Identification of Novel Splice Variants of Multiple Genes Using Genome-Wide Analysis of Alternative Splicing in Patients with Acute Myeloid Leukemia.

Abstract 1278

Poster Board I-300

Alternative splicing (AS) of pre-mRNA is a normal process for many genes that contributes to protein diversity. Splicing defects in genes associated with susceptibility and/or progression have been identified in a number of types of cancer including leukemias. Most studies of AS in patients with acute myeloid leukemia (AML) have been focused on individual genes. We performed a genome-wide analysis of AS in 27 patients with AML (12 PB, 10 BM cells and additional for 5 cases malignant blasts were enriched by sorting for CD34+ cells). In addition we have analyzed BM samples from 6 normal donors (ND). From 3 NDs CD34+ cells were sorted and 3 ND samples were unselected. Out of 27 patients, 23 samples were taken at diagnosis and the rest were obtained at relapse. In order to identify genes with differential AS between AML and NDs, patient and ND samples were hybridized to the Affymetrix Human Exon 10ST array. Results obtained from the exon array were analyzed using the XRAY software package (www.biotiquesystems.com). Of the 18726 genes with at least 4 “core” probe-sets (representing the most reliable genomic annotations) 7907 were found to be expressed above background in at least one sample and only these genes were analyzed for AS. After multiple testing corrections, a total of 2227 genes were identified as differentially spliced in one or more AML patient but not in NDs (p=0.01). Restricting results to genes that were differentially spliced in at least 35% of AML patients as compared to NDs, reduced the number of genes to 502 genes. More stringent analysis further reduced spliced gene number to 210 genes. Splice variants of these spliced genes were screened through a public Alternative Splicing and Transcript Diversity database (EMBL-EBI) and further studies focused on the genes with novel splice variants that have not been reported previously. For a selected genes exon array findings were validated using RT-PCR, qRT-PCR, and DNA cloning and sequencing on 18 of 27 patient and 6 ND BM samples, as well as on neutrophils and monocytes obtained from 4 ND peripheral blood samples. For further validation and functional studies several genes demonstrating AS were selected. For example, two different splice variants of NOTCH2 (NOTCH2-Va and –Vb) were detected in 16 (88%) of 18 AML samples. Both novel variants of NOTCH2 were cloned and sequenced from 10 AML patients. Sequencing analysis demonstrated that NOTCH2Va transcripts are result of one exon skipping, while NOTCH2Vb transcripts have two exons lost, both involving exons encoding extracellular regions of the protein. These AS events did not change coding sequence of the NOCTH2 gene; however, at the donor-splice sites two amino acids were changed. Similarly, 3 novel splice variants of FLT3 were found in 16 of 18 different patients. None of these variants were observed in ND BM or PB samples. FLT3 splice variants were cloned and sequenced from 7 AML patients, these splice variants are results of deletion of one entire exon and deletion of variable segments of adjacent exons. Expression of these specific NOTCH2 and FLT3 splice variants was not detected in 5 leukemia, 3 breast, 9 myeloma, and 3 lymphoma cell lines. Functional consequences of these splice variants as well as other novel splice variants identified by us through genome-wide AS analysis are in progress. Overall, our results from genome-wide AS analysis suggest that alternative splicing is a common event in AML, with some splice variants being detected in a significant number of patients. It will be of interest to evaluate alternative splicing in a large cohort of patients and define clinical correlations. Certain of the more common splice variants may generate new targets for the development of novel therapeutics.