Tcl1 Expression in Chronic Lymphocytic Leukemia Is Regulated by miR-29 and miR-181 Free

B-cell chronic lymphocytic leukemia (B-CLL) is the most common human leukemia in the world, accounting for ∼10,000 new cases each year in the United States (1). The TCL1 (T-cell leukemia/lymphoma 1) oncogene was discovered as a target of frequent chromosomal rearrangements at 14q31.2 in mature T-cell leukemias (2). Previously, we reported that transgenic mice expressing TCL1 in B cells develop B-CLL (3). These results suggested that deregulation of TCL1 may be a causal event in the pathogenesis of B-CLL. We and others also have shown that Tcl1 is a coactivator of the Akt oncoprotein, a critical molecule in the transduction of antiapoptotic signals in B and T cells (4, 5). A recent report suggested that high Tcl1 expression in human B-CLL correlates with unmutated VH status and ZAP70 positivity, suggesting that Tcl1-driven B-CLL is an aggressive form of B-CLL (6). Another study showed that the TCL1 transgenic model replicates the immunoglobulin V region rearrangements characteristic of the aggressive, treatment-resistant form of human B-CLL (7). One of the most significant genetic factors associated with poor prognosis in human B-CLL is the chromosome 11q deletion (8). Interestingly, B-CLL samples showing 11q deletion also display higher Tcl1 expression levels (6).

MicroRNAs are a large family of highly conserved noncoding genes thought to be involved in temporal and tissue specific gene regulation (9). We recently showed that microRNA expression profiles can be used to distinguish normal B cells from malignant B-CLL cells and that microRNA signatures are associated with prognosis and progression of chronic lymphocytic leukemia (10, 11). To determine whether Tcl1 expression is regulated by microRNAs in B-CLL, we studied microRNA expression patterns and Tcl1 protein expression in 80 B-CLL samples of three types of B-CLL: indolent B-CLL, aggressive B-CLL with normal chromosome 11, and aggressive B-CLL showing 11q deletion. We choose these three types of B-CLL because a recent study suggested a differential expression of Tcl1 in these three groups (6).

CLL samples and microRNA microchip experiments. Eighty B-CLL samples were obtained with informed consent from patients diagnosed with B-CLL from CLL Research Consortium institutions. Research was done with the approval of the Institutional Review Board of The Ohio State University. Briefly, blood was obtained from CLL patients, then lymphocytes were isolated through Ficoll/Hypaque gradient centrifugation (Amersham, Piscataway, NJ) and processed for RNA extraction using the standard Trizol method. Protein extraction was carried out as previously described (12). MicroRNA microchip experiments were done as previously described (11). Each microRNA microchip contained duplicate probes, corresponding to 326 human and 249 mouse microRNA genes. Statistical analysis was carried out as previously described (13). To identify statistically significant differentially expressed microRNA, class prediction analyses were done using BRB ArrayTools developed by Dr. Richard Simm and Amy Peng Lam.

DNA constructs, transfection, Western blotting, and luciferase assay. Full-length TCL1 cDNA including 5′ and 3′ untranslated region (UTR) cDNA was cloned into a pUSEamp vector (Upstate Biotechnology, Chicago, IL; used in Fig. 2B). MiR-29b and miR181b RNA duplexes were purchased from Ambion (Austin, TX). For miR-29 luciferase assays, a fragment of the 3′ UTR of TCL1 cDNA, including a region complimentary to miR-29 (Tcl1), was inserted using the XbaI site immediately downstream from the stop codon of luciferase into pGL3 vector (Promega, Madison, WI). For miR181 assays, full-length TCL1 cDNA was inserted into pGL3 vector in sense (Tcl1FL) or antisense (Tcl1FLAS) orientation. Transfections were carried out as previously described (14). Firefly and renilla luciferase activities were assayed with the dual luciferase assay system (Promega) and firefly luciferase activity was normalized to renilla luciferase activity. Cell lysate preparations and Western blot analyses were carried out using anti-Tcl1 monoclonal antibody (clone 27D6) as previously described (4). Each Western filter contained reference sample. Tcl1 protein expression was assessed using this sample as a reference. P values were two tailed and calculated by Fisher's exact test.

High expression of Tcl1 correlates with aggressive B-CLL phenotype. To evaluate Tcl1 and microRNA expression in B-CLL samples, we chose three groups of B-CLL: 23 samples of indolent B-CLL, 25 samples of aggressive B-CLL, and 32 samples of aggressive B-CLL showing 11q deletion. Detailed description of the samples can be found in Supplementary Table S1. MicroRNA microchip experiments revealed that three groups of B-CLL show significant characteristic differences in microRNA expression pattern (Table 1 and Supplementary Table S2). To determine Tcl1 protein expression in three groups of B-CLL, we carried out Western blot analysis using 27D6 Tcl1 monoclonal antibody. Results of these experiments are shown in Fig. 1A and B. Tcl1 expression was assessed as low, medium, high, and very high. Our experiments revealed low levels in 15 of 23 (65%) indolent B-CLLs, in 11 of 25 (44%) aggressive B-CLLs, and in 1 of 32 (3%) aggressive B-CLLs with 11q deletions, whereas high and very high Tcl1 expression was observed in 1 of 23 (4%) indolent B-CLLs, in 14 of 25 (56%) aggressive B-CLLs, and in 24 of 32 (75%) aggressive B-CLLs with 11q deletions (Fig. 1B). This finding suggests that Tcl1 overexpression correlates with aggressive B-CLL phenotype (P < 10⁻⁶) and 11q deletions (P = 10⁻⁴). Our results are consistent with the recently published study showing that high Tcl1 expression in human B-CLL correlates with unmutated VH status and ZAP70 positivity (6).

MiR-29 and miR-181 target Tcl1. To determine which microRNA(s) target TCL1, we used RNAhybrid software offered by Bielefeld University Bioinformatics Server and miRBase database (15). Among miR candidates targeting Tcl1, we found that miR-29b and miR-181b (Fig. 1C; several other sites with lower homology not shown) are also down-regulated in aggressive B-CLLs with 11q deletions (Table 1). The expression of these miRs was confirmed by real-time reverse transcription-PCR in a representative set of samples (Supplementary Fig. S1). Furthermore, it was previously shown that expression of members of miR-29 family could discriminate between CLL samples with good and bad prognosis (11). We thus proceeded to determine if these miRs indeed target Tcl1 expression using the TCL1 3′ UTR inserted downstream of luciferase open reading frame, as previously described (14). HEK293 cells were cotransfected with the miR-29b or scramble negative control, as indicated, and pGL3 construct containing a part of TCL1 cDNA, including a region homologous to miR-29 (Tcl1), or pGL3 vector alone as indicated. For miR-181 assays, full-length TCL1 cDNA was inserted into pGL3 vector in sense (Tcl1FL) or antisense (Tcl1FLAS) orientation. Figure 2A shows that Tcl1 mRNA expression is inhibited by miR-29 and miR-181. To confirm these findings, we cloned full-length TCL1 cDNA, including 5′ and 3′ UTRs, into cytomegalovirus mammalian expression vector and investigated whether miR-29b and miR181b affect Tcl1 protein expression levels. We cotransfected this construct with miR-29b, miR-181b, and pre-miR negative control (scramble) into 293 cells as indicated in Fig. 2B. These experiments revealed that coexpression of Tcl1 with miR-29 and miR-181 significantly decreased Tcl1 expression (Fig. 2B, lane 2 and 4 versus lanes 1 and 3). We therefore concluded that miR-29b and miR-181b target Tcl1 expression at mRNA and protein levels. Interestingly, we found an inverse correlation between miR-29b and miR-181b expression and Tcl1 protein expression in B-CLL samples (Fig. 3). For samples with highest miR-29b expression (top 20%), 10 of 12 had low or medium Tcl1 expression, whereas in samples with highest miR-181b expression (top 20%), 11 of 12 had low or medium Tcl1 expression. Likewise, for samples with high expression of both miR-29b and miR-181b, 4 of 4 had low or medium Tcl1 expression. In summary, for samples with high miR-29b and/or miR-181b expression, 17 of 20 showed low or medium Tcl1 expression (P = 0.04). In addition, none of the samples with high miR-29b and/or miR-181b expression showed high Tcl1 expression (P = 0.05). These results suggest that Tcl1 expression in B-CLL is, at least in part, regulated by miR-29 and miR-181.

In this report, we show that Tcl1 expression is regulated by miR-29 and miR-181 and this regulation is relevant to the three groups of B-CLL we studied. Although we observed a reverse correlation between Tcl1 protein expression and these two miRs, a significant proportion of B-CLL samples show low Tcl1 expression and low expression of miR-29 and miR-181 (Fig. 3). This suggests that, in these samples, Tcl1 expression is down-regulated transcriptionally or by other microRNAs. The fact that neither miR-29 nor miR-181 is located at 11q suggests that the region may contain an important regulator of the expression of these two miRs. Previously, a microRNA signature was published with 13 microRNAs that differentiate aggressive and indolent B-CLL (10, 11). Intriguingly, of the four down-regulated microRNAs in aggressive B-CLL, three are different isoforms of miR-29 (miR-29a-2, miR-29b-2, and miR-29c), strongly suggesting that miR-29 and TCL1 interactions play an important role in the pathogenesis of aggressive B-CLL. Interestingly, miR-181 is differentially expressed in B cells and TCL1 is mostly a B-cell–specific gene (16). This suggests that Tcl1 might be a target of miR-181 not only in B-CLL cells but also in normal B-lymphocytes. Additional studies are necessary to determine whether there is an inverse correlation between TCL1 and miR-181 expression at different stages of B-cell maturation. Because miR-29 and miR-181 are natural Tcl1 inhibitors, these miRs may be candidates for therapeutic agents in B-CLL-overexpressing Tcl1.

Grant support: Kimmel Cancer Research Foundation Award and CLL Global Research Foundation grant (G.A. Calin), CLL Global Research Foundation grant (Y. Pekarsky), and NIH grant PO1-CA81534 for the CLL Research Consortium (L. Rassenti, T. Kipps, and C.M. Croce).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.