Identification Of C-Terminal Truncated Splice Variants Of The Apoptosis-Stimulating Protein Of p53-2 (ASPP2) In Acute Leukemia Lacking The p53-Binding Site

Apoptosis seems to be a major means by which p53 suppresses tumorigenesis in hematopoietic cells. While loss-of-function mutations in p53 are frequently found in most neoplasms, non-complex karyotype acute leukemias are typically p53 wildtype and the p53-pathways are thought to be compromised through other mechanisms. ASPP2 is an independent haploinsufficient tumor suppressor which initiates induction of apoptosis after cellular damage in a p53 dependent manner (Kampa et al., PNAS 2009). We and others have found that attenuation/loss of ASPP2 is a common event in tumors. Here we report the identification of three novel splice variants of ASPP2, which are frequently found in acute leukemias: Tantalizingly, all variants lack the functionally relevant p53 binding sites.

The most prevalent isoform lacking exon 17 was termed ASPP2κ and detected in >30% of 80 investigated acute leukemia patients. Another variant termed ASPP2μ lacks exon 16/17 – and a third variant, ASPP2λ, lacks exon 17 and most of exon 18. Quantitative PCR on genomic DNA comparing exon 15 and exon 17 expression levels revealed that the variants derived from alternative splicing rather than mutation. All truncated isoforms were found to result in a reading frame-shift with a premature translation stop causing loss of most of the C-terminus harboring the p53-binding sites. Development of a highly sensitive isoform-specific qRT-PCR assay for ASPP2κ revealed low to none mRNA levels in physiologic mononuclear bone marrow cells – In contrast, high ASPP2κ mRNA levels were exclusively found in leukemia blasts. Analysis of the CD34+ fraction confirmed ASPP2κ-expression in leukemia stem/progenitor cells. In addition, analysis of samples derived from patients before and after induction chemotherapy revealed absence of ASPP2κ, resp. ASPP2μ, in patients achieving complete remission after hematopoietic reconstitution, again demonstrating leukemia-specificity of the truncated isoforms. We next generated isoform-specific antibodies for ASPP2κ directed to the hypothetical fusion site and confirmed their sequence-specificity by BlastN search. Subsequent co-immunoprecipitation experiments revealed genuine translation of the splice variants into protein isoforms. Comparative analysis of primary leukemia cells and healthy blood and bone marrow donors demonstrated that ASPP2κ protein is specifically expressed in leukemia. Preliminary functional analysis revealed that ASPP2κ is damage-inducible, suggesting a role of aberrant splicing of ASPP2 in response to cellular stress. Studies of the promoter region suggest alternative/aberrant promoter activation, which was not seen in physiologic cells. Transfection of ASPP2κ in murine pro-B Ba/F3 cells induced a more aggressive phenotype with mitotic failure, perturbed cellular proliferation and attenuated induction of apoptosis upon diverse cellular stressors. Expression of aberrant ASPP2 splice variants might be a underlying initiating events of tumorigenesis.

In conclusion, we identified three novel ASPP2 splice variants impairing proapoptotic signaling in native acute leukemia cells. Our data so far suggest aberrant ASPP2 signaling as an initiating event in oncogenesis, giving rise to genomic instability and mutations thereby triggering the actual leukemic clone. Further functional analyses are ongoing.