MicroRNAs (miRNA) are small noncoding, single-stranded RNAs that inhibit gene expression at a posttranscriptional level, whose abnormal expression has been described in different tumors. The aim of our study was to identify miRNAs potentially implicated in chronic myeloid leukemia (CML). We detected an abnormal miRNA expression profile in mononuclear and CD34+ cells from patients with CML compared with healthy controls. Of 157 miRNAs tested, hsa-miR-10a, hsa-miR-150, and hsa-miR-151 were down-regulated, whereas hsa-miR-96 was up-regulated in CML cells. Down-regulation of hsa-miR-10a was not dependent on BCR-ABL1 activity and contributed to the increased cell growth of CML cells. We identified the upstream stimulatory factor 2 (USF2) as a potential target of hsa-miR-10a and showed that overexpression of USF2 also increases cell growth. The clinical relevance of these findings was shown in a group of 85 newly diagnosed patients with CML in which expression of hsa-miR-10a was down-regulated in 71% of the patients, whereas expression of USF2 was up-regulated in 60% of the CML patients, with overexpression of USF2 being significantly associated with decreased expression of hsa-miR-10a (P = 0.004). Our results indicate that down-regulation of hsa-miR-10a may increase USF2 and contribute to the increase in cell proliferation of CML implicating a miRNA in the abnormal behavior of CML. (Mol Cancer Res 2008;6(12):1830–40)
MicroRNAs (miRNAs) are noncoding, single-stranded RNAs of 21 to 25 nucleotides recently implicated in the regulation of several biological processes, such as cell cycle, apoptosis, differentiation, and development or tumorigenesis (1, 2). miRNAs inhibit the expression of specific genes either by decreasing mRNA stability or by blocking mRNA translation (2-5). In silico studies have estimated and annotated more than 450 human miRNAs86, 7), many of them evolutionary conserved and localized in cancer-associated fragile genome regions (8). The fact that each miRNA may regulate multiple targets implies that a third of the protein-coding genes in humans may be regulated by miRNAs (9).
The first evidence for the involvement of miRNAs in human cancer was described in chronic lymphocytic leukemia in which deletion of a chromosomal region (13q14) was associated with down-regulation of miR-15a in the majority of chronic lymphocytic leukemia patients (68%; ref. 10). It was later shown that miR-15a and miR-16-1 negatively regulate expression of the BCL2 protein at the posttranscriptional level, inducing apoptosis in leukemic cells (11). Abnormal expression of miRNAs has currently been described in numerous solid tumors, such as lung cancer (12, 13), breast cancer (14), glioblastoma (15), hepatocellular carcinoma (16), endocrine pancreatic tumors (17), papillary thyroid carcinoma (18), testicular germ cell tumors (19), and colorectal cancer (20, 21), but also in hematologic malignancies, such as chronic lymphocytic leukemia (22), lymphomas (23), acute promyelocytic leukemia (24), and more recently acute lymphoblastic leukemia (25).
Chronic myeloid leukemia (CML) represents a remarkably useful model to study new mechanisms of tumorigenesis due to the fact that the hallmark of the disease is the presence of the Philadelphia chromosome that is present in >95% of the patients and is responsible for the disease (26-30). Information about the implication of miRNAs in the pathogenesis of CML is limited to the description of abnormal expression of certain miRNAs (miR-15a, miR-16, miR-142, miR-155, miR-181, miR-221, and let-7a) in the CML cell line K562 (31, 32) and up-regulation of the polycistronic miR-17-92 cluster in CML patients (33). Based on the analysis of the miRNA profile expression in patients with newly diagnosed CML, we have determined that hsa-miR-10a, hsa-miR-150, and hsa-miR-151 are down-regulated, whereas hsa-miR-96 is up-regulated in CML cells. Down-regulation of hsa-miR-150 and hsa-miR-151 is mediated by BCR-ABL1 expression, whereas down-regulation of hsa-miR-10a is BCR-ABL1 independent. In this study, we show that down-regulation of hsa-miR-10a results in increased expression of the upstream stimulatory factor 2 transcription factor (USF2), which contributes to the increased cell growth of CML cells.
A miRNA Expression Pattern Discriminates CML Patients from Healthy Donor Samples
We analyzed the expression profile of 157 miRNAs by quantitative real-time PCR (Q-RT-PCR) using Taqman methodology (PE Applied Biosystems) in mononuclear cells (MNC) and CD34+ cells from patients with CML (n = 6) at diagnoses and control healthy donors (n = 6). For comparison of miRNA expression, data on every sample were normalized using the expression of the housekeeping genes GUSB or U6 as described (34, 35). Relative quantification and supervised analysis comparing bone marrow (BM) MNCs of CML patients identified 53 differentially expressed miRNAs, 23 of them up-regulated and 30 down-regulated in CML patients with respect to healthy donors. Cluster unsupervised analysis based on differentially expressed miRNAs generated a tree that clearly distinguished between both groups (Fig. 1A). To assess a more homogeneous population of cells and because CML is a clonal stem cell disease, we did the same analysis using CD34+ cells from the same samples. Eleven miRNAs were differentially expressed, 10 of them down-regulated and 1 up-regulated in CML patients. Cluster unsupervised analysis based on differentially expressed miRNAs generated a similar tree distinguishing CML samples from normal donors (Fig. 1B).
When a further statistical analysis using Statistical Package for the Social Sciences was done comparing CML and healthy donor MNCs, only 34 of the 53 differentially expressed miRNAs maintained their statistically significant differences (P < 0.05; Supplementary Table S1). The same analysis done with CD34+ cells from patients and donors confirmed the differential expression of the 11 miRNAs (Supplementary Table S2). The results of the analysis were the same whether we used GUSB or U6 genes for normalization. We also did a comparison between the miRNA expression profile of CML MNC and CD34+ cells and healthy donor MNC and CD34+ cells, which showed 61 and 45 miRNAs to be differentially expressed in CML patients and normal donors, respectively, between MNC and CD34+ cells (Supplementary Fig. S1).
The analysis of those miRNAs differentially expressed both in BM MNC and CD34+ cells of CML patients in comparison with healthy donors showed that six miRNAs were differentially expressed in both comparisons, five of them down-regulated (hsa-miR-10a, hsa-miR-125a, hsa-miR-150, hsa-miR-151, and hsa-miR-199s) and one up-regulated (hsa-miR-96) in CML patients with respect to healthy donor samples, indicating a higher likelihood of these miRNAs to be involved in CML pathogenesis. To confirm the results, we analyzed the expression of five of these six miRNAs (hsa-miR-10a, hsa-miR-96, hsa-miR-125a, hsa-miR-150, and hsa-miR-151) in MNC and CD34+ cells from another 16 newly diagnosed CML patients and another 6 healthy donors by Q-RT-PCR using individual Taqman miRNA assays. hsa-miR-199s miRNA expression was not analyzed because this miRNA has been eliminated from the miRNA annotation list. The analysis confirmed the statistically significant differences in the expression of four of five miRNAs (hsa-miR-10a, hsa-miR-96, hsa-miR-150, and hsa-miR-151) between CML and healthy donors (Table 1), suggesting that these miRNAs could be directly implicated in the pathogenesis of CML.
|miRNA .||miRNA Sequence .||Genome Location .||P value CD34+ CML/HD .||P value MNCs CML/HD .|
|miRNA .||miRNA Sequence .||Genome Location .||P value CD34+ CML/HD .||P value MNCs CML/HD .|
Down-Regulation of hsa-miR-10a in CML Is Not Mediated by BCR-ABL1 Activity
As CML is characterized by the presence of the fusion protein BCR-ABL1, which shows an increased tyrosine kinase activity, we reasoned that abnormal expression of hsa-miR-10a, hsa-miR-150, hsa-miR-151, and hsa-miR-96 observed in CML patients could be dependent on BCR-ABL1 or its kinase activity. To test this hypothesis, we analyzed the expression of these miRNAs by Q-RT-PCR in the Mo7e and Mo7e-p210 cell line before and after treatment with imatinib, a specific inhibitor of BCR-ABL1 kinase activity, for 24 hours (36). Whereas Mo7e-p210 cells showed a reduced expression of hsa-miR-150 and hsa-miR-151 in comparison with Mo7e cells, treatment with imatinib induced an up-regulation of expression of both miRNAs in Mo7e-p210 cells. These results suggest that BCR-ABL1 kinase activity induces down-regulation of hsa-miR-150 and hsa-miR-151 in CML (Supplementary Fig. S2A and B). No expression of hsa-miR-96 was detected in any of the cell lines before or after treatment with imatinib. Whether a genetic abnormality is responsible for the lack of hsa-miR-96 expression in the parental Mo7e cell line is currently unknown.
Unlike what we observed in patients with CML in which expression of hsa-miR-10a was significantly down-regulated in 71% of the patients, the expression of hsa-miR-10a was increased in Mo7e-p210 and treatment with imatinib had no statistically significant effect on miRNA expression (Supplementary Fig. S2C). Expression of hsa-miR-10a was also increased in TCC-S cells, whereas BV173 and KU812 cell lines showed a reduced hsa-miR-10a expression in comparison with Mo7e (Fig. 2A). Expression of hsa-miR-10a in CML-derived cell lines was confirmed by Primer extension analysis (Fig. 2B). Quantification of the primer extension gave a similar result than the normalization of hsa-miR-10a expression obtained by Q-RT-PCR, ensuring the reliability of both techniques (data not shown). Although hsa-miR-10a is down-regulated in CML cells, which express BCR-ABL1, expression of hsa-miR-10a in these cells is unlikely to depend on BCR-ABL1 activity based on the following facts: (a) Mo7e-p210 does not show a down-regulation of hsa-miR-10a in comparison with Mo7e, (b) treatment with imatinib does not significantly affect the expression of hsa-miR-10a, and (c) CML-derived cell lines show different hsa-miR-10a levels.
USF2 Is a Potential Target of hsa-miR-10a
As biological significance of miRNA deregulation relies on their gene targets, we analyzed the predicted targets of hsa-miR-10a. The analysis was done using four algorithms, mirBase, miRanda, TargetScan, and PicTar, commonly used to predict human miRNA gene targets. Because any of the four approaches generate an unpredictable number of false positives, results were intersected to identify the genes commonly predicted by at least three of the algorithms. Results are shown in Supplementary Table S3.
We next analyzed whether any of the predicted targets of hsa-miR-10a were up-regulated in CML CD34+ cells. Thus, we compared the list of putative targets with the list of up-regulated genes found in gene array studies done comparing the gene expression profile of CD34+ cells from CML patients at diagnoses and healthy donor CD34+ cells (37). We found that the expression of transcription factor USF2 showed an inverse correlation with the expression of hsa-miR-10a being up-regulated in CML CD34+ cells (Supplementary Table S4). As predicted by PicTar,9
To show whether hsa-miR-10a could regulate the expression of USF2, RKO cells were transfected with the pre-miR-10a or anti-miR-10a molecules and the expression of hsa-miR-10a, HOXA1, and USF2 mRNA was analyzed by Q-RT-PCR. This cell line was used for technical reasons, as the colon carcinoma–derived RKO cell line can be efficiently transfected unlike leukemia cell lines. Expression of HOXA1 mRNA, a known hsa-miR-10a target (38, 39), was used as control. Transfection of RKO cells with Silencer FAM-labeled negative control showed a cell viability of 93% and a transfection efficiency of 88% by fluorescence-activated cell sorting (Supplementary Fig. S3). Transfection of RKO with anti-miR-10a induced a significant reduction of hsa-miR-10 expression level and an increase in the mRNA levels of HOXA1 and USF2 in comparison with cells transfected with the anti-miR scrambled control or negative control (Supplementary Fig. S4). In contrast, a significant increase of hsa-miR-10 expression level and a decrease of mRNA levels of HOXA1 and USF2 were found in RKO cells transfected with pre-miR-10 (Supplementary Fig. S4).
To further show the role of hsa-miR-10 in CML, nucleofection of the CML-derived KU812 cell line (viability was 78% and transfection efficiency was 96% by fluorescence-activated cell sorting using BLOCK-iT Fluorescent Oligo; Supplementary Fig. S5) was done. KU812 cells were chosen as they show low levels of hsa-miR-10, similar to what was observed in CML cells from patients (Fig. 2A). Transfection with anti-miR-10a induced a decrease in the expression of hsa-miR-10, which was smaller than the one observed in RKO cells, but a higher increase in the levels of HOXA1 and USF2 mRNA. Similarly, KU812 cells nucleofected with pre-miR-10 showed an increase in expression of hsa-miR-10a and a decrease in HOXA1 mRNA and USF2 mRNA and protein levels (Fig. 3A and B). These data indicate that hsa-miR-10a regulates the expression of USF2 (at least) through decrease in mRNA stability. In addition, we analyzed the expression of other genes presumably not regulated by hsa-miR-10a, such as GAPDH, RPS18, and CCND1, and found them not to be regulated by this miRNA (Supplementary Fig. S6).
To evaluate whether USF2 is a potential target of hsa-miR-10a, we did a Renilla luciferase reporter assay with a vector containing the putative USF2 3′-UTR target site downstream of a Renilla reporter gene (Renilla-USF2 3′-UTR vector). Predicted base pairing between hsa-miR-10a and a putative target site in the 3′-UTR of USF2 shows perfect complementarity between nucleotides 2 to 8 of hsa-miR-10a and the target, resulting in a seed of seven nucleotides (Fig. 4A). Renilla-USF2 3′-UTR vector, control vector containing firefly luciferase, and pre-miR-10a or negative precursor-miR control were transfected into RKO cell line. Firefly luciferase activity was evaluated to normalize for transfection efficiency. Renilla luciferase activity of RKO cells transfected with pre-miR-10a was significantly lower (P = 0.001) than RKO cells transfected with negative precursor-miR control (Fig. 4B). These data indicate that USF2 is a potential target of hsa-miR-10a.
USF2 Is Up-Regulated in CML Patients
To confirm our results in a more clinically relevant model, we analyzed the expression of USF2 in BM MNCs from a group of patients with newly diagnosed CML (n = 85) and in healthy donors (n = 21). Normalized ratios for USF2 expression determined in BM specimens, from healthy individuals (NUSF2 ratios) fell between 82% and 100% (mean NUSF2: 94 ± 7%). However, CML patient mean NUSF2 was 225 ± 205% (range, 89-1,057%), significantly higher than ratios found in healthy donors (P = 0.01). Moreover, a NUSF2 value above 125% (determined as the mean from healthy donors plus 3 SD) was chosen to define overexpression of USF2 in CML samples. Using this cutoff value, up-regulation of USF2 was found in 60% (51 of 85) of CML patients.
We also analyzed hsa-miR-10a expression in the same group of CML patients. Mean expression of hsa-miR-10a was 37% (range, 0.05-6,690%) being down-regulated in 60 of 85 (71%) CML patients (down-regulation was considered when expression of hsa-miR-10a was lower than the mean − 4 SD) compared with normal BM cells. CML patients with a decreased expression of hsa-miR-10a miRNA showed overexpression of USF2 mRNA, whereas CML patients with normal levels of hsa-miR-10a presented normal USF2 mRNA expression (hsa-miR-10a expression was 85 ± 159% in CML patients with up-regulation of USF2, and 515 ± 1,210% in patients with normal USF2 expression; P = 0.004). These results lead us to suggest that hsa-miR-10a down-regulation may cause increased levels of USF2 also in CML cells from patients.
Down-Regulation of hsa-miR-10a and Overexpression of USF2 Increase CML Cell Growth
We wanted to study whether down-regulation of hsa-miR-10a or up-regulation of USF2 could lead to increased cell growth of CML cells and, therefore, could be involved in the development or progression of CML. To evaluate this, hsa-miR-10a expression levels were altered in KU812 cells by nucleofection with pre-miR-10a or anti-miR-10 and cell growth was evaluated and compared with that of cells transfected with the pre-miR or anti-miR scrambled controls, respectively. Interestingly, an increase in hsa-miR-10a expression (Fig. 5A), which was associated with a decrease in USF2 protein expression (Fig. 5B), induced a decrease in cell growth (Fig. 5C), whereas a decrease in hsa-miR-10a expression resulted in an increase in cell growth (Fig. 5D).
Finally, to examine whether USF2 expression could participate in hsa-miR-10a–mediated regulation of cell growth, small interfering RNAs (siRNA) against USF2, or siRNA Silencer Negative Control-1, were nucleofected in KU812 cells. siRNAs against USF2 efficiently decrease USF2 expression as detected by Q-RT-PCR (Fig. 6A) or by Western blot analysis (Fig. 6B). Down-regulation of USF2 with siRNAs resulted in a decrease in cell growth compared with control cells (Fig. 6C and D). These results indicate that down-regulation of hsa-miR-10a and consequently overexpression of USF2 participate in the abnormal growth of CML cells.
Treatment of CML has notably been improved by the development of imatinib mesylate, a tyrosine kinase inhibitor that blocks the kinase activity of BCR-ABL1, defining this drug as the gold standard first-line therapy for CML patients (36, 40, 41). In contrast, it has been recently shown that imatinib mesylate does not inactivate all BCR-ABL1–activated signaling pathways that are essential for CML cell survival (42). This implies that persistent malignant progenitors can be a potential source of relapse in CML patients and that there is a need to improve our understanding of the biology of CML to provide new targets for therapy. The recent implication of miRNAs in the regulation of important biological processes altered in CML, such as cell cycle, apoptosis, and adhesion (2), establishes these small RNA molecules as potential players in CML pathogenesis.
The result of the current study identifies certain miRNAs (hsa-miR-10a, hsa-miR-96, hsa-miR-150, and hsa-miR-151) to be abnormally regulated in patients with CML and thus provides support to our hypothesis that miRNAs could be implicated in CML. While this article was in preparation, another study described the up-regulation of the polycistronic miR-17-92 cluster in CD34+ cells of CML patients (33). Unlike that study, we did not observe an up-regulation of miR-17-92 in CML CD34+ cells. The different results could be explained on the bases of the different strategies used for normalization. Whereas normalization of miRNA expression was done using the expression of hsa-miR-16 miRNA in the study from Venturini et al. (33), we used expression of GUSB and U6 genes for normalization as this strategy has been validated in recent studies (34). It is important to point out that the analysis of differentially expressed miRNAs in CD34+ cells of CML patients indicates that hsa-miR-16 is down-regulated, thus making this miRNA inappropriate for normalization of data. Furthermore, hsa-miR-16 negatively regulates the antiapoptotic protein BCL2 (11) that has also been implicated in CML (43). The use of a certain miRNA for normalization of the data based on the assumption that expression of the specific miRNAs is constant regardless of the tissue or the disease may not be a reliable way of normalizing the results (10, 12, 21, 44).
As we and others have previously shown, the presence of the p210 BCR-ABL1 fusion oncoprotein regulates directly or indirectly the mRNA and/or the protein expression of several genes that participate in the deregulation of the cell cycle, apoptosis, and adhesion observed in CML cells (45, 46). We decided to examine whether the abnormal expression of hsa-miR-10a, hsa-miR-96, hsa-miR-150, and hsa-miR-151 could be dependent of BCR-ABL1 and its kinase activity. The result showed that hsa-miR-150 and hsa-miR-151 decreased as a result of functional expression of BCR-ABL1. Further studies are required to address the molecular mechanism of BCR-ABL1–mediated down-regulation and whether down-regulation of hsa-miR-150 and hsa-miR-151 is essential for BCR-ABL1 role in CML.
The expression of hsa-miR-10a was independent of BCR-ABL1 activity and this could in fact suggest that abnormal expression of hsa-miR-10a is associated with the development of the disease. Interestingly, in a recent study in hematopoietic CD34+ cells, hsa-miR-10 has been designated as one of the 33 miRNAs implicated in hematopoiesis (hematopoietic-expressed miRNAs; ref. 47). According to this model, expression of these miRNAs would be responsible for controlling differentiation of hematopoietic cells so that once differentiation is activated certain miRNAs are down-regulated. Consistent with Georgantas et al. (47), hsa-miR-10a was normally expressed in BM MNC and CD34+ cells of healthy donor samples, but the expression of this miRNA was down-regulated in CML patient samples. Furthermore, reexpression of hsa-miR-10a in CML cells decreased cell growth, thus supporting the potential role of hsa-miR-10a in CML disease.
Different mechanisms, such as (22) localization of miRNA in a loss of heterozygosity region (8), epigenetic methylation of the CpG islands of miRNAs (48, 49), mutations in miRNA sequences (22), or altered miRNA maturation or stability (50), could be implicated in the regulation of miRNA expression. hsa-miR-10a is localized in 17q21.32 genome region, which is not a chromosome fragile site or a commonly deleted chromosome region in CML. Furthermore, a defined CpG island could not be found near the hsa-miR-10a chromosome localization, making unlikely that an epigenetic methylation mechanism participates in the regulation of hsa-miR-10a in CML. In addition, we did not detect an up-regulation of hsa-miR-10a expression after treatment of KU812 cells with the inhibitor of histone deacetylase trichostatin A or the demethylating agent 5-aza-2′deoxycytidine (data not shown), further indicating that hsa-miR-10a expression is not regulated by abnormal methylation or histone acetylation.
Down-regulation of hsa-miR-10a in CML could not be explained by the presence of gene mutations, as we did not find any point mutations in the 201-bp genomic region that includes the hsa-miR-10a sequence using any of the studied samples of CML CD34+ (data not shown). We also compared the expression of DROSHA and DICER genes, components of the miRNA maturation machinery, by Q-RT-PCR in 16 BM samples of Philadelphia chromosome–positive, chronic-phase CML patients and 10 healthy donors to determine whether the maturation of miRNA could be responsible for the abnormal expression of hsa-miR-10a. No statistically significant differences in the expression of DROSHA or DICER were found between CML and healthy donors, suggesting that alterations in the processing machinery of miRNAs are not responsible for the decreased expression of hsa-miR-10a in CML (data not shown). We also believe that hsa-miR-10a should be properly matured in CML cells as we failed to detect accumulation of hsa-miR-10a pre-miRNA or pri-miRNA precursors by primer extension (data not shown). Further studies will be required to identify the mechanisms that regulate the hsa-miR-10a expression in CML.
In conclusion, our study shows that the expression of the hsa-miR-10a miRNA is down-regulated in CML patients and that hsa-miR-10a regulates expression of USF2. Further, these abnormalities are independent of the BCR-ABL1 activity and are implicated in the abnormal growth of CML cells. Further studies will be necessary to identify other hsa-miR-10a targets and to show the exact role of hsa-miR-10a and its target genes in the development of CML.
Materials and Methods
Cell Lines and Patient Samples
Human-derived Mo7e, Mo7e transfected with p210 isoform of BCR-ABL1 (Mo7e-p210), and CML-derived cell lines TCC-S, KU812, and BV173 were grown in RPMI 1640 supplemented with 20% fetal bovine serum, 20 mmol/L HEPES buffer, and penicillin-streptomycin (all of them from BioWhittaker) with (Mo7e) or without (Mo7e-p210, TCC-S, KU812, and BV173) 10 μg/μL of recombinant human interleukin-3 (PeproTech) as described (45, 46). In some experiments, Mo7e-p210 cells were treated with 2 μmol/L imatinib for 24 h. Colon carcinoma–derived RKO cell line was grown in MEM (Life Technologies) supplemented with 10% fetal bovine serum and penicillin-streptomycin (BioWhittaker).
BM cells were obtained from patients with CML at diagnosis and from normal volunteer donors. All CML patients were 100% Philadelphia chromosome positive by conventional cytogenetic analysis. MNCs were isolated by sedimentation on Ficoll-Hypaque gradients. CD34+ cells were isolated using the MACS CD34+ isolation kit (Miltenyi Biotec) and the AutoMACS selection device as described (45). CD34+ purity was always above 90%. All studies were approved by the Investigational Review Boards at the University of Navarra. All the samples were obtained after informed consent of the patients and donors.
RNA Extraction and Reverse Transcription
Total RNA was extracted with Ultraspec (Biotecx) following the manufacturer's instructions. Total RNA (5 ng) was used to synthesize a specific cDNA of each analyzed miRNA according to the Taqman MicroRNA Assay protocol (PE Applied Biosystems). Reverse transcriptase reactions contained 1.5 μL of miRNA-specific stem-loop reverse transcriptase primer and 6 μL of Master Mix composed of 5 ng RNA, 1× reverse transcriptase buffer, 0.25 mmol/L of each deoxynucleotide triphosphate, 3.33 units/μL of MultiScribe reverse transcriptase, and 0.25 unit/μL of RNase inhibitor (PE Applied Biosystems). The 7.5 μL reactions were incubated in a PE Applied Biosystems 2720 Thermo Cycler in a 96-well plate for 30 min at 16°C, 30 min at 42°C, and 5 min at 85°C.
Expression Analysis by Q-RT-PCR
Expression of 157 miRNAs was analyzed using specific primers and Taqman probe for each miRNA according to the Taqman MicroRNA Assay protocol. Q-RT-PCR was done in a PE Applied Biosystems 7300 Sequence Detection system using 0.7 μL of reverse transcriptase product of each miRNA in a reaction volume of 10 μL with 1× Taqman Universal PCR master mix and 1 μL of primer and probe mix according to the Taqman MicroRNA Assay protocol (PE Applied Biosystems). The reactions were incubated in a 96-well optical plate at 95°C for 10 min followed by 40 cycles of 95°C for 15 s and 60°C for 10 min. The Ct data were determined using default threshold settings.
Validation of miRNAs was done by Q-RT-PCR using the standard curve method [User Bulletin N°2, part number (PN) 4303859; PE Applied Biosystems]. Individual Taqman miRNA assays for hsa-miR-10a (PN: 4373153), hsa-miR-96 (PN: 4373010), hsa-miR-125a (PN: 4373149), hsa-miR-150 (PN: 4373127), and hsa-miR-151 (PN: 4373179) were used (PE Applied Biosystems).
Expression of HOXA1 was analyzed using Assay-on-Demand (Hs 00171793_m1; PE Applied Biosystems). Q-RT-PCR for USF2, DROSHA, and DICER was done with the LightCycler technology using 1 μL cDNA in 20 μL reaction volume with 0.4 μm/L of each primer [USF2, 5′-AAAAATTGATGGAACCAGAACACC-3′ (forward) and 5′-TCCCGTCTTGCTGTTGTC-3′ (reverse); DROSHA, 5′-CAAGATGACCCAACTCCCT-3′ (forward) and 5′-TTCTGAACAATGGCAGTCCG-3′ (reverse); and DICER, 5′-TCTGCCAATTAACTCACCTC-3′ (forward) and 5′-TAACTATCCCTCAAACACTCTG-3′ (reverse)] and 2 μL of 10× LightCycler FastStart DNA Master SYBR Green I. The final Mg2+ concentration in the reaction mixture was adjusted to 3.5 mmol/L. The following program conditions were applied for Q-RT-PCR running: 95°C for 10 min; amplification program, consisting of 45 cycles at 95°C for 10 s, 62°C for 10 s, and 72°C for 10 s; melting program, 1 cycle at 95°C for 0 s, 40°C for 60 s, and 90°C for 0 s; and cooling program, 1 cycle at 40°C for 60 s. The temperature transition rate was 20°C/s, except in the melting program, which was 0.2°C/s between 40°C and 90°C. U6 gene was used as reference gene. To reduce the variation between different assays and samples, a procedure based on the relative quantification of target genes versus their controls/calibrators in relation to the reference gene was used. Calculations were automatically done by the LightCycler software (RealQuant, version 1.0, Roche). The normalized ratios of USF2 (NUSF2), DROSHA (NDROSHA), and DICER (NDICER) expression were obtained from the following equation and expressed as percentage of the control/calibrator:
Efficiencies (E) of each gene were calculated from the slopes of crossover points (Cp) versus DNA concentration plot according to the formula E = 10(−1/slope). ΔCp corresponded to the difference between control/calibrator Cp and sample Cp, either for the target or for the reference sequences. The selected controls/calibrators were BM specimens from healthy donors. They were considered as 100% expression.
hsa-miR-10a expression was analyzed by Primer extension technique as described previously (51) using (5′-TACACAAATTCGGATCTACA-3′) oligonucleotide for hsa-miR-10a and (5′-TGCTAATCTTCTCTGTATCGT-3′) oligonucleotide for U6. Six picomoles of γ-ATP–labeled hsa-miR-10a oligonucleotide were incubated with 3 μg of total RNA and only 0.25 μg of RNA with oligonucleotide U6. Samples were loaded onto a 14% polyacrylamide gel and separated by electrophoresis. Gels were dried and exposed to a screen that was developed in a Cyclone phosphorimager (Perkin-Elmer).
Normalization and Data Analysis
Expression of miRNAs was normalized using the expression of GUSB or U6 genes in each sample as previously described (34). Primers for expression analysis of GUSB were GUS-R (5′-CCGAGTGAAGATCCCCTTTTTA-3′) for the reverse transcription and primers GUS-F (5′-GAAAATATGTGGTTGGAGAGCTCATT-3′), GUS-R, and probe GUS-P (5′-FAMCCAGCACTCTCGTCGGTGACTGTTCATAMRA-3′) for the Q-RT-PCR. For expression of U6, we used Taqman RNU6B assay (PN 4373381; PE Applied Biosystems). Relative quantification of expression of analyzed 157 miRNAs was calculated with the 2−ΔΔCt method (User Bulletin N°2, PN 4303859; PE Applied Biosystems). The data are shown as ΔCt of the relative quantity of target miRNAs, normalized, and compared with expression in healthy donor samples.
An unsupervised cluster analysis was carried out using Cluster and TreeView software (52). Hierarchical clustering based on the average linkage method with the centered correlation metric was done. To identify genes with statistically significant changes in expression between both groups, we did a supervised analysis using the algorithm significant analysis of microarrays (53). All data were permuted over 100 cycles by using the two-class (unpaired) format.
Finally, miRNA expression was compared between CML and healthy donor samples using Statistical Package for the Social Sciences package, version 13.0 (SPSS). Differences between two independent groups were analyzed by the Student's t test or the Mann-Whitney U test depending on variable distribution (54). A P value of ≤0.05 was considered significant.
Cell Line Transfection
The pre-miR-10a precursor molecule, negative precursor-miR control, anti-miR-10a inhibitor, negative anti-miR control, USF2 siRNA (AM16708), and Silencer Negative Control-1 were purchased from Ambion. The colon carcinoma–derived RKO cell line was transfected with each of the precursor and inhibitor molecules at final concentration of 50 nmol/L using Lipofectamine RNAiMAX (Invitrogen Life Technologies) according to the manufacturer's instruction. In a different set of experiments, the CML-derived cell line KU812 was nucleofected using 100 μL of Cell Line Nucleofector Solution V (Amaxa GmbH) with each of the precursor-miR, inhibitor-miR molecules, USF2 siRNA, and Silencer Negative Control-1 at a final concentration of 50 nmol/L using program T-06 from the Nucleofector device (Amaxa GmbH). After 48 h, cells were harvested to analyze the expression of hsa-miR-10a, HOXA1, or USF2 by Q-RT-PCR. Transfection efficiency was determined by flow cytometry using the Silencer FAM-labeled Negative Control (Ambion) for the RKO cell line and BLOCK-iT Fluorescent Oligo (Invitrogen Life Technologies) for the KU812 cell line. Cell viability and total cell counts were determined at various times by trypan blue exclusion and found to be always above 80%.
hsa-miR-10a Genomic Sequence Analysis
DNA from CML cells was extracted using QIAamp DNA Mini kit (Qiagen). PCR for amplification of the 201-bp genomic sequence where hsa-miR-10a is located was done in a total volume of 25 μL using 1 μL genomic DNA, 1 unit AmpliTaq Gold DNA Polymerase (PE Applied Biosystems), 1.5 mmol/L MgCl2, 0.2 mmol/L deoxynucleotide triphosphates, and 20 pmol of each primer (10a-D: 5′-CTGGAAAATTTCTGGGCCAA-3′ and 10a-R: 5′-CCAGACTGTCCTCATTCAGAAAAA-3′). Amplification products were purified using QIAquick PCR Purification kit (Qiagen) and sequenced with ABI PRISM d-Rhodamine Terminator Cycle Sequencing kit in an ABI PRISM 377 DNA Sequencer (PE Applied Biosystems).
Renilla Luciferase Assay
Renilla luciferase construct Renilla-USF2 3′-UTR was made by ligating oligonucleotides (USF2-3′-UTR-D: 5′-CTAGAGCTGCCTCCTGCTCTCTGGAGGTACTGAGACAGGGTGCTGATGGGAAGGAGGAGC-3′; USF2-3′-UTR-R: 5′-GGCCGCCCTCCTTCCCATCAGCACCCTGTCTCAGTACCTCCAGAGAGCAGGAGGCAGCT-3′) containing the putative target site of USF2 3′-UTR into XbaI-NotI site of the Renilla reporter vector (pRL-SV40, Promega). Cells (250,000) of RKO cell line were cotransfected using Lipofectamine (Invitrogen) with 0.5 μg of Renilla-USF2 3′-UTR vector, 2 μg of the control vector containing firefly luciferase (pGL3-Promoter, Promega), and 50 mmol/L of pre-miR-10a or negative precursor-miR vector. Renilla luciferase activity was measured 48 h after transfection using Dual Luciferase System (Promega) in a Berthold Luminometer (Lumat LB 9507) as previously described (51).
Western Blot Analysis
Proteins extracted from the CML cell line KU812 nucleofected with USF2 siRNA and Silencer Negative Control-1 were analyzed by Western blot as previously described (55). The membranes, after being blocked, were incubated first with primary antibody for USF2 (diluted 1:1,000 for 2 h; Santa Cruz Biotechnology) or β-actin (1:4,000 for 1 h; Sigma) and then with alkaline phosphatase–conjugated secondary antibodies (1:10,000 for 1 h; Sigma). Bound antibodies were revealed by a chemiluminescent reagent (Tropix) and detected using Hyperfilm enhanced chemiluminescence (Amershan Biosciences). β-Actin was used as a loading control.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Grant support: Beca Ortiz de Landázuri 2006, Departamento de Salud-Gobierno de Navarra, Fondo de Investigación Sanitaria (Spain) PI060285, PI070602, PI070608, PI060003, PI030141, PI030661, PI021299, and ISCIII-RETIC RD06/0020 Junta de Andalucia 03/143; 03/144; 06/0356, 0004/2007; IMABIS (Málaga, Spain), Fundación de Investigación Médica Mutua Madrileña Automovilista; Asociacion Medicina e Investigacion; and “UTE project CIMA.”
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.
Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/).