Caspase-9: A Candidate Susceptibility Factor for Murine Alkylator-Induced Leukemia.

Abstract 1273

Poster Board I-295

Therapy-related acute myeloid leukemia (tAML) is caused by exposure to chemotherapies and radiotherapies and has a poor prognosis. To better understand the genetic factors involved in secondary leukemogenesis, we studied alkylator-induced leukemias in murine models. We performed genome-wide mRNA profiling and found that differential expression of apoptosis-related genes was correlated with strain-dependent differences in tAML susceptibility. To identify genetic variants associated with these gene expression changes, we examined copy number variants (CNVs), differences in the DNA copy number of genomic regions greater than one kilobase in length, across 20 different inbred murine genomes. Our studies indicate that CNVs are a major contributor to natural variation among inbred laboratory mouse strains, and our array-based expression data show that the presence of certain CNVs correlates with differences in apoptotic gene expression. One CNV identified on murine chromosome 4 is correlated with altered Caspase-9 (Casp9) expression levels in murine hematopoietic stem/progenitor cells. Casp9 is a gene downstream from extrinsic and intrinsic death-inducing signals crucial for the initiation of cellular apoptosis, and we propose that it may be an important factor that influences tAML susceptibility. Sequencing the CNV region confirmed in two tAML susceptible strains (DBA/2J and PL/J) the presence of a 1,705 bp CNV loss in Casp9 intron 6. In these strains, Casp9 expression is undetectable in flow sorted kit+/lineage- (KL) hematopoietic stem/progenitor cells as measured by microarray profiling, and confirmed in independent samples by qRT-PCR using assays targeting Casp9 exons 2-3 or exons 8-9. Full-length Casp9 cDNA clones could be isolated from mRNA libraries prepared from KL cells, but 35% of the transcripts isolated from DBA/2J and PL/J mice were a novel isoform lacking exon 2 that results in a frameshift and an early stop codon in exon 4. This premature translation termination codon is predicted to trigger nonsense-mediated mRNA decay, leading to the degradation of the novel isoform and thus preventing its translation. This mechanism may account for the low Casp9 expression levels in DBA/2J and PL/J mice. In addition, Casp9 cDNA sequencing also identified an unusually high SNP density in exon 2 for the full-length isoforms in these strains. At least two of the six SNPs result in the formation of putative exonic splice enhancers (ESE), sequences that enhance the splicing of the exon in which they reside. The presence of these novel ESEs could explain the excision of exon 2 and the creation of the novel Casp9 isoform in DBA/2J and PL/J mice. We hypothesize that cells with relatively low Casp9 expression would be more resistant to alkylator-induced apoptosis. Cells that fail to undergo apoptosis after genotoxic stress may be more likely to accumulate mutations and initiate leukemias. Preliminary data from flow cytometric apoptosis assays (by Annexin V staining) in flow sorted KL bone marrow cells after treatment with ENU, an alkylating agent, suggest that PL/J cells may be more resistant to ENU-induced apoptosis, compared to C57BL/6J cells with normal Casp9 expression (5.3±1.4% vs. 9.7±1.8% post-ENU AnnexinV+ cells, respectively). These results suggest that differences in Casp9 expression levels may play a role in influencing individual susceptibility to tAML, and that inherited genetic factors may explain the observed expression differences. Ultimately, our understanding of the role that genetics plays in determining susceptibility to secondary leukemias may allow us to define a process by which individuals who are more susceptible can be successfully identified and screened from potential treatments that are known to induce these cancers.