Genetic Alterations in MicroRNAs in Medulloblastomas

Abstract

MicroRNAs (miRNAs) regulate a variety of cellular processes via the regulation of multiple target genes. We screened 48 medulloblastomas for mutation, deletion and amplification of nine miRNA genes that were selected on the basis of the presence of potential target sequences within the 3′‐untranslated region of the MYCC mRNA. Differential PCR revealed deletions in miR‐186 (15%), miR‐135a‐1 (33%), miR‐548d‐1 (42%), miR‐548d‐2 (21%) and miR‐512‐2 (33%) genes, whereas deletion or amplification was detected in miR‐135b (23%) and miR‐135a‐2 (15%). In miR‐33b, deletion, amplification or a mutation at the precursor miRNA were detected in 10% of medulloblastomas. Overall, 35/48 (73%) medulloblastomas had at least one alteration. Real‐time RT‐PCR revealed MYCC overexpression in 11 of 37 (30%) medulloblastomas, and there was a correlation between MYCC overexpression and miR‐512‐2 gene deletion (P = 0.0084). Antisense‐based knockdown of miR‐512‐5p (mature sequence of miR‐512‐2) resulted in significant upregulation of MYCC expression in HeLa and A549 cells, while forced overexpression of miR‐512‐2 in medulloblastoma/PNET cell lines DAOY, UW‐228‐2, PFSK resulted in the downregulation of MYCC protein. Furthermore, the results of luciferase reporter assays suggested that miR‐512‐2 targets the MYCC gene. These results suggest that alterations in the miRNA genes may be an alternative mechanism leading to MYCC overexpression in medulloblastomas.

INTRODUCTION

Medulloblastoma (World Health Organization grade IV) is a common malignant brain tumor among children (21). The outcome for patients with medulloblastoma is still poor, even with multimodal treatments, including surgical resection followed by radiotherapy and chemotherapy (32).

Several molecular markers that predict the prognosis of medulloblastoma patients have been identified 31, 34, 37. Those predicting a favorable outcome include nuclear beta‐catenin immunoreactivity (12) and high expression of TrkC mRNA 11, 16, 17, while those predicting poor survival include amplification 1, 24 and overexpression 18, 19 of MYCC. High expression of MYCC mRNA has been significantly associated with tumor anaplasia 11, 39 and was identified as a significant prognostic factor for poor clinical outcome in medulloblastomas/primitive neuroectodermal tumor (PNET) 11, 18. MYCC amplification has been observed in a small fraction of medulloblastomas (5%–10%) 1, 18, 19, 24, while MYCC overexpression appears to be much more common (30%–50%) 11, 18, 19. MYCC mRNA expression levels in medulloblastomas/PNET vary widely, with a 22‐fold difference between the highest and lowest values, and do not correlate with amplification of the MYCC gene (18). These observations suggest that amplification of the MYCC gene cannot explain the overexpression of MYCC mRNA in most cases, suggesting that an alternative mechanism is involved.

MicroRNAs (mMiRNAs) are small non‐coding RNAs that target protein‐coding mRNAs by repressing translation or causing mRNA degradation 2, 3. Mature miRNAs comprise about 22 nucleotides, which derived from long transcripts of primary miRNAs (approximately 500–3000 bases) and precursor miRNAs (approximately 70 bases) (38). miRNAs play important roles in regulating a variety of cellular processes, including differentiation, proliferation and apoptosis, via the regulation of multiple target genes (8). More than 50% of miRNA genes are located at fragile sites and genomic regions that are frequently altered in cancer (8). Deletion or amplification of miRNA genes has been observed in a variety of human neoplasms (44), while mutation of miRNA genes appears to be rare 10, 41, 42.

Several studies have shown that miRNA‐expression profiling can distinguish medulloblastoma from normal cerebellar tissue, and also differentiate between different histologic subtypes (eg, anaplastic, classic, desmoplastic) 14, 33 and those with specific genetic alterations 13, 14, 33. Medulloblastomas that overexpress ErbB2 or MYCC have a distinct miRNA‐expression profile (14). Significantly decreased expression of miR‐124a has been associated with decreased expression of CDK6 protein in medulloblastoma (35). Expression profiling of 427 miRNAs in 90 medulloblastomas has revealed that expression of miR‐17/92 is highest in those tumors in which the sonic hedgehog (Shh) signaling pathway is activated (29). Downregulation of miR‐324‐5p was typically observed in medulloblastomas with 17p deletion (13). However, little is known whether miRNA genes are genetically altered in medulloblastomas. One study showed amplification of the miR‐17/92 polycistron proto‐oncogene in 6% of medulloblastomas (29).

The objective of the present study was to assess the frequency of genetic alterations of the miRNA genes in medulloblastomas. Using tissue samples from 48 cases of medulloblastoma, we assessed amplification, deletion, and mutation, of nine miRNA genes which were selected on the basis of the presence of potential target sequences within the 3′‐untranslated region of the MYCC messenger RNA (miR‐33b, miR‐135a‐1, miR‐135a‐2, miR‐135b, miR‐186, miR‐200b, miR‐512‐2, miR‐548d‐1 and miR‐548d‐2). The results were correlated with MYCC RNA expression as assessed by real‐time RT‐PCR.

MATERIALS AND METHODS

Samples

Samples of 48 medulloblastomas and 10 normal tissues were obtained from University Hospital, Zurich, Switzerland. The samples included 47 cases of classic medulloblastoma and one case of desmoplastic/nodular medulloblastoma, with 32 cases being in males and 16 cases in females. The mean age of the patients was 16.1 ± 13.1 years (range 1–60 years). The samples were fixed in buffered formalin and embedded in paraffin.

Selection of miRNAs for genetic analyses

According to a computer prediction of microRNA targets, the human MYCC 3′ UTR (467 bp) contains more than 30 potential miRNA target sites (http://www.targetscan.org, release 5.1, April 2009). Of these, we selected nine miRNA genes that are located in chromosomal loci reported to be deleted or amplified in medulloblastoma according to an array‐CGH database (http://www.progenetix.net/progenetix/). The miRNAs selected were: miR‐186 (located at 1p31.1), miR‐200b (1p36.33), miR‐135b (1q32.1), miR‐135a‐1 (3p21.1), miR‐548d‐1 (8q24.13), miR‐135a‐2 (12q23.1), miR‐33b (17p11.2), miR‐548d‐2 (17q24.2) and miR‐512‐2 (19q13.42).

Extraction of DNA and RNA from paraffin‐embedded histological sections

Medulloblastoma tissues (48 cases) and normal brain tissues (10 cases) surrounding medulloblastomas were manually scraped from the paraffin sections, and DNA was extracted using a previously‐reported protocol (22). Total RNA was extracted from 37 medulloblastomas using the RecoverAll™ total nucleic acid isolation kit (Ambion, Huntingdon, UK), with slight modifications based on previously published protocols 23, 25. Briefly, samples from four sections (5 µm) were placed in RNase‐free, 1.5 mL eppendorf tubes and deparaffinized with xylene. Samples were then washed with ethanol, centrifuged for 5 minutes at 14 000 rpm, air‐dried, and resuspended in 100 µL digestion buffer (Ambion) containing 4 µL proteinase K (60 units/µL). Samples were vortexed and incubated for 3 h at 50°C followed by 5 min at 80°C for RNA isolation. For DNA isolation, samples were incubated with protease for 16 h at 50°C. For nuclease digestion and final purification, either 60 µL of DNase (for RNA isolation) or RNase (for DNA isolation) was added and samples were incubated for 30 min at room temp. Samples were then washed twice with wash buffer and were centrifuged to remove residual fluid. Samples were eluted with 60 uL elution solution or nuclease‐free water at room temp (RNA) or 95°C (DNA).

Differential PCR to detect amplification and homozygous deletion of miRNA genes

MiRNA sequences were obtained from miRBase (The Welcome Trust Sanger Institute; http://microrna.sanger.ac.uk/sequences/search.shtml). Primers to amplify miRNA sequences covering precursor miRNA were designed using Primer3 (v.04.0) software (http://frodo.wi.mit.edu) (Table 1).

Table 1.

Primer sequences of miRNAs.

miRNA gene	Location	Primer sequences	Annealing temperature	PCR product
miR‐33b	17p11.2	F: 5′‐GGTGGATGGTGTGGTAGGA‐3′	62°C	216 bp
		R: 5′‐CTCTGGGAGGGGCAGGAT‐3′
miR‐135a‐1	3p21.1	F: 5′‐GAAGAAGTGCCTGCAAGAGC‐3′	60°C	230 bp
		R: 5′‐GGAATAGAGGAACGGCTGTG‐3′
miR‐135a‐2	12q23.1	F: 5′‐GTTTTGCATCCGACCAAGATA‐3′	60°C	223 bp
		R: 5′‐AAAGGAACACCAGGCAGGTAG‐3′
miR‐135b	1q32.1	F: 5′‐AGCTTCTCGCTTCCCTATGAG‐3′	60°C	217 bp
		R: 5′‐AAGCAAAGCCTCCTTCTGGT‐3′
miR‐186	1p31.1	F: 5′‐CCCATCATATTCTTCCCAAAC‐3′	59°C	209 bp
		R: 5′‐TTGACATTCACATGCTTCAGG‐3′
miR‐200b	1p36.33	F: 5′‐TACTGAGCTTCCCAGCGAGT‐3′	60°C	217 bp
		R: 5′‐CTGTGTGGGAGGGGAGTGT‐3′
miR‐512‐2	19q13.42	F: 5′‐TAGAGGATGTGCCTGCAGTTT‐3′	60°C	219 bp
		R: 5′‐AAGCACCACGGTTTAGCTTTT‐3′
miR‐548d‐1	8q24.13	F: 5′‐GCTAGCAGAAAACTGTGTTGG‐3′	62°C	225 bp
		R: 5′‐TCATGAGAAGAAAATGCCACA‐3′
miR‐548d‐2	17q24.2	F: 5′‐GGGGGTATATGATGTCAACCTG‐3′	60°C	217 bp
		R: 5′‐TGTGGACCTACGTCTTCATTTG‐3′

To detect amplification and homozygous deletion of miRNA genes, differential PCR was performed as described previously (36), using the beta‐actin sequence as a reference. Primer sequences for beta‐actin were 5′‐CTG TGG CAT CCA CGA AAC TA‐3′ (sense) and 5′‐AGG AAA GAC ACC CAC CTT GA‐3′ (antisense) to produce a 187‐bp fragment. After PCR (33 cycles), the products were separated on 8% acrylamide gels. Gels were stained with ethidium bromide. Quantitative analysis of the signal intensity was performed with Molecular Imager and the Quantity One Analysis Software (Bio‐Rad, Hercules, CA, USA). The mean ratio of miRNA gene to beta‐actin of normal DNA (10 samples of normal tissue) was about 1.0. The threshold for evidence of miRNA gene amplification was calculated using the formula 2 × mean + 3 × standard deviation (SD) (36). Samples in which the ratio of miRNA genes to beta‐actin was ≤0.2 were considered to show homozygous deletion of miRNA genes (28).

SSCP and DNA sequencing to detect miRNA gene mutations

Pre‐screening for miRNA mutations was carried out using single‐strand conformational polymorphism (SSCP) as described previously (40) using the primers described earlier (Table 1). Samples that showed mobility shifts on SSCP analysis were further analyzed by direct DNA sequencing (ABI 3100 PRISM DNA Sequencer, Applied Biosystems, Foster City, CA, USA).

Differential PCR to detect amplification of the MYCC gene

To detect MYCC gene amplification, differential PCR was performed as described previously (36). The primer sequences used were 5′‐TCT GGA TCA CCT TCT GCT GG‐3′ (sense) and 5′‐TGT TGC TGA TCT GTC TCA GG‐3′ (antisense), designed to produce a 180‐bp fragment. The glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH, ENSG00000111640) sequence was used as a reference. The primers used to amplify GAPDH were 5′‐AAC GTG TCA GTG GTG GAC CTG‐3′ (sense) and 5′‐AGT GGG TGT CGC TGT TGA AGT‐3′ (antisense), which produced a 160‐bp sequence. The PCR was carried out with 33 cycles, with an annealing temperature of 60°C. The PCR products were loaded on 8% acrylamide gels and stained with ethidium bromide. Quantitative analysis of the signal intensity was performed using the Molecular Imager and Quantity One Analysis Software (Bio‐Rad). The mean ratio of MYCC to GAPDH for DNA from 10 normal tissues was 0.97, with an SD of 0.04. The threshold for evidence of MYCC gene amplification was calculated using the formula 2 × mean + 3 × SD (36); a ratio of MYCC: GAPDH of >2.06 was considered to indicate MYCC gene amplification.

Real‐time RT‐PCR to detect expression of the MYCC gene

Synthesis of cDNA was carried out using 500 ng of total RNA according to the protocol of SuperScript III First‐Strand Synthesis System (Invitrogen, Life Technologies, CA, USA). Real‐time quantitative reverse transcription‐PCR (q‐RT‐PCR) was conducted to detect MYCC expression as previously described 4, 23 using the iCycler iQ5 Detection System (Bio‐Rad). iQ SYBR Green Supermix Kit was purchased from Bio‐Rad. Primers for the analysis of MYCC and GADPH mRNA were designed to amplify sequences spanning introns to eliminate possible false‐positive amplification caused by DNA contamination. The primer sequences used were 5′‐TTC GGG TAG TGG AAA ACC AG‐3′ (sense) and 5′‐TCC TGT TGG TGA AGC TAA CG‐3′ (antisense) to amplify a 66‐bp fragment of MYCC; 5′‐CAA TGA CCC CTT CAT TGA CC‐3′ (sense) and 5′‐CTT CCC GTT CTC AGC CTT G‐3′ (antisense) to amplify a100‐bp fragment of GAPDH. Real‐time RT‐PCR was performed according to the manufacturer's recommendations. Briefly, 40 ng (2 µL) cDNA was used as template, 5 µL 2 × SYBR Green Supermix with 0.5 µL of sense and of antisense primers, in a total volume of 10 µL mixture with nuclease‐free water. The PCR conditions were 3 minutes at 95°C for denaturation, followed by 40 cycles of 10 s at 95°C and 30 s at 60°C. Fluorescence data were specified for collection during the 60°C step. The amount of MYCC, normalized to the reference control GAPDH, was related to the commercially available calibrator, human cerebellum (pooled from 10 male/female Caucasians, age range 22–68 years; Clontech, Mountain View, CA, USA) (37). The relative expression of MYCC mRNA was calculated using the comparative cycle threshold (CT) method. The amount of MYCC relative to the calibrator, is given by 2^−ΔΔCT, where ΔΔCT = ΔCT (sample) − ΔCT (calibrator) 23, 25, 26.

Knockdown of miR‐512‐5p

HeLa cells and A549 human alveolar basal epithelial cells were used to assess the effect of miR‐512‐5p (mature sequence of miR‐512‐2 gene) on MYCC expression. Antisense of mature miR‐512‐5p 3′ end sequence was used to block the endogenous expression of miR‐512‐5p via a pGC‐FU lentiviral vector (GeneChem, Shanghai, China). Scrambled sequence (5′‐TTT CTC CGA ACG TGT CAC GT‐3′) was used as a negative control (NC). Cell infections were performed at the optimum conditions with multiplicity of infection (MOI) treatments. For HeLa cells, MOI of lentivirus were 10 and 20, and for A549 cells, MOI were 20 and 40. Cells were harvested 5 days after infection, and RNA and protein were extracted. Expression of miR‐512‐5p was quantified using a Taqman microRNA assay kit. MYCC expression was measured by quantitative PCR. Western blotting was performed to detect the level of MYCC protein expression using MYCC antibody (Ab11917; Abcam, Cambridge, MA, USA). To test proliferation ability, the 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) assay was carried out 2 days after infection in 96‐well plates (2000 cells/well). The absorbance [optical density (OD) 490 nm) was determined in triplicate every 24 h over 5 days.

In silico prediction of genes targeted by miR‐512‐5p

The analysis of the predicted targets of miR‐512‐5p in the MYCC 3′UTR was carried out using the Targetscan algorithms (http://www.targetscan.org). We identified miR‐512‐5p target site resides at 100–106 nt of the MYCC 3′ UTR.

Luciferase reporter assay

HEK 293 cells were plated in triplicate into six‐well plates. On the second day, the cells were transfected with 1.4 µg of miR‐512‐2 precursor miRNA expression clone (gguacuucucagucuguggcacucagccuugagggcacuuucuggugccagaaugaaagugcugucauagcugagguccaaugacugaggcgagcacc) in pEZX‐MR03 vector with GFP [pEZX‐miR‐512‐2] and co‐transfected with either 1.0 µg of MYCC 3′ UTR miRNA target sequence expression clone in pEZX‐MT01 vector [pEZX‐MT01‐MYCC UTR‐fluc] or with control vector [pEZX‐MT01]. The cells were transferred to a 96‐well plate 18 h after transfection and cultured for another 24 h. Firefly luciferase expression is regulated by binding of the targeting miRNA to the 3′ UTR target sequence. Both firefly luciferase and Renilla luciferase activities were measured with a colorimetric assay. Firefly luciferase activity was then normalized with Renilla luciferase activities in the same well.

Transfection with pre™‐miR‐512‐2

DAOY and UW‐228‐2 medulloblastoma cells and PFSK PNET cells were seeded in six‐well plates at a cell density of 2 × 10⁵. Transfection of pre™‐miR‐512 (Ambion AM17100 ID PM10944) or NC referred to as “scrambled” (Ambion AM17110) in a final concentration of 30 nmol/L was performed using siPORT NeoFX transfection reagent (Ambion) according to the manufacturer's recommendations. Cells were harvested 24 h after transfection for subsequent quantitative RT‐PCR and Western blot analysis. Six random fields were counted for each sample. Experiments were repeated twice independently.

Quantitative RT‐PCR to detect miR‐512‐5p expression

Total RNA was extracted using the mirVana miRNA isolation kit (Ambion AM1561, Austin, TX, USA) following the manufacturer's instructions. 10 ng of RNA enriched for small RNAs was used as template for reverse transcription, which was triggered by TaqMan microRNA assay reverse‐transcription primers. Quantitative RT‐PCR was performed under conditions optimized for the ABI7900HT instrument, using TaqMan universal PCR Master Mix, No AmpErase UNG (Applied Biosystems). Probe‐primer specific for miR‐512‐2 (Applied Biosystems) was hsa‐miR‐512‐5p. The relative gene expression was calculated for each gene of interest by using the ΔΔCT method, where CT values were normalized to the housekeeping gene RNU6B.

Western blot analysis

Extracts of total protein were obtained from 0.5–1.5 × 10⁶ cells lysed with radioimmunoprecipitation assay (RIPA) buffer [50 mM Tris‐Cl, pH 6.8, 100 mM NaCl, 1% Triton X‐100, 0.1% sodium dodecyl sulphate (SDS)] supplemented with Complete Mini Protease Inhibitor Cocktail (Roche‐Applied Sciences, Indianapolis, IN, USA) and with the phosphatase inhibitors β‐glycerophosphate (20 mM) and Na3Vo4 (200 µM). Proteins were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis and Western blotting on polyvinylidene fluoride (PVDF) membranes (GE Healthcare UK, Buckinghamshire, UK). After binding of antibodies, the signals were detected by chemiluminescence using Super‐Signal West Femto Maximum Sensitivity Substrate (Pierce‐Thermo Scientific, Rockford, IL, USA). Antibody specific for MYCC was purchased from Cell Signaling Technology, Danvers, MA, USA (#9402). Beta‐actin (Sigma‐Aldrich, St. Louis, MO, USA) was used as loading control.

Statistical analyses

Chi‐squared tests and Fisher's exact tests were performed to assess correlations between overexpression of MYCC and genetic alterations in the selected miRNAs using the statistics software package SAS version 9.1 (SAS, Cary, NC, USA). We declared statistical significance at the traditional threshold of P < 0.05.

RESULTS

Genetic alterations in miRNA genes

Differential PCR revealed deletions in miR‐186 (15%), miR‐135a‐1 (33%), miR‐548d‐1 (42%), miR‐548d‐2 (21%) and miR‐512‐2 (33%) in the 48 medulloblastomas analyzed (Table 2). Deletions or amplifications were found in miR‐135b (23%) and miR‐135a‐2 (15%). For miR‐33b, deletion, amplification or a point mutation (a G to A substitution at 76 nt in precursor miR‐33b; 41 bp downstream of the mature miR‐33b) were detected in five (10%) medulloblastomas (Table 2, Figure 1). This mutation was not detected in normal brain tissue surrounding the medulloblastoma (Figure 1). No alteration in miR‐200b was detected in any of the medulloblastomas (Table 2). Overall, 35 out of 48 (73%) medulloblastomas had at least one alteration in the nine miRNA genes analyzed.

Table 2.

miRNA genetic alterations and MYCC expression in medulloblastomas. Abbreviations: Amp = amplification; Del = deletion; + = overexpression; — = no alteration; nd = not determined.

Case No.	Data base ID	Age (years)	Sex	miR‐33b	miR‐135a‐1	miR‐135a‐2	miR‐135b	miR‐186	miR‐200b	miR‐512‐2	miR‐548d‐1	miR‐548d‐2	MYCC Overexpression
T11	16429	1	M	—	Del	Del	Del	Del	—	Del	Del	Del	+
T45	20504	1	M	—	—	—	—	—	—	—	—	—	nd
T3	14827	2	F	—	—	—	Amp	—	—	—	—	—	—
T15	17578	2	M	—	Del	—	—	—	—	Del	—	—	+
T36	19785	2	F	—	—	—	—	—	—	—	Del	—	—
T39	19965	2	M	—	—	—	—	—	—	Del	Del	—	+
T44	20503	2	M	—	Del	—	—	Del	—	Del	Del	Del	+
T22	18523	4	M	—	—	—	—	—	—	—	—	—	nd
T47	21160	4	F	—	Del	—	—	—	—	—	Del	—	—
T23	18581	5	F	—	—	Amp	—	—	—	—	—	—	—
T6	15747	6	M	—	—	—	—	—	—	—	—	—	+
T18	18020	6	F	—	Del	—	—	—	—	Del	Del	—	—
T25	18745	6	M	—	—	—	—	—	—	—	—	—	nd
T28	19063	6	F	—	—	—	—	—	—	Del	—	—	nd
T42	20285	6	M	—	—	—	Amp	—	—	Del	—	Del	+
T12	16497	7	M	—	—	—	—	—	—	Del	Del	Del	+
T7	15811	8	M	—	Del	Amp	—	—	—	—	Del	—	nd
T26	18883	8	M	—	Del	—	—	Del	—	Del	Del	Del	+
T46	20508	8	M	—	Del	—	—	—	—	Del	Del	—	—
T4	15171	9	M	—	—	—	Amp	—	—	—	—	—	—
T32	19604	9	M	Mut*	—	Del	—	—	—	—	—	—	—
T43	20422	10	M	—	Del	—	—	Del	—	Del	Del	—	—
T2	14430	11	M	—	Del	—	—	Del	—	Del	Del	Del	—
T9	15876	11	M	—	—	—	—	—	—	—	—	—	nd
T16	17796	14	F	—	Del	—	—	—	—	—	Del	—	nd
T5	15318	16	F	—	—	—	—	—	—	—	—	—	—
T14	16730	16	M	—	Del	Del	Del	Del	—	Del	Del	Del	—
T8	15829	17	M	Amp	—	Amp	Amp	—	—	—	—	Del	—
T19	18041	17	M	—	—	—	Amp	—	—	—	Del	Del	—
T21	18230	18	F	—	—	—	—	—	—	—	—	—	—
T24	18589	22	M	Amp	—	—	—	—	—	—	Del	—	—
T37	19823	22	M	—	—	Amp	Amp	—	—	—	—	—	—
T48	21539	22	F	—	—	—	—	—	—	—	—	—	—
T29	19147	24	M	—	—	—	—	—	—	—	—	—	nd
T17	18002	26	F	Del	—	—	—	—	—	—	—	—	—
T33	19612	26	M	—	Del	—	Amp	—	—	—	Del	—	—
T1	14327	27	F	—	—	—	—	—	—	—	Del	—	nd
T40	20256	27	M	—	—	—	—	—	—	Del	—	—	—
T13	16648	28	M	—	—	—	—	—	—	—	—	—	+
T10	16027	30	F	—	—	—	—	Del	—	—	—	—	—
T34	19643	31	M	—	—	—	—	—	—	—	—	—	—
T27	18886	32	M	—	Del	—	—	—	—	Del	Del	—	+
T38	19877	32	F	—	—	—	—	—	—	—	—	—	—
T31	19397	34	F	—	—	—	—	—	—	—	—	—	—
T41	20284	34	M	Del	Del	—	—	—	—	—	—	—	+
T30	19354	44	M	—	—	—	—	—	—	—	—	Del	—
T35	19781	60	M	—	Del	—	Del	—	—	Del	Del	—	+
T20	18136	?	F	—	—	—	Amp	—	—	—	—	—	nd
				5/48 (10%)	16/48 (33%)	7/48 (15%)	11/48 (23%)	7/48 (15%)	0/48 (0%)	16/48 (33%)	20/48 (42%)	10/48 (21%)	11/37 (30%)

^* G‐>A mutation at nucleotide 76 (G76A) in precursor miRNA (miR‐33b) gene.

Figure 1.

Direct sequencing showing a somatic G‐>A mutation at 76 nt in precursor miR‐33b (41 bp downstream from the mature miR‐33b sequence) in a medulloblastoma (case T32; left). Normal brain tissue surrounding the medulloblastoma shows no mutation (right).

Amplification and overexpression of MYCC

None of the 48 medulloblastomas analyzed by differential PCR showed amplification of the MYCC gene. Real‐time RT‐PCR revealed overexpression of MYCC in 11 out of the 37 medulloblastomas analyzed (30%; Table 2; Figure 2A). The frequency of miRNA gene alterations in medulloblastomas with and without MYCC overexpression were similar (9/11, 82% vs. 20/26, 77%), but there was a correlation between MYCC overexpression and deletion of miR‐512‐2 (P = 0.0084).

Figure 2.

A. Real‐time RT‐PCR showing a medulloblastoma (case T13) with MYCC overexpression, and another medulloblastoma (case T36) without MYCC overexpression. B. Data on normal cerebellum RNA. C. Relative cycle threshold (CT) values for MYCC and glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) at different cDNA concentrations. The slopes of the curves are similar, suggesting equal efficiencies of the two PCR reactions at CT. D. Relative levels of MYCC expression calculated using the comparative CT method. Abbreviation: N = normal cerebellum.

Knockdown of miR‐512‐5p increases MYCC expression in tumor cells

After blocking the endogenous expression of miR‐512‐5p (mature sequence of miR‐512‐2) via a pGC‐FU lentiviral vector, miR‐512‐5p expression was not detectable in HeLa or A549 cells, indicating that the knockdown was successful. Knockdown of miR‐512‐5p significantly upregulated the expression of MYCC at both the mRNA and protein levels in HeLa cells and in A549 cells (Figure 3A,B). The results of the MTT assay showed that the knockdown of miR‐512‐5p significantly promoted the proliferation of HeLa and A549 cells (P ≤ 0.002; Figure 3A,B).

Figure 3.

Knockdown of miR‐512‐5p (mature sequence of miR‐512‐2) leads to upregulation of MYCC expression at both the mRNA and protein levels in (A) HeLa cells and (B) A549 cells (P ≤ 0.002). MTT assay showed that the knockdown of miR‐512‐5p promoted proliferation of HeLa and A549 cells. Abbreviations: KD = cells in which miR‐512‐5p was knocked down; NC = negative control.

The inhibitory effect of miRNA‐512‐2 on a target sequence (3′ UTR) expression

Targeting of the MYCC gene by miR‐512‐2 was confirmed by a reporter assay for luciferase activity. Co‐transfection of a precursor miR‐512‐2 expression vector [pEZX‐miR‐512‐2] with the vector containing the 3′UTR of the MYCC gene [pEZX‐MT01‐ MYC UTR‐fluc] caused a reduction in luciferase activity compared with the controls (Figure 4)

Figure 4.

Co‐transfection of HEK 293 cells with pEZX‐vector containing MYCC 3′UTR together with a plasmid encoding miR‐512‐2 showed decreased luciferase activity (P‐value 0.0020 vs. pEZX‐control vector transfected cells). Both firefly luciferase and Renilla luciferase activities were measured with a colorimetric assay. Firefly luciferase activity was then normalized with Renilla luciferase activities in the same well.

Overexpression of miR‐512‐2 decreases MYCC protein level in tumor cells

To assess whether miR‐512‐2 affects MYCC expression, we determined the levels of MYCC protein in three medulloblastoma/PNET cell lines transfected with either pre‐miR‐512 or pre‐miR‐control (control). Quantitative RT‐PCR showed clear overexpression of miR‐512‐2 after 24 h of transfection with pre‐miR‐512 in all three medulloblastoma/PNET cell lines compared with the controls (Figure 5A). Western blot analysis revealed that medulloblastoma/PNET cells transfected with pre‐miR‐512 had a significantly decreased level of MYCC protein as quantified by densitometric analysis (25% in DAOY and PFSK, and 50% in UW‐228‐2 cells; Figure 5B).

Figure 5.

miR‐512‐2 overexpression reduces MYCC protein in medulloblastoma cells. A. Quantitative RT‐PCR analysis of miR‐512‐2 transfected with either pre‐miR‐512 or pre‐miR‐control shows pre‐miR‐512 is overexpressed in medulloblastoma cell lines. Y‐axis represent fold miRNA changes normalized to the endogenous control (RNU6B) cells. B. Western blot analysis revealed that overexpression of miR‐512‐2 downregulated MYCC expression. Values represent fold decrease of MYCC protein relative to control (n = 3; ± SD).

DISCUSSION

As is the case for oncogenes and tumor suppressor genes, abnormal expression of miRNA in cancer may be caused by genetic alterations. Calin et al (7) demonstrated that deletions of the miR‐15 and miR‐16 genes at 13q14 are frequent (68%) in chronic lymphocytic leukemia, which was associated with a significant reduction in their expression. Similarly, Davidson et al (9) showed that miRNA‐218 was associated with gene dosage in lung squamous cell carcinomas. O'Hara et al (30) assessed changes in gene copy number, pre‐miRNA and mature miRNA levels for primary effusion lymphomas, demonstrating that >50% of cases showed gene amplification and concordant changes in pre‐mature miRNA and mature miRNA levels of the miR‐153‐1, miR‐218‐1, miR‐107, miR‐188 and miR‐125a genes. Huse et al (20) reported that amplification of the miR‐26a‐2 gene in 12% of high‐grade gliomas was associated with overexpression of miR‐26a. Thus, although many miRNAs are post‐transcriptionally regulated, alterations at the DNA level may also be the cause of abnormal expression of some miRNAs. In the present study, we provide evidence that genetic alterations of the miRNA genes are frequent in medulloblastomas. More than two thirds of the medulloblastomas analyzed showed genetic alterations (mostly amplification or deletion) in at least one of the nine miRNA genes analyzed.

Missense mutations in miRNA genes in human cancer appear to be very rare, and if occur, they are mostly located in precursor miRNA genes 10, 15, 42, 43. While some miRNA mutations do not exert detectable effects on miRNA function (10), one study has shown a significant reduction of the expression of miRNA let‐7e in a prostate cancer carrying a G to A transition at 19 nt downstream of the miRNA let‐7e gene (41). In the present study, a point mutation in the miR‐33b gene was detected in a medulloblastoma. The mutation was located in the sequence encoding precursor miRNA, and DNA extracted from the normal brain tissue surrounding this medulloblastoma was the wild‐type sequence, indicating that the mutation is somatic.

While MYCC overexpression is considered to be common in medulloblastoma, the frequency varies considerably according to different studies 11, 19, 37. Using in situ hybridization, Herms et al (19) showed MYCC overexpression in 46% of classic medulloblastomas and in 23% of desmoplastic medulloblastomas. A similar study by Eberhart et al using in situ hybridization (11) showed that MYCC is overexpressed in 31% of all medulloblastomas analyzed and that the frequency of MYCC overexpression was significantly higher in large‐cell/anaplastic medulloblastoma (52%) than in classic (12%) or desmoplastic medulloblastomas (14%). Using real‐time quantitative RT‐PCR, Rutkowski et al (37) reported that 53 of 101 cases (52%) of medulloblastomas had MYCC overexpression. With real‐time RT‐PCR using primers that exclude the possibility of false‐positive results caused by contaminated DNA (see Materials and Methods), we showed that MYCC is overexpressed in approximately 30% of medulloblastomas in the present study.

It has been reported that several miRNAs such as miR‐24 and miR‐145 regulate MYCC (5). Other miRNAs affect expression of MYCC regulators, which may indirectly affect MYCC expression (5). The interplay of MYCC and miRNAs appears to be highly complex, as MYCC also reprograms the expression of several miRNAs (6). In the present study, we selected nine miRNA genes on the basis of the presence of potential target sequences within the 3′‐untranslated region of the MYCC mRNA as well as on the basis of chromosomal loci reported to be deleted or amplified in medulloblastomas according to an array‐CGH database (http://www.progenetix.net/progenetix/). We observed an association between deletion of the miR‐512‐2 gene and overexpression of MYCC. Antisense‐based knockdown of miR‐512‐2 resulted in significant upregulation of MYCC expression at the mRNA and protein levels in HeLa and A549 cells, while forced overexpression of miR‐512‐2 in medulloblastoma/PNET cell lines DAOY, PFSK and UW‐228‐2 resulted in clear downregulation of MYCC protein. Furthermore, the results of luciferase reporter assays suggested that miR‐512‐2 targets the MYCC gene. However, miR‐512‐2 may target many other genes that can directly or indirectly affect the expresson of MYCC, and MYCC is also likely to be regulated by many other factors, whether genes or miRNAs.

It is noteworthy that both deletion and amplification were detected in several miRNA genes, that is, miR‐135b, miR‐135a‐2 and miR‐33b (Table 2). These results are in agreement with combined data on >350 medulloblastomas using array CGH (http://www.progenetix.net/progenetix/), which show that chromosomal locus at 17p11.2 (locus of miR‐33b) showed both loss (18%–25%) and gain (30%–32%) (27), locus at 1q32.1 (miR‐135b) showed both loss (19%) and gain (2%), and that at 12q23.1 (miR‐135a‐2) showed both loss (8%) and gain (2%) in medulloblastomas. Similarly, gain or loss of several miRNA genes (miR‐132, miR‐212 and miR‐22) has been detected in ovarian cancer (44). Of 57 miRNA genes analyzed in ovarian cancer cell lines, 23 (40%) showed both gains and losses in DNA copy number (44). These suggest that the consequences of alterations in miRNAs may be highly complex; both overexpression and downregulation may cause abnormal expression of multiple target genes associated with the pathogenesis and progression of tumors.

In summary, genetic alterations involving deletion or amplification of miRNA genes are frequent, and may play a role in the pathogenesis of medulloblastoma. Deletion of the miR‐512‐2 gene was observed in one third of medulloblastomas, and was associated with overexpression of MYCC. In vitro study further demonstrated that miR‐512‐2 targets MYCC expression in tumor cells. These results suggest that alterations in the miRNA genes that target MYCC may be an alternative mechanism leading to overexpression of MYCC in medulloblastomas.