Article Figures & Data
Figures
- Figure 1.
Global identification of differentially expressed miRNAs between gastric normal tissues and tumors. A, identification of differentially expressed miRNAs. Top, expression heat map showing expression of 80 differentially expressed miRNAs between and 40 normal gastric tissues (gray bar) and 40 GCs (blue and yellow bars; FDR < 0.01, using SAM, see Materials and Methods). Bottom, differential expression of miRNAs between intestinal-type and diffuse-type GCs (Int-GC and Diff-GC, respectively) analyzed from a separate SAM run (FDR < 0.01). B, heat map showing differential expression of 16 candidate tumor-suppressor miRNAs. Expression of miR-486 is highlighted in red type.
- Figure 2.
Expression of candidate tumor-suppressor hsa-miR-486-5p in gastric tumors and cell lines. A, genomic location of miR-486. miR-486 is located on chromosome 8p11 within the ANK1 gene, between exons 41 and 42. miR-486 is transcribed from the same strand from an alternative promoter located in intron 40 of the ANK1 gene. B, reciprocal expression of miR-486 compared with 3 previously reported oncogenic miRNAs (mir-17, mir-21, and mir-27a) across gastric normal tissues and GCs. C, quantitative PCR (qPCR) analysis showing relative expression of miR-486 in 29 primary GC tissues compared with matched adjacent normal tissues. Quantifications were measured by TaqMan real-time PCR. Each column represents an individual tumor/normal pair. Fold changes (tumor/normal) were transformed to log 2 values (y-axis). P values denotes the significance of fold change observed. D, expression of miR-486 and mir-17 in primary gastric tissues (normals and tumors) and GC cell lines. The x-axis depicts primary GCs (first column, n = 40, median), 15 GC cell lines, and primary normal gastric tissues (blue arrow and box, last column, n = 40, median). Expression of miR-486 is in green whereas mir-17 is in pink (color legend). Error bars indicate respective standard deviations across tumors or normals. R denotes the Pearson correlation between the 2 miRNAs. P values denoting the significance of the correlation coefficient R. Cell lines selected for functional analysis are highlighted in red (YCC3, AGS, SCH, YCC6).
- Figure 3.
miR-486 expression modulates multiple pro-oncogenic traits. A, restoration of miR-486 expression in GC cell lines. YCC3, SCH, and AGS cells were transfected with control (Ctr) or miR-486 precursors, qRT-PCR was carried out to assess relative miR-486 expression. B, miR-486 expression suppresses cellular proliferation in GC cell lines. miR-486–transfected cell lines were assessed for cell proliferation at 24, 48, and 72 hours posttransfection. Triplicate experiments were conducted for each set. *, P < 0.05, t test; points, mean; bars, SD. C, miR-486 expression suppresses anchorage-independent cell growth. Left, soft-agar colony formation assay reduction showing reduced in colony numbers in SCH and YCC3 stably transfected with miR-486 or empty vector control. Right, the quantification of colonies observed. Triplicate experiments were conducted for each cell lines. Columns, mean; bars, SD; *, P < 0.05, t test. D, miR-486 reduces motility and invasiveness of AGS cells. Top, migration of AGS cells stably expressing miR-486 or vector controls, measured by a Transwell migration assay. Columns, mean; bars, SD; *, P < 0.05. Bottom, invasion of AGS cells stably expressing miR-486 or vector controls, measured by a Matrigel assay. Insert, phase contrast microscopy of the crystal-violet stained cells in Matrigel. Columns, mean; bars, SD; **, P = 0.059. E, inhibition of miR-486 expression promotes cellular proliferation in YCC6. Top, miR-486 inhibitor and negative control inhibitor [anti–miR-486 (red) and anti–miR-ctr (blue)] transfected cells were assessed for miR-486 expression by qRT-PCR. Bottom, cell proliferation levels were assayed at 24 and 48 hours posttransfection. Triplicate experiments were conducted for each set. Points, mean; bars, SD; significance for the difference in growth denoted by *, P < 0.05.
- Figure 4.
Genomic Loss of miR-486 in primary GCs. A, recurrent genomic loss of miR-486. Red bars highlight the miR-486 locus. Top, genomic location of miR-486 on Chr 8p11.21. Middle, histogram showing frequency of genomic loss in this region across 106 primary GCs. Twenty-eight percent of GCs (30 tumors) are observed to exhibit loss of the miR-486 locus (blue horizontal bars below histogram). Genomic loss of the miR-486 locus in individual samples. Only samples with miR-486 loss are shown. The color gradient depicts the extent of copy number deletion. Bottom, genome browser view of the Chr 8p11.21 region showing miR-486 and adjacent genes such as ANK1. B and C, focal deletion of miR-486 in 2 gastric tumor samples: 2000088 and 990187. The copy number log-ratio data are shown together with segments identified. The miR-486 locus (with focal deletion) is highlighted in red.
- Figure 5.
Direct regulation of Olfactomedin-4 (OLFM4) by miR-486. A, miR-486 target prediction by miRanda v3.0 and TargetScan 5.1 algorithms. Seventeen targets were common to both prediction programs (genes in box). OLFM4 is highlighted in red. B, OLFM4 is highly expressed in primary gastric tumors compared with matched normal tissues (P < 0.001). C, OLFM4 and miR-486 in intestinal-type primary GCs and matched normals. We observed a significant negative correlation of miR-486 to OLFM4 expression (R = −0.61852; P = 0.003106) in 11 intestinal-type GCs and 7 matched normal tissues analyzed. D, OLFM4 protein levels are regulated by miR-486. Western blot analyses of OLFM4 protein in cells transfected with miR-486 or negative control mimics. OLFM4 protein levels ware lower in YCC3 cells expressing miR-486 as compared with control-miR–expressing cells. Similar reductions in OLFM4 protein were also observed in miR-486–expressing AGS cells. Quantitative real-time PCR showed that OLFM4 transcripts are also reduced in cells transfected with miR-486 as compared with controls. (*, t test for OLFM4 expression in miR-486 vs. control cells). E, OLFM4 is a direct target of miR-486. The predicted miR-486 target region found in the OLFM4 mRNA 3′-UTR was cloned downstream of luciferase in a pMIR-Report-luciferase reporter vector. Reporter constructs were cotransfected with miR-486 and negative control mimic molecules into AGS cells. Luciferase reporter assays were normalized to β-galactosidase activities and experiments were conducted in triplicates. Data were plotted after normalized against the negative control miRNA mimics. Columns, mean; bars, SD. F, silencing of OLFM4 in GC cells YCC3 and SCH by siRNA reduces cell proliferation capacity. YCC3 and SCH cells were transfected with siRNAs against OLFM4 or scrambled controls (scr, negative control). Experiments were conducted in triplicates. y-Axis denotes the absorbance at 490 nm (cell proliferation) and x-axis is the assay time-points. Points, mean; bars, SD; significance for the difference in growth is denoted by **.
Tables
- Table 1.
High OLFM4 expression is associated with intestinal-type GC
Lauren's classification Intestinal Diffuse Mixed Total OLFM4 low expression 37 42 8 87 OLFM4 high expression 52 24 11 87 Total 89 66 19 174 NOTE: A Pearson χ2 statistic was used to test the association of the expression of OLFM4 with GC histologic subtype (174 tumors). High OLFM4 expression was positively associated with intestinal-type GC (Pearson χ2 test, P = 0.019).
Supplementary Data
Supplementary Figures S1-S4; Supplementary Table S1.
Files in this Data Supplement:
- Supplementary Data - Supplementary Figures S1-S4; Supplementary Table S1.
Volume 17, Issue 9
