MicroRNA-155 regulates casein kinase 1 gamma 2: a potential pathogenetic role in chronic lymphocytic leukemia

Chronic lymphocytic leukemia (CLL) is the most common leukemia of adults in Western countries. There are approximately 19 000 new cases diagnosed in the United States in 2016 (https://seer. cancer.gov/statfacts/html/clyl.html). MicroRNAs (miRs) have been implicated as one of the key contributors in the pathogenesis of CLL. One of these miRNAs is miR-155, which is probably the bestcharacterized miRNA involved in B-cell maturation and function. MiR-155 has a critical role in normal immune function, including innate response, regulation of the germinal center response and formation of class-switched plasma cells and negative regulation of activation-induced cytidine deaminase. However, miR-155 is also an oncogenic miRNA overexpression of miR-155 in mice or its precursor BIC in chickens led to the development of lymphomas. Furthermore, abnormal expression of miR-155 has been observed in a number of lymphoid malignancies, including diffuse large B-cell lymphoma, classical Hodgkin lymphoma, primary mediastinal B-cell lymphoma and CLL. In various studies, miR-155 has been found consistently overexpressed in CLL cells compared with normal B cells. Although there does not appear to be definitive correlations between miR-155 expression levels and individual CLL prognostic factors, high pretreatment levels of miR-155 in CLL cells or in plasma have been demonstrated to be associated with shorter need-for-treatment interval and failure to achieve complete response, respectively. The association of adverse clinical outcome in CLL with high miR-155 levels appears to be linked to the capacity of miR-155 to enhance the sensitivity of CLL cells to B-cell receptor ligation. Thus it is likely that miR-155 has important roles in the pathobiology of CLL. Identification of targets for miR-155 will facilitate the understanding of how its deregulation contributes to the pathogenesis of CLL. To identify physiologically relevant targets of miR-155 in CLL, we measured miR-155 levels of purified CLL B cells from the peripheral blood of 38 patients using quantitative reverse transcriptase-PCR, as well as 5 sets of naive B cells and 6 sets of memory B cells isolated from hyperplastic tonsils. The results were ranked according to miR-155 levels and IGVH status (Supplementary Figure S1). MiR-155 levels in CLL with high miR-155 expression (11 in total) are about fivefold that of CLL with low miR-155 expression (11 in total). The six top-ranked and five bottom-ranked CLLs in each of the IGVH-unmutated and -mutated subgroups were selected based on miR-155 expression for global cDNA expression microarray analysis (supplementary Table S1) using Affymetrix HG-U133 Plus 2.0 microarrays (Cat no. 900466, Affymetrix Inc., Santa Clara, CA, USA) containing 454 000 probe sets covering 420 000 characterized human genes. A comparison between the expression profiles of CLL with high miR-155 vs low miR-155 using significance analysis of microarray analysis (5% false discovery rate) revealed 10 probe sets derived from 8 genes that are downregulated (Supplementary Table S2). Three of these genes: CSNK1G2, ZNF652, and KLF3, harbor evolutionarily conserved miR-155 binding sites according to prediction by TargetScan (v 7.1, June 2016). To further validate CSNK1G2 as a miR-155 target, we generated luciferase reporter plasmids with human CSNK1G2 3′-untranslated region containing either wildtype or mutated miR-155-binding site to perform luciferase reporter assays. The results confirmed that miR-155 could repress the reporter gene expression in 293 T cells and that this repressive effect could be relieved by mutation in the seed region of the miR-155-binding site (Supplementary Figure S2). Within the original cohort, CLLs with low miR-155 had significantly higher levels (~1.6-fold) of CSNK1G2 mRNA compared with CLLs with high miR-155 (Figure 1a), which validates the results of microarray analysis and is consistent with a negative regulatory effect of miR-155 on CSNK1G2 mRNA levels. For additional validation, a cohort of 43 additional CLL samples were tested Supplementary Table S3. We did not detect any significant correlation between the miR-155 levels and CSNK1G2 mRNA expression levels for the new cohort when the entire cohort was analyzed (Figure 1b). When CSNK1G2 mRNA expression levels were compared between the subgroup of six patients with the highest miR-155 levels and the subgroup of six patients with the lowest miR-155 levels, the high miR-155 subgroup showed a trend of lower CSNK1G2 levels (Figure 1c). The difference was not statistically significant (P= 0.17). However, when the cohort was subdivided into IGVH-unmutated and -mutated groups, we found a moderate correlation (R = 0.42) between miR-155 and CSNK1G2 mRNA levels for the IGVH-mutated group but not for the IGVHunmutated group, and the CSNK1G2 mRNA expression was significantly higher for the low miR-155 expression in the IGVHmutated subgroup (P= 0.04, Figures 1d–g). We were unable to determine whether similar negative correlation can be seen between miR-155 and CSNK1G2 protein, as western blotting of protein extracts isolated form primary CLL cells using commercially available antibodies has been unsuccessful in detecting specific bands despite repeated attempts. We also enforced miR-155 expression in CLL cells to determine its effect on CSNK1G2 expression. MiR-155 mimic (Cat. no. 472490001, Exiqon, Vedbæk, Denmark) was introduced into fresh purified CLL B cells from the peripheral blood of patients with IGVH mutation and relatively low miR-155 levels. A 1.6-fold Increase in miR-155 resulted in reduction of CSNK1G2 and SOCS1 mRNA (as positive control) levels by 17% and 31%, respectively, compared with CLL cells transfected with negative control microRNA (miRCURY LNA microRNA mimic Negative Control 4, Exiqon, Cat. no. 479903-001) (Figures 2a and b). These findings provided direct experimental evidence of a role of miR-155 in inhibiting CSNK1G2 expression, particularly in IGVH-mutated CLL cells. Our findings suggest that CSNK1G2 is a physiologically relevant miR-155 target in CLL cells. Previously, CSNK1G2 has been predicted to be a miR-155 target in diffuse large B-cell lymphoma based on transcriptome profiling and also by Argonaute-2 RNA immunoprecipitation followed by next-generation sequencing in 293 T cells with enforced miR-155 expression. CSNK1G2 is an isoform of casein kinase I, an evolutionarily conserved family of serine/threonine kinases implicated in multiple cellular functions, such as differentiation, proliferation and chromosomal segregation. One of the targets of CSNK1G2 is the collagen type IV alpha 3 binding protein (COL4A3BP/CERT), phosphorylation of which can lead to its inactivation and consequent downregulation of endoplasmic reticulum-to-Golgi transport of ceramide and Citation: Blood Cancer Journal (2017) 7, e606; doi:10.1038/bcj.2017.80