Discovery and Functional Implications of a Novel Mir-29b/Mir-29a Polymorphism in Acute Myeloid Leukemia.

MicroRNAs (miRNAs) are associated with cytogenetics and molecular subtypes of acute myelogeneous leukemia (AML). We have previously shown that miR-29 expression is down-regulated in cytogenetically normal AML (CN-AML) with wild type NPM1 and in t(11q23) primary blasts. Functionally, restoration of miR-29b in AML cell lines and primary samples induces apoptosis and dramatically reduces tumorigenicity in a xenograft leukemia model (Garzon et al, EHA 2008). Despite, these studies supporting a tumor suppressor role of miR-29b in AML, little is known about how miR-29 expression is down-regulated in AML. Since, miRNAs could be target for mutation, here we propose to screen mutations that could affect miR-29 expression and function.

The miR-29 family is comprised of three isoforms arranged in 2 clusters; miR-29b-1 and -a located on chromosome 7q32 and miR-29b-2 and -c located on chromosome 1q23. To screen for mutations, the entire genomic region from blasts of 100 primary AML samples corresponding to the miR-29b-1 and -a cluster, including 200 bp at the 5' and 3' ends was amplified and sequenced using the Applied Biosystems DNA sequencing system. When a deviation from the normal sequence was found, a panel of DNA from the blood of 50 control subjects was screened to identify polymorphisms. Patient characteristics include: CN-AML: 62 (FLT3-ITD 10/43, NPM1 mutated (34/62); inv16: 10; t(8;21): 2; t11q23: 2; complex karyotype (CK): 10; monosomy 7(-7): 7; other cytogenetics: 7. miR-29 expression we performed by miRNA Taqman assays as per manufacturer recommendations.

We identified a thymidine (T) base deletion within the miR-29b-1 and -a cluster precursor miRNAs (at -264 bp from the 5' position of miR-29a in chromosome 7q32) in 17/100 patients. The (T) base deletion was observed in 4/10 inv16 and 6/62 CN-AML patients, while the other 7 cases were distributed among CK (2/10), -7 (3/7), 11q23 (1/2) and other cytogenetics (1/7). In 2 patients, normal cells from the buccal mucosa were heterozygous for this abnormality. The frequency of this germline abnormality in the normal population was 16% (8/50 cases). Next, we investigated the miR-29b and -a expression in 35/100 primary AML samples, where RNA was also available. Although miR-29a and -b levels were not significantly different in polymorphism (n=10) versus wild type (WT) (n=35) samples, we observed that the miR-29a/miR-29b ratios were significantly lower in the polymorphism than WT (43.5 vs. 24.9 respectively, P-value=0.007, t-test). To characterize further this abnormality, we cloned the polymorphism harboring miR-29b and -a cluster from 1 patient into p-Retro Super plasmid and transfected into K562 cells (lack miR-29 expression) along with WT and empty vector constructs. Northern blotting after 24-48 hours revealed an accumulation of the precursor miR-29a while the mature miR-29a level was decreased by 2 fold. The level of the mature miR-29b was unchanged. To asses whether this polymorphism affects miR-29 targeting efficiency, we co-transfected a reporter luciferase construct containing the 3' untranslated region of the known miR-29 target, MCL-1 with the WT, empty vector and polymorphism harboring miR-29b and -a cluster and performed luciferase assays. Interestingly, relative normalized luciferase activities were less inhibited with the polymorphism cluster than the WT construct (relative reduction WT:80%, Polym:63%. Likewise, MCL-1 protein down-regulation elicited by the ectopic WT cluster overexpression was stronger than the one observed for the polymorphism harboring cluster (b-actin/MCL-1 rations 0.35 vs 0.48, respectively).

Our results identify a novel germline polymorphism within the miR-29b and —a cluster in AML. The frequency of this polymorphism in AML is similar to the normal population. However, the increased frequency observed in the inv16 subgroup (4/10) warrant further confirmation in a large cohort of patients. Functionally, this polymorphism affects the expression ratio of miR-29b and —a by dampening the processing of miR-29a and impacts negatively in the ability of this cluster to target the oncogene MCL-1.

No relevant conflicts of interest to declare.