MicroRNA profiling of peripheral nerve sheath tumours identifies miR-29c as a tumour suppressor gene involved in tumour progression

Abstract

Background: Neurofibromatosis type 1 is one of the most common familial diseases, the hallmark of which is the development of multiple neurofibromas. These are benign nerve sheath tumours, which can transform into malignant peripheral nerve sheath tumours (MPNST).

Methods: The aim of this study was to identify differentially expressed microRNA (miRNA) in neurofibromas and MPNST obtained from patients with neurofibromatosis type 1 using microarray analysis. Differential expression was validated by reverse transcription quantitative-PCR, and functional studies were performed after transfection of miRNA oligonucleotide mimics into MPNST cells.

Results: Sixteen miRNA were significantly differentially expressed in MPNST compared with NF, and of these fourteen were downregulated in MPNST: these included miR-30e*, miR-29c*, miR-29c, miR-340*, miR-30c, miR-139-5p, miR-195, miR-151-5p, miR-342-5p, miR-146a, miR-150, miR-223, let-7 a and let-7 g with a false discovery rate of q=8.48E-03 for the least significant miRNA. In contrast, miR-210 and miR-339-5p were upregulated in MPNST compared with neurofibromas. Prediction softwares/algorithms identified a list of genes targeted by miR-29c including extracellular matrix genes and matrix metalloproteinase (MMP)-2, all of which are reported to be involved in cell migration and invasion. Functional studies in a MPNST cell line, sNF96.2, using a mimic of the mature miR-29c showed reduced invasion, whereas there was no change in proliferation. Zymography of the manipulated cells showed that MMP2 activity was also reduced when miR-29c expression was forced in sNF96.2.

Conclusion: We provide evidence that reduction of miR-29c has a pivotal role in the progression of nerve sheath tumours and results by increasing the invasive/migratory properties of nerve sheath tumours.

Figure 1

MicroRNA profiling of neurofibromas and MPNST. (A) Multidimensional scaling of the 10 NF (light grey) and 9 MPNST (dark grey). (B) MicroRNA expression profiling and hierarchical clustering analysis of 10 neurofibromas and 9 MPNST representing the eight most significant differentially expressed miRNA with a cutoff of q=8.48E-03 for the least significant miRNA. Each row represents the relative levels of expression of a single miRNA and each column shows the expression levels for a single sample. (C) Heat-map of the miR-29 family members in 10 neurofibromas and 9 MPNST samples. All miR-29 family members were significantly differentially expressed with a P<0.05. The yellow and orange colours indicate relatively high and low expression, respectively.

Figure 2

miR-29c target genes. (A) sNF96.2 cells were transfected with either a miR-29c mimic or scrambled mimic in six-well plates. Two days post-transfection, cells were studied for expression of selected genes by RT–qPCR. Target genes expression is given as fold difference in sNF96.2 re-expressing miR-29c compared with the expression in the scrambled mimic cells. Bars represent mean and s.d. of three independent experiments relative to control (scrambled mimic). *P<0.01 and **P<0.001 to control. (B) Heat-map of selected target genes from our previously published GEM (Henderson et al, 2005). Expression levels are displayed in a colour spectrum that extends from light green (low expression) to bright red (high expression).

Figure 3

miR-29c target genes involved in migration/invasion. sNF96.2 cells were transfected with either a miR-29c mimic, scrambled mimic, or oligofectamine only in six-well plates. (A, B) Two days post-transfection, cells were detached and ∼25 000 cells were transferred to individual wells of transwell plates. (A) Representative photomicrographs of cells transfected with miR-29c, scrambled mimic, and oligofectamine only on membranes. The cells were fixed and stained with gelRed after 6 h incubation. Migrated cells on the membrane were counted in four fields where the cells were evenly distributed. (B) The graph shows the average number of cells from each field (n=4) counted in three independent experiments (total n=12), the bars represent the mean. *P<0.05 miR-29c mimic vs control (scrambled mimic) and was calculated in GraphPad Prism doing paired t-test. (C) Representative images of the scratch assay 2 days post-transfection with miR-29c mimic and scrambled mimic. The images were monitored by time-lapse video-microscopy for ∼60 h. Photomicrographs were taken at time 0 (2 days post-transfection) and time 59:45 h.

Figure 4

Proteolytic activity in sNF96.2 cells expressing miR-29c. (A) The supernatants from cells transfected with miR-29c and scrambled mimics were analysed by gelatin zymography in 0.1% gelatin–10% acrylamide gels with non-reducing conditions. Zones of gelatinolytic activity were revealed by negative staining. A different pattern of bands was observed when the cells were transfected with the miR-29c mimic and the scrambled mimic. A molecular weight ladder was used to identify, which MMPs was present or absent. Based on the molecular weight and the type of substrate (gelatin) used we concluded that the bands represented MMP9, MMP2, and MMP3 in the sNF96.2 cell line when transfected with the scrambled mimic, but that two of these corresponding to MMP2 and MMP9 were absent when cells were transfected with the miR-29c mimic. (B) Western blot analysis and RT–qPCR showed a reduction in MMP2 protein and mRNA expression in cells transfected with miR-29c mimic compared with control. In contrast MMP3 and MMP9 that are known not to be direct targets of miR-29c did not reveal a change in protein expression, and protein and mRNA expression, respectively.