Arrested Differentiation in Acute Myeloid Leukemia (AML) with Silenced Ankyrin Repeat and SOCS Box Protein 3 (ASB3) Expression

Abstract 1232

Acute myeloid leukemia (AML) is a hematological malignancy of the myeloid lineage characterized by an accumulation of abnormal immature white blood cells as a result of arrested myeloid differentiation. In the clinical setting, promyelocytic leukemic cells can be induced to differentiate into mature granulocytes by all-trans retinoic acid (ATRA), which acts by targeting oncogenic transcription factors. Expression profiling has previously identified ASB2, a member of the ankyrin repeat and SOCS box protein family as an ATRA target gene. ASBs negatively regulate cytokine signaling by mediating the ubiquitination of a broad range of target proteins. Furthermore, ASB proteins may regulate hematopoietic differentiation as suggested by ASB2 mediated degradation of the mixed lineage leukemia (MLL) protein and HOX gene down-regulation. MLL fusion proteins, present in approximately 10% of AML cases, have been found to be resistant to ASB2 mediated degradation which may contribute to leukemogenesis. Given that ASB2 plays a role in modulating differentiation and is ATRA induced it is feasible that other family members behave similarly.

To understand the role of ASB proteins in hematopoietic differentiation we examined the expression profiles of ASB genes during normal myeloid differentiation using cells isolated at each stage of differentiation towards the granulocyte lineage. Alongside this, both ATRA-induced and DMSO-induced differentiation in AML cell lines (HL-60 and NB4) was assessed. In addition, ASB expression was analysed in the MILE study, which encompasses expression profiling of 16 leukemic subtypes, an independent cohort of clinically phenotyped patient samples, and normal bone marrow. Candidate ASBs who exhibited significantly different expression were identified by bioinformatic analysis of the MILE study and patient samples. One candidate (ASB3) was selected for silencing by shRNA and the resultant phenotype was assessed by proliferation ability, apoptosis potential and differentiation capacity. Furthermore, ASB3 silencing was analysed by Affymetrix microarrays to identify potential targets of ASB3.

ASBs were found to be differentially expressed throughout normal hematopoietic differentiation with the highest expression in immature cells. In particular, ASB3 exhibited a two-fold decrease (p<0.001) in expression as myeloid cells matured. Several ASBs showed differential expression between AML patient samples and normal bone marrow with ASB3 exhibiting aberrant over-expression in AML (p≤0.005). Furthermore, Kaplan Meier plots using the MILE dataset indicated AML patients with FAB AML M0-M2 possess lower ASB3 expression which was associated with reduced overall survival and a poorer prognosis (p<0.05). ATRA or DMSO induced differentiation of HL-60 and NB4 cells altered ASB3 expression ten-fold (p<0.001) suggesting that the association of ASB3 with differentiation is not ATRA specific. Silencing of ASB3 expression (70% ±10%) in HL-60 and NB4 cells resulted in decreased proliferation and a 17% ±6% increase in apoptosis (p<0.01). In addition, reduced ASB3 expression markedly inhibited ATRA-induced differentiation of HL-60 and NB4 cells and ASB3 expression levels remained at a low level. Microarray analysis identified 78 differentially expressed genes between control and ASB3 silencing (p<0.05) including key genes specifically involved in apoptosis and hematopoietic differentiation e.g. PBX1, RUNX2 and CD44 were up-regulated during ASB3 silencing while ZNRF1, HOXA7 and CSRP2 were down-regulated when ASB3 was silenced.

As ASB3 is highly expressed in myeloid progenitors and loss of expression blocks ATRA-induced differentiation it is possible ASB3 may play a role in myeloid differentiation. Little is known about the function of ASB3 although it has been suggested that it is a negative regulator of tumor necrosis factor receptor II (TNF-R2). Our study, and others, would suggest that it may have a role in the regulation of transcription factors important for differentiation e.g. HOX genes, as with ASB2. Furthermore, poor prognosis AML exhibited lower ASB3 expression predominantly in the most immature AML subtypes which suggests that without ASB3 the AML cells are unable to differentiate. Additional studies are required to develop our understanding of the function of ASB3 in myeloid differentiation and its contribution to leukemogenesis.