MiR-22 is frequently downregulated in medulloblastomas and inhibits cell proliferation via the novel target PAPST1

Abstract

Medulloblastoma is the most frequent malignant central nervous system tumor in children. MicroRNAs (miRs) are small, non-coding RNAs that target protein-coding and non-coding RNAs, and play roles in a variety of cellular processes through regulation of multiple targets. In the present study, we analyzed miR-22 expression and its effect in cell proliferation and apoptosis in medulloblastomas. Quantitative reverse transcription PCR (RT-PCR) revealed significantly lower expression of miR-22 in 19 out of 27 (70%) medulloblastomas, D341, DAOY, ONS-76 medulloblastoma cell lines, compared with normal cerebellum. Forced expression of miR-22 by lentiviral vector transfection reduced cell proliferation and induced apoptosis, while knockdown of miR-22 increased proliferative activity in DAOY and ONS-76 cells. DAOY cells with miR-22 overexpression in nude mice yielded tumors smaller than those originated from control DAOY cells. Microarray analysis in DAOY cells with forced miR-22 expression showed significant changes in expression profiles, PAPST1 being the most significantly (10 folds) downregulated gene. Quantitative RT-PCR revealed PAPST1 mRNA upregulation in 18 out of 27 (67%) medulloblastomas. In addition, a luciferase reporter assay in ONS-76 and DAOY cells suggested that miR-22 directly targets the PAPST1 gene, and lentivirus-mediated knockdown of PAPST1 suppressed proliferation of DAOY and ONS-76 medulloblastoma cells. These results suggest that frequently downregulated miR-22 expression is associated with cell proliferation in medulloblastomas, and this may be at least in part via PAPST1, which is a novel target of miR-22.

Keywords: PAPST1; apoptosis; cell proliferation; medulloblastoma; miR-22.

Figure 1

Relative miR‐22 expression in medulloblastomas (A) and D341, DAOY, ONS‐76 medulloblastoma cells (B) determined by quantitative RT‐PCR. Values represent the fold of miR‐22 mRNA relative to normal cerebellum set at 1 (n = 3, ± SD). Note that most (19 out of 27; 70%) medulloblastomas, all medulloblastoma cell lines analyzed show lower miR‐22 expression (<20%) compared with normal cerebellum (P < 0.05). Representative fluorescent in situ hybridization (C). The left panel shows normal cells displaying a balanced 17p (orange) and RARA (green) signal. The right panel shows the loss of 17p signal in a medulloblastomas (C). Normal, normal cerebellum.

Figure 2

Laser scanning confocal microscope analysis 3 days after miR‐22 transfection with lentivirus, indicating high efficiencies and stable miR‐22 expression in DAOY and ONS‐76 medulloblastoma cells (A). The forced expression (B) and knockdown (C) of miR‐22 were confirmed by quantitative RT‐PCR (n = 3, ± SD, *P < 0.05, **P < 0.01). NC, negative control.

Figure 3

CCK‐8 assay showing that miR‐22 expression inhibits proliferation of DAOY cells (A), the effects was maximum at 72 h (n = 6, ± SD, *P < 0.05, **P < 0.01) and ONS‐76 cells (B), the effects was maximum at 48 h (n = 6, ± SD; **P < 0.01) compared with controls transfected with empty vector. The knockdown expression increased the proliferation of DAOY cells (C, n = 3, ± SD, *P < 0.05, **P < 0.01) and ONS‐76 cells (D, n = 3, ± SD, *P < 0.05, **P < 0.01).

Figure 4

Flow cytometric assay stained with AnnexinV‐PE/7‐AAD, showing increased fractions of apoptosis in DAOY and ONS‐76 cells transfected with miR‐22 (A). Mean fractions of apoptosis in DAOY (B) and ONS‐76 (C) cells. The values represent apoptotic cells (n = 5, ± SD, **P < 0.0001 for DAOY cells; *P < 0.02 for ONS‐76 cells).

Figure 5

Western blot shows induction of expression of cleaved caspase‐3 in DAOY and ONS‐76 cells transfected with miR‐22 (upper panel). Quantification of Western blot data (n = 4, ± SD), showing that significantly higher cleaved caspase‐3 expression in DAOY and ONS‐76 cells transfected with miR‐22 (*P < 0.05) (lower panel).

Figure 6

Representative images of nude mice 8 weeks after inoculation of DAOY cells transfected with miR‐22, showing that miR‐22‐expressed tumors grow more slowly than those with empty vectors (A). Quantification of volume of tumors in nude mice by caliper measurements, indicating that tumors from DAOY cells transfected with miR‐22 are significantly smaller than those transfected with empty vector. The bar represent the average volume of the tumor in each group (n = 5, ± SD; *P < 0.04) (B). Transmission electron microscopy reveals apoptotic DAOY cells (indicated by arrow) transfected with miR‐22. The condensed chromatin block is located in the middle of the nucleus and abutted on the nucleus membrane. The cell membrane and mitochondria are intact (C). Quantitative RT‐PCR showing that tumors originating from miR‐22‐expressing DAOY medulloblastoma cells in nude mice have significantly higher miR‐22 expression (n = 3, ± SD; *P < 0.04) (D).

Figure 7

Quantitative RT‐PCR, showing PAPST1 overexpression in medulloblastomas (A) and D341, DAOY, ONS‐76 medulloblastoma cells (B). Forced expression of miR‐22 significantly reduces PAPST1  mRNA expression in DAOY and ONS‐76 medulloblastoma cells (n = 3, ± SD; **P < 0.01) (C). Western blot showing PAPST1 downregulation in DAOY and ONS‐76 medulloblastoma cells (n = 3, ± SD; *P < 0.05) (D).

Figure 8

Quantitative RT‐PCR showing that tumors originating from miR‐22‐expressing DAOY medulloblastoma cells in nude mice show significantly lower PAPST1  mRNA expression (n = 3, ± SD; **P < 0.01) (A) than those originating from control DAOY cells. Western blot analysis showed similar observations at the protein level (n = 3, ± SD; **P < 0.01) (B).

Figure 9

PAPST 1 immunohistochemistry. No immunoreactivity was observed in normal cerebellum (A) or in IgG negative control in medulloblastomas (B). Cytoplasmic PAPST1 expression in medullobalstoma cells (C,D).

Figure 10

Summary of miR‐22 putative target binding sites in 3' Untranslated Regions (UTR) of PAPST 1, predicted by FindTar3 Online Prediction (A). Construction of the luciferase reporter containing the full‐length wild‐type (WT) 3' UTR or 3' UTR sequence with mutations at the two putative binding sites (MUT) of the PAPST 1 gene (B). Luciferase reporter assay confirmed that PAPST 1 is negatively regulated by the interaction between miR‐22 and complementary sequences in the 3'UTR of PAPST 1 in DAOY and ONS‐76 medulloblastoma cells (C,D).

Figure 11

Lentivirus shRNA‐mediated knockdown of PAPST 1. The knockdown expression by miR‐22 was determined by quantitative RT‐PCR in DAOY and ONS‐76 medulloblastoma cells (A) (n = 3, ± SD, *P < 0.05, **P < 0.01). Reduced expression was also detected at the protein level by Western blot (B) (n = 3, ± SD; **P < 0.01). In the CCK‐8 assay, lowered expression of PAPST1 led to decreased cell proliferation in DAOY (C) and ONS‐76 (D) medulloblastoma cells (n = 3, ± SD, *P < 0.05, **P < 0.01). NC, negative control shRNA.