Targeting insulin-like growth factor-binding protein-3 by microRNA-125b promotes tumor invasion and poor outcomes in non-small-cell lung cancer

Abstract

Insulin-like growth factor-binding protein-3 acts as a tumor suppressor that inhibits the PI3K/AKT signaling pathway due to blocking insulin growth factor-1 binding to its receptor. We hypothesized that insulin-like growth factor-binding protein-3 might be targeted by microRNA-125b and promote tumor invasion and poor outcome in non-small-cell lung cancer via activation of the PI3K/AKT signaling pathway. Real-time polymerase chain reaction and immunohistochemistry were performed to determine the level of microRNA-125b, insulin-like growth factor-binding protein-3 messenger RNA, and phosphorylated-AKT expression in 105 tumors from non-small-cell lung cancer patients. Low insulin-like growth factor-binding protein-3 messenger RNA levels and positive phosphorylated-AKT expression were more commonly found in patients with high microRNA-125b tumors than low microRNA-125b tumors. A poorer overall survival and relapse-free survival were observed in patients with high microRNA-125b tumors than low-microRNA-125b tumors in p53-mutated patients, but not in p53-wild-type patients. Mechanistically, microRNA-125b promotes invasion ability in p53-mutated cells via the PI3K/AKT activation by targeting of insulin-like growth factor-binding protein-3, but this effect was not observed in p53-wild-type cells. An increase in phosphorylated-AKT expression due to targeting of insulin-like growth factor-binding protein-3 by microRNA-125b was responsible for cell invasion in p53-mutated cells. In conclusion, the microRNA-125b level promotes invasive ability in p53-mutated cells via PI3K/AKT activation by targeting of insulin-like growth factor-binding protein-3, thereby resulting in p53-mutated non-small-cell lung cancer patients with poor outcomes.

Introduction

MicroRNA (miRNA)-125b is located at chromosome 11q23-24 and its expression was associated with tumor progression in breast, ovarian, and lung cancer.^1–5 However, miRNA-125b plays a dual role in lung tumorigenesis. For example, miRNA-125b promotes tumor metastasis by targeting tumor protein p53-induced nuclear protein 1.⁴ Conversely, miR-125b may target metastasis-associated gene 1 to suppress tumor invasion and migration.⁶ Therefore, the role of miRNA-125b in lung tumorigenesis remains controversial.

P53 is a central regulator of apoptosis and proliferation. The dual roles of miRNA-125b in human tumorigenesis raised the possibility that genes in the p53 network might be regulated by miRNA-125b. Insulin-like growth factor-binding protein-3 (IGFBP3) is one of the 20 targets of miRNA-125b in the p53 network using by a gain- and loss-function screen.⁷ An early report showed that IGFBP is directly regulated by p53.⁸ MiRNA-125b inhibits p53 expression by directly binding with p53 messenger RNA (mRNA) 3′-untranslated region to inhibit p53-dependent cell apoptosis.⁹ However, the downregulation of miRNA-125b also results in apoptosis through p53-independent pathway in p53-mutated cells.^10,11 We therefore expected that the oncogenic and tumor suppressor role of miRNA-125b might depend on p53 mutational status via targeting of IGFBP3.

IGFBP3 is a member of the IGFBP family that binds insulin growth factor-1 (IGF1) and blocks its mitogenic and antiapoptotic actions.¹² In the absence of IGFBP3, IGF1 binds to the IGF1 receptor (IGF1R) to activate the PI3K/AKT signaling pathway.^13,14 The PI3K/AKT signaling pathway plays a crucial role in cell survival and proliferation.¹⁵ We therefore suggested that the low expression of IGFBP3 caused by miRNA-125b targeting might play a role in cell survival and proliferation via activation of the PI3K/AKT signaling pathway.

Our preliminary data from non-small-cell lung cancer (NSCLC) patients indicated a negative correlation between miRNA-125b and IGFBP3 mRNA expressions. Moreover, a small subset of NSCLC patients showed a shorter overall survival (OS) when their tumors showed high expression of miRNA-125b rather than low miRNA-125b expression in p53-mutated patients, but this effect was not observed in p53-wild-type (WT) patients. We therefore hypothesized that miRNA-125b might promote tumor invasion and poor outcome in p53-mutated NSCLC via targeting of IGFBP3 due to activation of the PI3K/AKT signaling pathway.

Materials and methods

Tumor specimens of NSCLC patients

Tumor tissues from 105 NSCLC patients were retrieved from the Department of Thoracic Surgery, Taichung Veterans General Hospital (Taichung, Taiwan), between 1998 and 2004, with the approval of the Institutional Review Board (Institutional Review Board, Chung Shan Medical University Hospital, CSMUH No: CS11177). The inclusion criteria for patients were as follows: primary diagnosed with lung cancer, no metastatic disease at diagnosis, no previous diagnosis of carcinoma, no neoadjuvant treatment before primary surgery, and no evidence of disease within 1 month of primary surgery. The tumour, node, and metastasis (TNM) stage; tumor type; and stage of each collected specimen were histologically determined according to the World Health Organization’s (WHO) classification system. Clinical parameters and OS data were collected from chart review and the Taiwan Cancer Registry, Ministry of Health and Welfare, Executive Yuan, R.O.C.

Sample preparation

Lung tumor specimens were collected by surgical resection, and surgically resected tissues were stored at −80°C at the Division of Thoracic Surgery, Taichung Veterans General Hospital. To achieve most accurate comparability of histological and molecular evaluations (H&E-stained tissues’ images), the frozen tissue samples were cut into 10-µm sections with a cryotom (LEICA Biosystems, Wetzlar, Germany). After 10 serial cuts, one tissue section was transferred to a glass slide and stained with H&E. Subsequently, a pathologist visually quantified the proportion of tumor compartments in the specimen. The tissue sample used in this study had a tumor cell percentage of over 75%.

Cell lines

All cell lines used in this study are NSCLC cells. The H1703, H358, H1355, CH27, H1975, A549, H1650, and H460 cell lines were purchased from the American Type Culture Collection; the CL1-5 and CL3 cells were kindly provided by Dr P.C. Yang (Department of Internal Medicine, National Taiwan University Hospital); PC9 cells were kindly provided by Dr James C.-H. Yang (Department of Clinical Medicine, National Taiwan University Hospital, Taiwan), and H157 cells were kindly provided by Dr J.Y. Chen (Institute of Biomedical Sciences, Academic Sinica, Taipei, Taiwan). The H1703, CL1-5, H358, H157, H1355, PC9, H1975, H1650, CL3, and H460 cells were cultured in RPMI-1640 medium (Hyclone, Waltham, MA, USA) supplemented with 10% fetal bovine serum (Hyclone). The CH27 and A549 cells were cultured in Dulbecco’s modified Eagle’s medium (Hyclone) with 10% fetal bovine serum. All cell lines were grown at 37°C in humidified air with 5% CO₂.

Plasmids and transfections

The IGFBP3 shRNA and pLKO.1 vectors were obtained from the National RNAi Core Facility (Academia Sinica, Taipei, Taiwan). The IGFBP3 complementary DNA (cDNA) clone (SC119779) was purchased from OriGene (Rockville, MD, USA). The constitutively active myrAkt delta4-129 (Plasmid #10841), control pECE (Plasmid #26453), and mutated p53 (R175H; Plasmid #16436) plasmids were purchased from Addgene (Cambridge, MA, USA). The p53-WT constructs were provided by Dr Jiunn-Liang Ko (Chung Shan Medical University, Taichung, Taiwan).¹⁶ These plasmids were transiently transfected into NSCLC cells (1 × 10⁶) using the Turbofect reagent (Formentas, Glen Burnie, MD, USA). After 48 h, the cells were harvested for subsequent experiments.

MiRNA-125b mimics and inhibitor transfection

Cells were grown to confluence in 6-well plates. The miR-125b mimics (10–50 nM/well; Ambion, Foster City, CA, USA), miRNA-125b inhibitors (10–100 nM/well; Ambion), and negative control (Ambion) cells were transfected using Lipofectamine 3000 transfection reagent (Invitrogen, Foster city, CA, USA) according to the manufacturer’s protocol. Transfection efficiency was evaluated by the real-time polymerase chain reaction (PCR).

Immunohistochemistry analysis

The immunohistochemical procedures and quantification methods were as described previously.¹⁷ Specimens were formalin fixed and paraffin embedded. Briefly, 3-µm sections were cut, mounted on glass, and dried overnight at 37°C. All sections were then deparaffinized in xylene, rehydrated through alcohol, and washed in phosphate-buffered saline. This buffer was used for all subsequent washes. Sections were heated in a microwave oven twice for 5 min in citrate buffer (pH 6.0). The primary antibody and the incubation time was 60 min at room temperature followed by a conventional streptavidin–peroxidase method (LSAB Kit K675; DAKO, Carpinteria, CA, USA). Signals were developed with 3, 3′-diaminobenzidine for 5 min and counter-stained with hematoxylin. Negative controls were obtained by leaving out the primary antibody. The intensities of signals were evaluated independently by three observers. Immunostaining scores were defined as the cell staining intensity (0 = nil, 1 = weak, 2 = moderate, and 3 = strong) multiplied by the percentage of labeled cells (0%–100%), leading to scores from 0 to 300. A score over 150 was rated as “high” immunostaining, whereas a score less than 150 was rated as “low.” The immunostaining results for phosphorylated (p)-AKT expression in lung tumors were obtained from previous reports.¹⁷

RNA isolation and real-time PCR

Total cellular RNA was extracted using TRI reagent according to the manufacturer’s instructions (Invitrogen). The purified total RNA was used as a template for the synthesis of cDNA using the High-Capacity cDNA Reverse Transcription Kits (Life Technologies, Foster City, CA, USA). The following primer sequences were used for amplification of the IGFBP3 gene: the forward primer, 5′- AAGACAGCCAGCGCTACAAAG -3′, and the reverse primer, 5′- TACGGCAGGGACCATATTCTG-3′. The expression level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control for mRNA. The real-time PCR was performed using Maxima SYBR Green/ROX qPCR Master Mix (Life Technologies), and reactions were run on a 7500 Real-Time PCR Systems (Life Technologies). The relative gene expression was analyzed using the 2^−(ΔΔCt) method and normalized to the control. All experiments were performed in duplicate three independent times. The median value (0.07, range: 0.001–2724.51) of these tumors was used to categorize patients with high or low IGFBP3 mRNA expression.

Real-time reverse transcription PCR analysis of miRNA-125b mRNA expression levels

DNase I-treated total RNA (10 ng) was subjected to miRNA PCR analysis with the TaqMan^® miRNA Reverse Transcription Kit (Life Technologies), miRNA assays (miRNA-125b: TM:000449, RNU6B: TM: 001093; Life Technologies), and a Real-Time Thermocycler 7500 (Life Technologies). RNU6B was used as the small RNA reference housekeeping gene. The median value (3.20, range: 0.01–1658.72) of these tumors was used to categorize patients with high or low miRNA-125b mRNA expression.

Boyden chamber assay

Cell invasion ability was analyzed using a Falcon^® cell culture insert for 24-well plate with 8.0-µm pore transparent PET membrane (BD Biosciences, Bedford, MA, USA). Cells (1 × 10⁵) were plated onto the upper compartment of chamber coated with Geltrex^®. The lower compartment was filled with complete medium. The chamber was incubated at 37°C for 20 h. Invaded cells on the bottom of inserts were fixed with 100% ethanol, stained with Giemsa, and observed under a microscope. The number of invaded cells was counted and averaged in three random microscopic fields for each sample group. All experiments were performed in three independent times.

Western blotting

Cells were harvested and lysed in radioimmunoprecipitation assay (RIPA) lysis buffer. Equal amounts of total protein from each sample were electrophoretically separated on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and then electrotransferred to polyvinylidene difluoride (PVDF) membranes. After blocking, the membranes were probed with primary antibodies. The antibodies were used against AKT (Phosphorylated Serine 473), IGFBP3 (GeneTex Inc., Irvine, CA, USA), and AKT (pan) (GeneTex Inc.). Then, the membrane was incubated with the appropriate secondary horseradish peroxidase-conjugated antibody. Bound antibody was visualized using the SuperSignal West Pico Chemiluminescent Substrate (Pierce Biotechnology, Rockford, IL, USA). β-actin (Sigma, St Louis, MO, USA) was used as the loading control. The ImageJ software (National Institutes of Health, USA) was used for western blot analyses.

Statistical analysis

The statistical analysis was performed using the SPSS statistical software program (Version 15.0 SPSS Inc., Chicago, IL, USA). The associations between miR-125b and IGFBP3, as well as p-AKT, were analyzed using a chi-square test. A multivariate Cox regression analysis was performed for both OS and relapse-free survival (RFS). The analysis was stratified for all known prognostic variables (age, gender, smoking status, tumor type, and stage of cancer) and for miR-125b expression.

Results

MiRNA-125b levels are associated with poor OS and RFS in NSCLC patients, particularly in p53-mutated patients

The 105 tumors from surgically resected NSCLC patients were enrolled to determine the levels of miRNA-125b and IGFBP3 using real-time PCR. The median value of miRNA-125b and IGFBP3 levels was used as a cutoff point to divide patients into “high” and “low” subgroups. High miRNA-125b expression more commonly occurred in non-squamous and p53-WT patients than in squamous and p53-mutated patients (65.9% vs 39.3%, p = 0.007 for tumor histology; 63.5% vs 37.7%, p = 0.008 for p53 status; Supplementary Table 1). However, miRNA-125b levels were not associated with clinical parameters of age, gender, smoking status, or stage in this study population.

We examined whether the prognostic value of miRNA-125b could be associated with p53 status. Kaplan–Meier analysis indicated that the OS and RFS periods were shorter in the high miRNA-125b tumors than in the low miRNA-125b tumors in all studied and p53-mutated patients (all studied patients: p < 0.001 for OS, p = 0.001 for RFS; p53-mutated patients: p = 0.001 for OS and RFS; Supplementary Figure 1(a) and (b)); however, no prognostic value of miRNA-125b was revealed in p53-WT patients (Supplementary Figure 1(c)). Cox regression analysis further indicated that the prognostic significance of miRNA-125b level on OS and RFS was still observed in all studied and p53-mutated patients (all studied patients: p < 0.001 for OS and RFS; p53-mutated patients: p = 0.001 for OS, p = 0.002 for RFS; Table 1). These results showed that miRNA-125b level may be associated with poor OS and RFS in NSCLC patients, particularly in p53-mutated patients.

Table 1.

Cox regression for the prognostic value of miRNA-125b and various parameters on OS and RFS.

Variables	OS				RFS
	CaseNo.	AdjustedHRa	95% CI	p-value	CaseNo.	AdjustedHRa	95% CI	p-value
All studied patients
MiRNA-125b
Low	52	1			52	1
High	53	3.066	1.736–5.415	<0.001	53	2.460	1.493–4.055	<0.001
Age (years)
≤65	37	1			37	1
>65	68	1.159	0.668–2.010	0.600	68	1.155	0.704–1.893	0.568
Gender
Female	19	1			19	1
Male	84	0.828	0.374–1.833	0.642	84	1.181	0.566–2.464	0.657
Smoking status
Nonsmokers	44	1			44	1
Smokers	61	1.504	0.799–2.830	0.206	61	1.117	0.632–1.975	0.702
Tumor type
Non-squamous	44	1			44	1
Squamous	61	1.297	0.744–2.26	0.359	61	1.200	0.724–1.987	0.479
Stage
I, II	54	1			54	1
III	49	3.253	1.868–5.667	<0.001	49	2.166	1.334–3.517	0.002
p53-mutated patients
MiRNA-125b
Low	33	1			33	1
High	20	3.682	1.666–8.137	0.001	20	3.220	1.516–6.843	0.002
Age (years)
≤65	19	1			19	1
>65	34	2.386	0.929–6.125	0.071	34	1.876	0.846–4.161	0.121
Gender
Female	7	1			7	1
Male	46	0.565	0.151–2.106	0.395	46	1.509	0.409–5.557	0.537
Smoking status
Nonsmokers	15	1			15	1
Smokers	38	1.458	0.538–3.948	0.458	38	0.645	0.235–1.767	0.394
Tumor type
Non-squamous	16	1			16	1
Squamous	37	1.018	0.384–2.697	0.972	37	0.818	0.349–1.918	0.644
Stage
I, II	29	1			29	1
III	24	2.189	0.989–4.842	0.053	24	2.170	1.051–4.480	0.036

OS: overall survival; HR: hazard ratio; RFS: relapse-free survival; CI: confidence interval.

a

HR adjusted for age, gender, smoking status, and stage.

MiRNA-125b levels are negatively correlated with IGFBP3 mRNA, but positively related with p-AKT expression in lung tumors from NSCLC patients

We examined the possibility that a decrease in IGFBP3 could be responsible for the miRNA-125b-mediated poor prognosis in NSCLC patients due to activation of the PI3K/AKT signaling pathway. This possibility was tested by evaluating the IGFBP3 mRNA levels and p-AKT expression in lung tumors using real-time PCR and immunohistochemistry, respectively. Tumors with high levels of IGFBP3 mRNA were frequently associated with low miRNA-125b levels (75.0% vs 25.0%), whereas tumors with low levels of IGFBP3 mRNA were associated with high miRNA-125b levels (73.6% vs 26.4%, p < 0.001; Table 2). The reverse correlation between IGFBP3 and miRNA-125b was observed in p53-mutated and p53-WT patients (p < 0.001 for p53-mutated, p = 0.008 for p53-WT). A positive correlation of p-AKT with miRNA-125b level was observed in all studied and p53-mutated patients, but not in p53-WT patients (p = 0.001 for all studied, p < 0.001 for p53-mutated; Table 2). However, the negative correlation of p-AKT expression with IGFBP3 mRNA levels was apparent only in the all studied and p53-mutated patients, not in the p53-WT patients (p = 0.025 for all studied and p53-mutated; Table 2). These results observed from NSCLC patients seemed to support the possibility that miRNA-125b may promote lung tumor invasion via activation of the PI3K/AKT signaling pathway through a targeting of IGFBP3.

Table 2.

Correlation of miR-125b with IGFBP3 and p-AKT expression and their each correlation in lung cancer patients.

	No.	IGFBP3			p-AKT
		Low	High	p-value	Low	High	p-value
All studied patients
MiR-125b
Low	52	13 (25.0%)	39 (75.0%)	<0.001	35 (67.3%)	17 (32.7%)	0.001
High	53	39 (73.6%)	14 (26.4%)		19 (35.8%)	34 (64.2%)
IGFBP3
Low	52				21 (40.4%)	31 (59.6%)	0.025
High	53				33 (62.3%)	20 (37.7%)
p53-mutated patients
MiR-125b
Low	33	7 (21.2%)	26 (78.8%)	<0.001	25 (75.8%)	8 (24.2%)	<0.001
High	20	16 (80.0%)	4 (20.0%)		5 (25.0%)	15 (75.0%)
IGFBP3
Low	23				9 (39.1%)	14 (60.9%)	0.025
High	30				21 (70.0%)	9 (30.0%)
p53-WT patients
MiR-125b
Low	19	6 (31.6%)	13 (68.4%)	0.008	10 (52.6%)	9 (47.4%)	0.477
High	33	23 (69.7%)	10 (30.3%)		14 (42.4%)	19 (57.6%)
IGFBP3
Low	29				12 (41.4%)	17 (58.6%)	0.438
High	23				12 (52.2%)	11 (47.8%)

IGFBP3: insulin-like growth factor-binding protein-3; p-AKT: phosphorylated-AKT.

MiRNA-125b may promote cell invasion by targeting IGFBP3 in p53-mutated NSCLC cells, but not in p53-WT NSCLC cells

We used six p53-mutated and six p53-WT NSCLC cell lines to verify whether miRNA-125b levels were negatively correlated with IGFBP3 in p53-mutated cells, but not in p53-WT cells. Western blotting indicated that IGFBP3 protein expression was correspondent with its mRNA expression in the panel of NSCLC cells (Figure 1(a) upper panel). The IGFBP3 mRNA expression evaluated by real-time PCR in the p53-mutated cells was almost undetectable in the H1703 and CL1-5 cells that had high expression of miRNA-125b, but the expression of IGFBP3 mRNA gradually increased in the H358, H157, H1355, and PC9 cells that had lower expression of miRNA-125b. In general, a reverse correlation between IGFBP3 mRNA and miRNA-125b was still observed in six p53-WT cell lines except H460 cells (Figure 1(a) lower panel).

Figure 1.

MiRNA-125b promotes invasion ability in p53-mutated cells, but not in p53-WT cells. (a) The miRNA-125b and IGFBP3 mRNA levels in 12 lung cancer cell lines with mutant or wild-type p53 were examined by quantitative real-time PCR. The IGFBP3 protein levels in 12 lung cancer cell lines with mutant or wild-type p53 were examined by western blotting. The relative miRNA-125b expression levels were normalized to the H1703 cells; the relative IGFBP3 mRNA expression levels were normalized to the PC9 cells. (b) The miRNA-125b and IGFBP3 mRNA levels in H1703 (left panel) and PC9 (right panel) lung cancer cell lines were determined by quantitative real-time PCR after transfection with a miRNA-125b inhibitor or mimic, respectively. The relative miRNA-125b and IGFBP3 expression levels were normalized to the cells treated with microRNA inhibitor or mimic negative control. The significance was compared to microRNA inhibitor or mimic negative control. The IGFBP3 protein levels were evaluated by western blotting using a specific antibody. (c) The cell invasion ability in H1703 and PC9 lung cancer cell lines was analyzed by transwell assay after transfection with a miRNA-125b inhibitor or mimic, respectively. The relative invasiveness of the cells was normalized to that shown by the cells treated with a microRNA inhibitor or a mimic negative control. The significance was compared to the microRNA inhibitor or mimic negative control.

[Figure omitted. See PDF]

We examined whether IGFBP3 could be targeted by miRNA-125b in NSCLC cells. MiRNA-125b expression levels were decreased and increased in high and low miRNA-125b expressing H1703 and PC9 cells, respectively, when treated with a miRNA-125b inhibitor and a mimic transfection (*p < 0.05, **p < 0.01, Figure 1(b) upper panel). IGFBP3 expression was increased by the miRNA-125b inhibitor in H1703 cells and was decreased by the mimic in PC9 cells in a dose-dependent manner (*p < 0.05, **p < 0.01, Figure 1(b) lower panel). The Boyden chamber assays indicated that the invasive ability was decreased by the miRNA-125b inhibitor in H1703 cells and was increased by the mimic transfection in PC9 cells in a dose-dependent manner (Figure 1(c)). However, the miRNA-125b inhibitor and the mimic transfection did not statistically modulate the invasive ability ls in p53-WT CH27 and H460 cells (Supplementary Figure 2). These results indicated that miRNA-125b may promote invasion ability via targeting of IGFBP3 in p53-mutated cells, but not in p53-WT cells.

The PI3K/AKT activation by a decrease in IGFBP3 expression is responsible for miRNA-125b-induced cell invasion in p53-mutated cells, but not in p53-WT cells

We explored the possibility that the targeting of IGFBP3 by miRNA-125b could be responsible for invasiveness in p53-mutated cells via activation of the PI3K/AKT signaling pathway. H1703 and PC9 cells were transfected with miRNA-125b inhibitor and mimic and/or co-transfected with a small hairpin (sh)RNA and an expression vector of IGFBP3. The Boyden chamber assays indicated that invasive ability was significantly reduced by the miRNA-125b inhibitor in H1703 cells and was elevated by the mimic in PC9 cells (Figure 2(a)).

Figure 2.

MiRNA-125b promotes invasion ability via PI3K/AKT activation by IGFBP3 targeting in p53-mutated cells, but not in p53-WT cells. (a) The cell invasion ability in H1703 and PC9 lung cancer cell lines was analyzed by transwell assays after transfection with IGFBP3 shRNA and miRNA-125b inhibitor or IGFBP3 expression vector and miRNA-125b mimic, respectively. The significance was compared to the microRNA inhibitor or mimic negative control or miRNA-125b inhibitor or mimic alone. (b) The protein levels of IGFBP3, phosphorylated-AKT, and total AKT were determined in H1703 and PC9 lung cancer cell lines by western blot analysis after transfection with IGFBP3 shRNA and miRNA-125b inhibitor or IGFBP3 expression vector and miRNA-125b mimic, respectively. The β-actin was used as a loading control. (c) The cell invasion ability in H1703 and PC9 lung cancer cell lines was analyzed by transwell assays after transfection with IGFBP3 and a myr-AKT expression vector or treatment with IGFBP3 shRNA and the AKT inhibitor LY294002, respectively. The significance was compared to the microRNA inhibitor or the mimic negative control or IGFBP3 expression or shRNA vector alone.

[Figure omitted. See PDF]

We next examined whether an increase in invasion ability by miRNA-125b-mediated IGFBP3 reduction could occur through activation of the PI3K/AKT signaling pathway. Western blotting indicated that the expression of p-AKT was markedly increased by IGFBP3 silencing and/or miRNA-125b inhibitor treatment in H1703 cells when compared with H1703 cells transfected with non-specific shRNA (NC) or H1703 cells transfected with miRNA-125b inhibitor alone (Figure 2(b) left panel). Conversely, p-AKT expression was significantly increased by transfection with an miRNA-125b mimic and/or an IGFBP3 expression vector when compared with PC9 NC cells or PC9 cells transfected with a miRNA-125b mimic (Figure 2(b) right panel). The invasive ability was also markedly suppressed by IGFBP3 overexpression; however, the invasive ability induced by constitutive AKT activation (myr-AKT) was nearly unchanged by IGFBP3 overexpression in H1703 cells (Figure 2(c) left panel). Conversely, the invasive ability was increased by IGFBP3 silencing in PC9 cells, but was decreased by treatment with LY294002 (a PI3K/AKT inhibitor) and was unchanged by a combination treatment with IGFBP3 shRNA plus LY294002 in PC9 cells (Figure 2(c) right panel). These results suggest that the miRNA-125b-induced cell invasion due to targeting of IGFBP3 depends on the activation of the PI3K/AKT signaling pathway in p53-mutated cells, but not in p53-WT cells.

Targeting of IGFBP3 by miRNA-125b activates the PI3K/AKT signaling pathway depends on p53 status

We examined the possibility that targeting of IGFBP3 by miRNA-125b activated the PI3K/AKT signaling pathway depends on p53 status. P53-null H358 cells were transfected with R175H mutant p53 or a p53-WT expression vector and/or co-transfected with a miRNA-125b mimic or inhibitor. Western blotting indicated that p-AKT expression in H358 cells was elevated by the miRNA-125b mimic or by transfection with the R175H mutant p53 plus the miRNA-125b mimic, but it was almost completely suppressed by p53-WT transfection. The phosphatase and tensin homolog (PTEN) expression was expectedly increased by p53-WT transfection (Figure 3 upper panel).¹⁷ However, the expression of p-AKT in H358 cells was unchanged by transfection with the P53-WT plus the miRNA-125b mimic or with the R175H mutant p53 (Figure 3 upper panel). The expression level of IGFBP3 mRNA and protein was negatively correlated with the increase in miRNA-125b expression due to miRNA-125b mimic transfection in H358 cells (Figure 3). However, p-AKT expression was nearly completely suppressed by treatment of H358 cells with a miRNA-125b inhibitor, p53-WT, P53-WT plus miRNA-125b inhibitor, or H175H mutant p53 plus miRNA-125b inhibitor, and it was only slightly elevated by transfection with mutant p53 when compared with H358 VC cells. A similar negative correlation was observed between miRNA-125b and IGFBP3 mRNA levels in H358 cells subjected to these same treatments (Figure 3 lower panel). The results suggest that activation of the PI3K/AKT signaling pathway through targeting of IGFBP3 by miRNA-125b is dependent on p53 status.

Figure 3.

The PI3K/AKT activation by targeting of IGFBP3 by miRNA-125b is dependent on p53 status. The protein levels of IGFBP3, p53, PTEN, phosphorylated-AKT, and total AKT in H358 lung cancer cell lines were evaluated by western blot analysis after transfection with indicated combination of p53-WT, p53-mutated (R179H) plasmids, miRNA-125b inhibitor (miRNA-125bi), and miRNA-125b mimic (miRNA-125bm). The β-actin was used as a loading control.

[Figure omitted. See PDF]

Discussion

We here provided evidence that miRNA-125b promotes invasive ability in p53-mutated cells, but not in p53-WT cells via activation of PI3K/AKT signaling pathway due to targeting of IGFBP3 (Figures 1 and 2; Supplementary Figure 2). MiRNA-125-targeted IGFBP3 is regardless of p53 mutational status; however, the PI3K/AKT activation via targeting of IGFBP3 by miRNA-125b depends on p53 status (Figure 3). Among 11 targets in the p53 network, IGFBP3 was only targeted by miRNA-125b in p53-mutated H1703 and PC9 cells, but this effect was not observed in p53-WT CH27 and H460 cells (Supplementary Figure 2). No change in IGFBP3 expression was observed due to p53-WT cells with low miRNA-125b expression. Surprisingly, p53 expression was not influenced by miRNA-125b inhibitor and mimic transfection in these four cell types. This finding supported the possibility that cell invasion mediated by miRNA-125b might be through p53-independent pathway.

In this study population, high IGFBP3 occurred commonly in p53-WT patients, but this was not consistent with the finding of tested cell lines. The conflicting should be further verified by a larger study population and an increase in tested cell lines. Nevertheless, IGFBP3 expression was negatively associated with p-AKT expression in p53-mutated patients, but not in p53-WT patients (Table 2). We therefore suggest that tumor invasion and poor outcome mediated by miRNA-125b in p53-mutated patients may be through targeting of IGFBP3 and, in turn, activating the PI3K/AKT signaling pathway.

PTEN transcription is directly regulated by p53.¹⁸ We thus expected that the low PTEN expression caused by p53 mutation/deletion may activate the PI3K/AKT signaling pathway. However, no clear association was apparent between p-AKT expression and PTEN expression in these six tested p53-mutated cells (Supplementary Figure 3), which suggested that the PI3K/AKT signaling pathway in these p53-mutated cells might be activated through targeting of IGFBP3 by miRNA-125b. We observed p-AKT expression only in p53-null H358 cells transfected with the miRNA-125b mimic or with the mutant p53 (R175H) plus miRNA-125b mimic. However, no p-AKT expression was evident in H358 cells transfecting the p53-WT expression vector or with p53-WT plus miRNA-125b mimic, even though IGFBP3 expression in H358 cells was decreased by the miRNA-125b mimic (Figure 3). This was due to an increase in PTEN expression by p53-WT transfection in H358 cells. Therefore, we suggest that the tumor invasion induced by miRNA-125b via PI3K/AKT activation by IGFBP3 targeting is dependent on the p53 status.

In summary, we have provided evidence that miRNA-125 may act as an oncogene to promote cell invasion in p53-mutated NSCLC via activation of the PI3K/AKT signaling pathway due to targeting of IGFBP3. The possible route for the mechanistic action of miRNA-125b in cell invasion is proposed in Supplementary Figure 4. Consistent findings were observed in NSCLC patients and indicated that miRNA-125b levels were negatively correlated with IGFBP3 mRNA and were positively associated with p-AKT expression in p53-mutated patients, but not in p53-WT patients. The miRNA-125 levels may independently predict poorer OS and RFS in p53-mutated NSCLC patients. We therefore suggest that a clinical use AKT inhibitor (perifosine) might potentially improve tumor regression and outcomes in p53-mutated NSCLC patients whose tumors expressed high miRNA-125b level.

H.-H.W. and Y.-C.W. contributed equally to this work. H.L. was guarantor of the integrity of the study. H.-H.W., Y.-C.W., D.-W.W., and H.L. contributed to study concepts. All authors contributed to study design. H.-H.W., Y.-C.W., and D.-W.W. contributed to data acquisition. All authors contributed to data analysis. H.-H.W., Y.-C.W., and D.-W.W. contributed to statistical analysis. H.-H.W., Y.-C.W., D.-W.W., and H.L. contributed to manuscript preparation. H.-H.W., Y.-C.W., D.-W.W., and H.L. contributed to manuscript editing. H.-H.W., Y.-C.W., D.-W.W., and H.L. contributed to manuscript review.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This work was jointly supported by grants from the Aim for the Top University Project-Cancer Translational Center and the Ministry of Science and Technology in Taiwan, ROC (105-2320-B-038 -057).