Patient Characteristics
In total, 50 meningioma patients were recruited in this study. There were WHO grade I 30(60%), WHO grade II 15 (30%), and WHO grade III 5 (10%). There were 17 males and 33 females, aged 24-76 years, with an average age of (56.82±11.58) years.
Expression and correlation of miR-153-5p and GAB1 in normal meninge and meningioma
Since our previous study showed that GAB1 3'UTR gene polymorphism was associated with the risk of meningioma, and we predicted the direct interaction between miR-153-5p and GAB1 3'UTR on the TargetScan website (http://www.targetscan.org).In order to further explore the role of GAB1 and miR-153-5p in meningioma, we detected the expression of GAB1 and miR-153-5p in meningioma.
In contrast to normal meninge tissues, meningioma tissues expressed remarkably lower miR-153-5p and elevated GAB1 mRNA (Figure 2A1, A3). In comparison with WHO grade I, there were also lower miR-153-5p and higher GAB1 expression in WHO grade II and WHO grade III. However, there is no difference between the latter two(Figure 2A2, A4). The immunohistochemistry results showed that the expression of GAB1 was mainly found in the cellular cytoplasm (Figure 2, B1-4). Immunohistochemical staining (++) of 28 cases (56%) and immunohistochemical staining (+++) of 7 cases (14%) of meningioma tissue group; 1 cases of normal dura mater immunohistochemical staining (++) cases (10%) and immunohistochemical staining (+++) in 0 case (0%). The number of GAB1 positive cells in the meningioma tissue group was higher than that in the normal dura mater group (P<0.0001) (Figure2 C). It was exhibited that the relative expression of miR-153-5p in the tissues of meningioma patients was negatively linked with the relative expression of GAB1 mRNA by Pearson correlation (r=-0.521, P<0.001), which could be considered that both of them had a moderate degree of correlation(Figure 2D). The ROC curves were drawn (Figure2 E-F) and the results were compared as follows: The ROC curves of miR-153-5p and GAB1 were remarkably different (P < 0.001) in meningioma, both of them had higher accuracy for meningioma diagnosis.
Using the Mann-Whitney U test, the results showed that the dura mater invasion group had down-regulated miR-153-5p mRNA expression and up-regulated GAB1 mRNA expression compared with the dura mater non-invasion group, (all P < 0.05, Figure 3A,B). There were also significant differences in tumor grade, (all P < 0.05, Figure 2A2,B4). However, there was no significant correlation between the mRNA expression of both and the patient's age, gender, and peritumoral edema (P>0.05). Binary logistic regression was used to assess whether both are risk factors for dural invasion and tumor grade, both miR-153-5p downregulation and GAB1 upregulation were risk factors for dural invasion and tumor grade (all P < 0.05,Table 1 ).
Table 1 Correlations of miR-153-5p and GAB1 mRNA expression with clinical characteristics in meningioma patients
clinical
characteristics
|
case
|
miR-153-5p
|
P
|
GAB1 mRNA
|
P
|
Age (years)
|
|
|
|
|
|
≥50 vs <50
|
38 vs 12
|
1.39(0.29±6.60)
|
0.680
|
0.85(0.66±1.10)
|
0.224
|
Gender
|
|
|
|
|
|
Male vs female
|
19 vs 31
|
1.38(0.35±5.50)
|
0.647
|
0.95(0.77±1.17)
|
0632
|
Tumor size(cm)
|
|
|
|
|
|
d≤4.5 vs d>4.5
|
25 vs 25
|
0.54(0.14±2.12)
|
0.376
|
0.99(0.81±1.20)
|
0.899
|
Peritumoral edema
|
|
|
|
|
|
Yes vs No
|
30 vs 20
|
1.47(0.37±5.81)
|
0.582
|
1.10(0.90±1.35)
|
0.361
|
Dural invasion
|
|
|
|
|
|
Yes vs No
|
20 vs 30
|
6.17(1.20±31.82)
|
0.030
|
0.76 (0.61±0.95)
|
0.017
|
Tumor grade
|
|
|
|
|
|
WHOⅠvs WHOⅡ +Ⅲ
|
30 vs 20
|
0.01(0.00±0.16)
|
0.001
|
1.83(1.32±2.54)
|
0.000
|
Stratified analysis of miR-153-5p, GAB1 mRNA, and protein expression differences according to different genotypes
Under the optimal model (recessive model), stratified analysis was performed according to the genotypes of GAB1 4 SNPs (rs3805246, rs3828512, rs3805236, and rs1397529), respectively. miR-153-5p and GAB1 mRNA expression levels were compared in different genotype subgroups. The results were not significantly different among them(Figure 4).
Similarly, GAB1 protein expression level was compared in different genotype subgroups. We found there was a statistically significant in GAB1 protein expression at rs1397529 (AC vs AA). The protein expression level of GAB1 is up-regulated in meningioma patients with AC genotype compared with patients with AA genotype(P<0.05). There were no statistical differences in other sites(P>0.05). The result was shown in Table 2.
Table 2 Expression of GAB1 protein stratified analysis according to different genotypes
SNPs
|
Genotype
|
GAB1 protein
|
Z
|
P
|
-
|
+
|
++
|
+++
|
rs3805246
|
GG+AG
|
3
|
8
|
23
|
6
|
-0.607
|
0.544
|
AA
|
0
|
4
|
5
|
1
|
rs3828512
|
AA+ AG
|
3
|
9
|
23
|
6
|
-0.183
|
0.855
|
GG
|
0
|
3
|
5
|
1
|
rs3805236
|
AA+AG
|
3
|
10
|
21
|
6
|
-0.472
|
0.637
|
GG
|
0
|
2
|
7
|
1
|
rs1397529
|
AC
|
1
|
1
|
8
|
1
|
-2.264
|
0.024
|
AA
|
2
|
11
|
20
|
6
|
Transfection efficiency of miRNA and siRNA
The transfection results indicated that FAM NC (fluorescein amidite negative control) transfected IOMM-Lee cells, and the transfection efficiency was higher after 6 hours of transfection, and the highest transfection efficiency was about 70%. The optimal transfection conditions were: FAM NC 2 μL: Lipo2000 2 μL (Figure 5A). In comparison to the NC groups, miR-153-5p contents were remarkably increased in IOMM-Lee cells inoculated with miR-153-5p mimics (P <0.001). By contrast, miR-153-5p expressions revealed that the miR-153-5p mimic and inhibitor were very efficient at transfection and so appropriate for further experiments (Figure 5B). Total RNA along with protein was extracted 48 hours after si-GAB1 transfection. RT-qPCR analysis revealed that si-GAB1 cell transfects had a greater than 60% reduction in GAB1 mRNA expression (P <0.01). This established that the siRNA was capable of successfully inhibiting GAB1 expression and could be employed in further experiments (Figure 5C).
GAB1 was a Direct Target of miR-153-5p
Subsequently, we further screened the downstream target of miR-153-5p through the database in TargetScan( http://www.targetscan.org), which predicted that miR-153-5p could bind with four positions of the 3'UTR of GAB1 (Figure 6A). Because previous investigations have found that the gene polymorphism of GAB1 is related to meningioma susceptibility, and the SNP rs1397529 is located in the 3'UTR, the site-directed mutation was used firstly to introduce base substitution to this miR-153-5p binding site of luciferase reporters to produce mutated plasmids using Site-Directed Gene Mutagenesis Kit (Beyotime).
There was no significant difference in luciferase activity between the GAB1 3'UTR wild-type group or mutant group and the NC group(P>0.05,Figure 6B). Interestingly, when all four sites predicted to target miR-153-5p binding to the GAB1 3'UTR were modified with complementary mutations, miR-153-5p was able to regulate luciferase activity in the wild-type GAB1 3UTR group(P<0.001,Figure 6C), and the regulatory relationship disappeared when the binding site was mutated(P<0.0001,Figure 6C).
miR-153-5p inhibits malignant meningioma cells proliferation, migration and invasion by targeting GAB1
CCK-8 analysis showed that compared to the NC group, GAB1 siRNA, miR-153-5p overexpression, and GAB1 siRNA+miR-153-5p inhibitor significantly inhibited cell proliferation, while miR-153-5 silencing significantly promoted cell proliferation(Figure 7A).The results of the Transwell assay showed that compared with the NC group, GAB1 siRNA, miR-153-5p overexpression, and GAB1 siRNA+miR-153-5p inhibitor significantly reduced cell invasion, while miR-153-5 silencing inhibited cell invasion (Figure 7B, 7C).According to the results of cell scratch migration assay, compared with the NC group, GAB1 siRNA, miR-153-5p overexpression, and GAB1 siRNA+miR-153-5p inhibitor significantly inhibited cell migration. However, miR-153-5 silencing significantly promoted cell migration migration (Figure 7D, 7E).Both miR-153-5p overexpression and GAB1 silencing could inhibit the proliferation, migration and invasion of malignant meningioma cells, and GAB1 silencing partially blocked the effect of miR-153-5p silencing on proliferation, migration and invasion of malignant meningioma cells, which suggested that miR-153-5p might inhibit malignant meningioma cell proliferation, migration, and invasion by targeting GAB1.
miR-153-5p exerts its functions by inhibiting the GAB1- MEK/ERK axis.
We examined the expression of the signature genes of the GAB1-MEK/ERK axis to further understand miR-153-5p's mechanism of action. Overexpression of miR-153-5p resulted in a significant decrease in GAB1 mRNA and protein expression in malignant meningioma cells (all P < 0.05,Figure 8A-C). Meanwhile, miR-153-5p inhibitor resulted in increased GAB1 mRNA expression, but did not reach statistical significance(P>0.05,Figure 8A).Instead, miR-153-5p inhibitor significantly increased GAB1 protein expression (P < 0.05,Figure 8B,8C). There was an overexpression of miR-153-5p, a decrease in the expression of GAB1 protein, and a significant decrease in the levels of ERK1/2, MEK1, and p-ERK1/2 protein expressions compared to the NC group(all P < 0.01,Figure 8B,8C). The expression of p-MEK1 protein was also decreased. The p-MEK1 protein expression also decreased, and the result was not statistically significant. After miR-153-5p was silenced, the expression of GAB1 protein was increased, and MEK1, ERK1/2, p-MEK1, and p-ERK1/2 expressions were also up-regulated at the same time, especially the expression of ERK1/2 reached statistical significance (P < 0.05,Figure 8B,8C). miR-153-5p inhibited expression of GAB1, thereby regulating activation of MEK/ERK pathways.