an essential factor for granulocytic differentiation of common myeloid progenitors [3,4]; in fact, Cebpa-null mice lack neutrophils and eosinophils but retain monocytes [3]. Fetal liver cells from Cebpa-null mice do not form colonies in response to granulocyte colony-stimulating factor (G-CSF) but develop into immature myeloid cells in response to GMCSF [3]. Accordingly, ectopic C/EBP expression induces granulocytic differentiation and arrests erythroid differentiation of human CD34 cells [5]. The mechanism whereby C/EBP induces granulocytic differentiation of myeloid progenitors depends on its transcription activatory and on growth suppressive functions that require the integrity of the C/EBP basic region-leucine zipper and transcriptional activation domains that are involved in DNA binding, homodimerization and interaction with E2F, p21 and Cdk2/Cdk4 [6-8]. Notably, inhibition of E2F activity by C/EBP leads to downregulation of c-Myc, an important step required for granulocytic differentiation [9,10]. Two different groups of investigators have initially reported CEBPA mutations in AML patients [11,12]. These mutations were usually associated with the FrenchAmerican-British (FAB) subtypes M1 or M2, suggesting a block at a specific stage of the myelocytic differentiation. Approximately half of the described mutations were clustered in the N-terminus, leading to either a truncated nonfunctional protein or increased expression of the dominant negative isoform p30C/EBP [11,12]. The remaining mutations were identified at the C-terminus of C/EBP and predicted mutant proteins lacking DNA binding or homodimerization activities [11,12]. Interestingly, in these and subsequent studies, some patients presented biallelic mutations at the C-terminus, whereas others were either heterozygous for distinct mutations or found to have a N-terminal mutation associated with a C-terminal mutation [13,14]. The prognostic value of these findings has only recently been recognized. Three studies, including that of Frohling et al [15] reported in this issue of the Journal of Clinical Oncology, have linked CEBPA mutations with a favorable outcome in AML. Preudhomme et al [16] reported 22 different CEBPA mutations in 15 (11.1%) of 135 patients enrolled in the Acute Leukemia French Association (ALFA)-9000 trial. Most of these mutations were associated with FAB M1 or M2 phenotypes and intermediate cytogenetic risk by the MRC classification [17]. In this study, the presence of CEBPA mutations was a favorable predictor for overall and disease-free survival in both univariate and multivariate analyses. Similar data were subsequently reported by the Dutch-Belgium HematologyOncology Cooperative Group (HOVON), which identified 12 AML patients (4.3%) with CEBPA mutations at the C-terminus of the 277 patients enrolled onto the HOVON AML-4 and AML-29 trials [14]. Finally, the study by Frohling et al has now shown that mutations of CEBPA are independent predictors for favorable clinical outcome in a homogeneously treated population of AML patients with normal cytogenetics [15]. JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 22 NUMBER 4 FEBRUARY 15 2004