Abstract
MDS comprise a spectrum of senescence-dependent stem cell malignancies with cytogenetic abnormalities characterized by segmental or numerical chromosome deletion or duplication. Accumulation of genetic events with age has been implicated in MDS pathogenesis. MiRNAs are short non-coding double-stranded RNAs that regulate the posttranscriptional gene expression. Expression profiling has shown that miRNA signatures are tumor specific and may have a role in leukemia pathogenesis. We hypothesized that altered regulation of miRNA expression in hematopoietic progenitors plays a critical role in the physiology of hematopoietic senescence and in MDS pathogenesis. To this end, we compared the miRNA microarray profile of MDS BM-MNC with normal controls, and validated differences by Northern blot and real-time RT-PCR. Twenty MDS patients (10 IPSS Low or Int-1 risk [LR], 10 Int-2 and high risk [HR]) and 20 normal controls (NC) (10 younger, ages 18–35 years [YC] and 10 older >70 years [OC]) were investigated. Total RNA was extracted and 5 mg were labeled with biotin and used for microarray analysis on a miRNA chip (OSUCC 3rd generation). Arrays were normalized with Cyclic Loess (Bioconductor). Statistical comparisons were made using the Significance Analysis of Microarray (SAM) Excel plugin. MiRNA predictors and 10-fold cross-validation were computed with use of Prediction Analysis of Microarray (PAM). Fifty-six miRNAs were up-regulated and 8 down-regulated in HR MDS vs. LR MDS with a >2 fold change and q value <0.01 (false positive rate) indicating that miRNA over-expression is a common event in MDS and possibly implicated in disease progression. Results of PAM (predictor analysis of microarray) revealed that HR MDS and LR MDS could be segregated by miRNA profile. A signature consisting of seven up-regulated miRNAs correctly discriminated between HR and LR groups with a misclassification error of 0 (10-fold CV). All seven miRNAs in this signature belong to the miR-181 family, which play key roles in hematopoiesis. Furthermore, our studies show divergent miRNA profiles between HR MDS and normal controls (NC), with 117 up-regulated miRNAs and 12 downregulated in HR MDS (>2-fold change, q<0.01). A miRNA signature consisting of 15 up-regulated miRNAs correctly discriminates these two groups with a misclassification error of 0 (threshold of 4.6). LR MDS and NC samples had greater similarity and a larger number of miRNAs was needed for discrimination with misclassification error of 11.4%. Similar data were obtained in comparisons of OC with YC; however, several members of let-7 miRNA family implicated in the down-regulation of RAS were consistently down-regulated in OC. We conclude that miRNA up-regulation is a frequent event in MDS. Up-regulation of miR-181 is sufficient to discriminate between HR and LR disease. Down-regulation of let-7 miRNA with age suggests a possible role in age-related predisposition to myeloid malignancy.
Disclosure: No relevant conflicts of interest to declare.
Author notes
Corresponding author