Diagnostic microRNAs in myelodysplastic syndrome.

OBJECTIVE: The myelodysplastic syndromes (MDS) are aging-associated disorders characterized by ineffective maturation of hematopoietic elements, which are often diagnostically challenging. This study identifies microRNAs (miRNA) and miRNA targets that might represent diagnostic markers for MDS.
MATERIALS AND METHODS: This study utilized a total of 42 MDS samples and 45 controls. A discovery set of 20 frozen bone marrow mononuclear cell samples (10 MDS, 10 controls) was profiled on a custom Agilent miRNA microarray. Classifier miRNAs were validated in a separate set of 49 paraffin-embedded particle preparations by real-time polymerase chain reaction (24 MDS, 25 controls). Target prediction analysis was compared to a de novo transcriptional profile of MDS derived from the Microarray Innovations in Leukemia study. c-Myb and Sufu were further investigated by immunohistochemical stains on a set of 26 paraffin-embedded samples.
RESULTS: We identified 13 miRNAs of interest from the discovery set, 8 of which proved statistically significant on real-time polymerase chain reaction verification. These eight miRNAs were then examined in an independent real-time polymerase chain reaction validation set. Notably, hsa-miR-378, hsa-miR-632, and hsa-miR-636 demonstrated particularly high discrimination between MDS and normal controls. Target prediction identified potential targets of miRNA regulation that correspond to many of the genes that characterize MDS. Immunohistochemical staining performed on a third validation set confirmed that c-Myb and Sufu are differentially expressed in MDS.
CONCLUSIONS: Our data utilize both discovery and validation sets and two complementary platforms to identify miRNAs associated with MDS. We have analyzed predicted targets and identified c-Myb and Sufu as potential diagnostic markers of MDS.