Transcription profiling of mouse lymphoid neoplasia shows changes in processing of 3 prime-UTR characterize clinically distinct tumor types

We used a novel probe-level microarray analysis, revealing connections between mRNA processing and lymphoid neoplasia, in a mouse leukemia model. Characteristic differences in mRNA processing, primarily in the 3’-untranslated region, distinguished histologically similar tumor subtypes with different survival characteristics. Gene sets with specific processing in each tumor subtype defined signatures useful for tumor subclassification, as demonstrated by internal cross-validation with up to 80% discrimination accuracy. A combination of mRNA expression and sequence analysis suggested that differences in isoform abundance likely arose from both alternative polyadenylation and differential degradation. Experiment Overall Design: 5 types of samples were analyzed with multiple biological replicates. Progenitor B-cells, mature B-cells were the control. LPC, APC, APN were the tumor cells. LPC (Lig4/P53 deficient cells developed pro-B lymphoma with c-myc amplification), APC (Artemis/P53 deficient cells developed pro-B lymphoma with c-myc amplification), APN (Artemis/P53 deficient cells developed pro-B lymphoma with N-myc amplification)