Overexpression of amplified in breast cancer 1 (AIB1) gene promotes lung adenocarcinoma aggressiveness in vitro and in vivo by upregulating C-X-C motif chemokine receptor 4

Patients and tissue specimens

Thirty pairs of lung adenocarcinoma and their adjacent normal tissue samples (10 localized, 10 regional and 10 metastatic cases) were obtained with informed consent under institutional review board-approved protocols between January 2012 and December 2012 from Sun Yat-sen University Cancer Center, Guangzhou, China. Tumors without regional lymph nodes or distant metastases, tumors with regional lymph node metastases but without distant metastases, and tumor with distant metastases were defined as localized, regional and metastatic cases, respectively. Paraffin-embedded pathological specimens from 183 lung adenocarcinoma patients treated between October 1994 and February 1998 were obtained from the archives of the Department of Pathology at the same institution. All the patients were treated with initial surgical resection with a curative or palliative intent. The cases were selected consecutively based on the availability of resection tissue and follow-up data. Tumor differentiation grades and pathological tumor-node-metastasis (TNM) status were assessed according to the criteria of the World Health Organization and the 8th edition of the TNM classification of the International Union Against Cancer (UICC, 2015). The medical ethics committee of the Cancer Center of Sun Yat-sen University approved this study.

Construction of tissue microarrays (TMAs)

TMAs were constructed according to the method described previously [21]. The tissues (183 lung adenocarcinomas and 30 normal lung tissues from the same patients) were sampled using a tissue arraying instrument (Beecher Instruments, Silver Spring, MD, USA).

Immunohistochemistry (IHC)

Endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide for 15 min. Tissue slides were boiled in 10 mmol/L citrate buffer (pH 6.0) (Beyotime, Shanghai, China) in a pressure cooker for 10 min (AIB1) or microwave-treated for 10 min for antigen retrieval. The slides were incubated with anti-AIB1 [Clone 34, BD Transduction Laboratories, San Jose, CA, USA, diluted 1:50 in phosphate buffer saline (PBS)] and anti-CXCR4 (Clone 2074, Abcam, Cambridge, UK, diluted 1:1000 in PBS) overnight at 4 °C. Subsequently, the slides were sequentially incubated with biotinylated rabbit antimouse immunoglobulin (Dako, Carpinteria, CA, USA) at a concentration of 1:100 for 30 min at 37 °C and then reacted with a streptavidin-peroxidase (Dako) conjugate for 30 min at 37 °C and 3′-3′ diaminobenzidine (Dako) as a chromogen substrate. The nucleus was counterstained using Meyer’s hematoxylin (Sigma, St. Louis, MO, USA).

Since the positive nuclei staining of normal lung tissues ranged from 0% to 10% of the epithelium, normal expression and overexpression of AIB1 were identified when the nuclei of ≤ 10% and > 10% of tumor cells were positively stained, respectively. To evaluate CXCR4 IHC staining, a previously validated semi-quantitative scoring criterion was used [22, 23]. A staining index (values 0–9) was calculated by multiplying a score reflecting the intensity of CXCR4-positive staining (negative = 0, weak = 1, moderate = 2, and strong = 3) and a score reflecting the proportion of immunopositive cells of interest (< 10% = 1, 10% to 50% = 2, and > 50% = 3.

Cell lines and culture conditions

Four lung adenocarcinoma cell lines (A549, H1975, H2073 and PC9) were cultured in RPMI1640 (Gibco, Grand Island, NY, USA) medium with 10% newborn calf serum. (Gibco, Grand Island, NY, USA) Another lung adenocarcinoma cell line, H1993, was maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS) (Gibco). All 5 cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA).

Protein extraction and Western blotting

The protein was extracted from the lung adenocarcinoma cells using Radio-Immunoprecipitation Assay (RIPA) Lysis Buffer (Beyotime) at 4 °C. Protein concentrations were measured by the Bicinchoninic Acid Protein Assay (BioRad, Hercules, CA, USA). Equal amounts of whole-cell lysates were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride membrane (Millipore, Bedford, MA, USA) followed by incubation with primary mouse monoclonal antibodies against human AIB1 (1:1000 dilution), CXCR4 (1:500 dilution), tumor necrosis factor (ligand) superfamily member 10 (TNFSF10) (1:500 dilution), matrix metallopeptidase 11 (MMP11) (1:1000 dilution), matrix metallopeptidase 2 (MMP2) (1:500 dilution), and vascular endothelial growth factor A (VEGFA) (1:1000 dilution) (BD Transduction Laboratories) overnight at 4 °C. β-Actin was used as an internal control (1:1000 dilution, BD Transduction Laboratories). After washing, the polyvinylidene fluoride (PVDF) membranes were incubated with secondary antibody (goat anti-mouse, 1:10,000 dilution, Cell Signaling Technology, Danvers, MA, USA) for 2 hat room temperature. The immunoreactive proteins were detected with enhanced chemiluminescence detection reagents (Amersham Biosciences, Uppsala, Sweden) according to the manufacturer’s instructions.

Knockdown of AIB1 and CXCR4 by lentiviral short hairpin RNA (shRNA)

We synthesized the sequences of AIB1 to construct lentiviral shRNA1 (5′-GGTCTTACCTGCAGTGGTGAA-3′) and shRNA2 (5′-AGACTCCTTAGGACC GCTT-3′), which have been previously found to efficiently knock down endogenous AIB1 expression in human cancer cells [14]. The shRNA sequence for CXCR4 is 5′-ACCGCGATCAGTGTGAGTATATAAAGTTCTCTTATATACTCACACTGATCGCTTTTTC-3′, which was also previously validated [24]. Virus packaging was performed by the transient transfection of 293FT cells with a transfer plasmid and three packaging plasmids: pMDLg/pRRE, pRSV-REV, and pCMV-VSVG, which were kindly provided by Professor Peng Xiang (Center for Stem Cell Biology and Tissue Engineering, Sun Yat-sen University). Seventy-two hours after transfection, the lentiviral particles were collected, filtered, and then concentrated. Subsequently, we infected the lung adenocarcinoma cell lines with the lentivirus in a 24-well plate. Four days after infection, the knockdown efficiency was examined by Western blotting.

Plasmid constructs and transfection

The construction of a plasmid expressing human AIB1 (pcDNA-AIB1) was conducted as described in our previous study [25]. In brief, full-length human AIB1 cDNA was amplified by PCR (primer: 5′-GTCATATGATGAGTGGATTAGGAGAAAAC -3′ (forward) and 5′- CGAGATCTTCAGCAGTATTTCTGATCAGG-3′ (reverse), initial denaturation at 95 °C for 10 min and 35 cycles of 95 °C for 15 s, 55 °C for 30 s, and 72 °C for 5 min) and cloned into the NheI and EcoRI site pcDNA3.1 (+) expression vector (Invitrogen, Carlsbad, CA), then transfected into A549 cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. Cells transfected with empty vector were used as controls. Stable AIB1-expressing clones were selected by Geneticin (Rache Diagnostics, Indianapolis, IN) (500 μg/mL).

RNA interference (RNAi)

Short interfering RNAs (siRNAs) specifically directed against the CXCR4 gene (1: 5′-AUCACGUAAAGCUAGAAA-3′, 5′-GGGAUCAUUUCUAGCUUU-3′; 2: 5′-GCUGUUUAUGCAUAGAUA-3′, 5′-GAGAGAUUAUCUAUGCAU-3′) [26] and corresponding scrambled siRNAs (Ribo bio, Guangzhou, China) were transfected into A549 cells in six-well plates using Lipofectamine 2000 transfection reagent (Invitrogen) according to the manufacturer’s instructions.

Migration and invasion assays

Cell migration was assessed by measuring the movement of cells into a scratch created by a 200 ml pipette tube. The degree of wound closure was observed after 24 h and photographed under a microscope. The fraction of cell coverage across the line was measured to determine the migration rate. Wound repair = [(Diameter of the wound before migration − Diameter of the wound after migration)/Diameter of the wound before migration] × 100%. Each independent experiment was repeated three times.

For invasion assays, cells (3 × 10⁵) were added to a Matrigel invasion chamber (BD Biosciences, Becton Dickson Labware, Flanklin Lakes, New Jersey, USA) in the insert of a 24-well culture plate. Fetal bovine serum was added to the lower chamber as a chemoattractant. After 24 h, invasive cells located on the lower side of the chamber were fixed and stained with crystal violet, air dried, and photographed. The invasive cells were counted in five fields under an inverted microscope. Experiments were performed in triplicate with a minimum of 40 grids (400 magnification) per filter counted.

Real-time PCR

RNA was extracted from H1993 AIB1_shRNA2 and H1993_control_shRNA using Trizol (Invitrogen) and was cleaned using the RNeasy MinElute Cleanup Kit (Qiagen, Valencia, California, USA). The concentrations of the RNA samples were measured by NanoDrop 2000 (Thermo Fisher Scientific, Waltham Massachusetts, US). Subsequently, total RNA was reverse transcribed using Super- Script III reverse transcriptase (Invitrogen), and cDNA was amplified by PCR using 2×Super Array PCR Master Mix (SuperArray Bioscience, Frederick, Maryland, USA). Real-time PCR was then performed on each sample using the Human Tumor Metastasis RT2 ProfilerPCR array (SuperArray Bioscience) in an Opticon DNA Engine ABI PRISM7900 system (Applied Biosystems, Foster City, CA, USA), according to the manufacturer’s instructions. Data were normalized to glyceraldehyde phosphate dehydrogenase (GAPDH) levels by the ∆∆Ct method [27].

In vivo metastasis model

Animal experiments were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory animals (NIH Publications No. 8023, revised 1978). Eight 4-week-old Balb/c nude mice, which were purchased from Shanghai Slac Laboratory Animal Co. Ltd. (Shanghai, China), were injected with A549-Vec, A549-AIB1, or A549-AIB1 + CXCR4 shRNA cells. Briefly, 2 × 10⁵ cells (mixed with 100 μL PBS) were injected intravenously through the tail vein into each Balb/c nude mouse in a laminar flow cabinet. Six weeks after cell injection, the mice were killed by cervical dislocation. Their livers and lungs were harvested, fixed in 4% paraformaldehyde, and embedded in paraffin. Subsequently, serial 2-μm-thick sections of the whole lungs and livers were obtained and examined by hematoxylin and eosin (H&E) staining to identify the metastases. All the procedures were performed in accordance with the guidelines of the laboratory animal ethics committee of Sun Yat-sen University.

Statistical analysis

Statistical analysis was performed with SPSS software (SPSS Standard version 19.0, SPSS Inc. Chicago, IL). Receiver operating characteristic (ROC) curve analysis was applied to determine the cutoff scores of AIB1 expression for distinguishing localized, regional, and metastatic lung adenocarcinomas. The sensitivity, specificity, and areas under the ROC curves (AUC) were calculated. The association of AIB1 protein expression with clinicopathologic features and the correlations between molecular features were assessed by the Chi square test. Survival curves were assessed by the Kaplan–Meier method and compared by the log-rank test. Two-sided P values of less than 0.05 were considered to indicate statistical significance.

IHC staining of AIB1 expression in the localized, regional, and metastatic stages of lung adenocarcinoma tissues

The mean positive rates of AIB1 expression in lung adenocarcinoma at the localized, regional, and metastatic stages of lung adenocarcinoma were about 20.0%, 40.0%, and 70.0% respectively, compared with 3.0% (range 0.0%–10.0%) for normal tissues (P < 0.001, Fig. 1a–e). The sensitivity, specificity, and area under the ROC curve (AUC) values of AIB1 expression levels for regional versus nonregional stages were 83.3%, 93.3%, and 0.927, respectively (P < 0.001, Fig. 1f), whereas the sensitivity, specificity, and AUC values of AIB1 expression levels for metastatic versus nonmetastatic stages were 80.0%, 78.0%, and 0.890, respectively (P < 0.001, Fig. 1g).

Association between the expression of AIB1 and lung adenocarcinoma patient clinicopathologic features and survival

Silencing of AIB1 by RNA interference inhibits lung adenocarcinoma cell migration and invasion in vitro

Of the five lung adenocarcinoma cell lines analyzed, H1975, H1993, H2073, and PC9 cells showed relatively high levels of endogenous AIB1 protein expression, whereas A549 cells showed relatively low levels of AIB1 protein expression (Fig. 2a left).Two lung adenocarcinoma cell lines, H1975 and H1993, were then treated with two specific shRNAs against AIB1, and the shRNAs could efficiently knock down endogenous AIB1 in lung adenocarcinoma cells (Fig. 2a right). The knockdown of AIB1 caused an apparent suppression of cell migration in both H1975 and H1993 cell lines, as shown by using a wound-healing assay (P < 0.01, Fig. 2b). The ablation of endogenous AIB1 markedly reduced the invasive ability of both H1975 and H1993 cell lines in Matrigel invasion assays (P < 0.05, Fig. 2c).

Ectopic overexpression of AIB1 by plasmid transfection promotes lung adenocarcinoma cell migration and invasion in vitro

To determine whether the ectopic overexpression of AIB1 could enhance the migration and invasion capacity of lung adenocarcinoma cells, the A549-AIB1 cell line, which overexpressed AIB1, was constructed and used to perform wound-healing and invasion assays (Fig. 3a left) The wound-healing assay showed that the ectopic overexpression of AIB1 enhanced A549 cell migration at the edge of the exposed regions (P < 0.01, Fig. 3a right). The Matrigel invasion assay demonstrated that the invasive capacity of the A549-AIB1 cells was greater than that of the control A549-vector cells (P < 0.05, Fig. 3b).

AIB1 up-regulates CXCR4 expression in lung adenocarcinoma cells

To identify potential downstream targets regulated by AIB1 that were involved in lung adenocarcinoma cell invasion and/or metastasis, the mRNA expression profiles of shAIB1-transfected H1993 cells were compared with those of the control H1993 cells using a Human Tumor Metastasis RT2 Profiler TM PCR Array containing 84 cell metastasis-related genes. The results showed that a total of five downregulated genes (3.5-fold) were identified in shAIB1-transfected H1993 cells (Table 2, Additional file 1: Table S1). Subsequently, these five downstream targets (CXCR4, TNFSF10, MMP11, MMP2, and VEGFA) were selected and analyzed by Western blotting (Fig. 4a). Consistent with the mRNA expression in the real-time PCR array, decreased protein expression of CXCR4 was shown by Western blotting in H1993 cells after AIB1 knockdown (Fig. 4b).

Gene symbol	Gene name	Location	Fold change
CXCR4	Chemokine (C-X-C motif) receptor 4	2q21	− 22.49
TNFSF10	Tumor necrosis factor (ligand) superfamily, member 10	3q26	− 8.37
MMP11	Matrix metallopeptidase 11	22q11.23	− 4.63
MMP2	Matrix metallopeptidase 2	16q13-q21	− 4.31
VEGFA	Vascular endothelial growth factor A	6p12	− 3.85

CXCR4 mediates the invasiveness of AIB1-induced lung adenocarcinoma

To determine whether CXCR4 is a functional downstream target involved in AIB1-induced lung adenocarcinoma cell aggressiveness, two RNAis were used to silence CXCR4 expression in AIB1-overexpressing A549 cells (Fig. 4c). We found that after the siCXCR4 treatment of A549-AIB1 cells, the AIB1-induced migration and invasive capacities of the A549 cells were dramatically inhibited (P < 0.05, Fig. 4c, d). In addition, we upregulated the expression of CXCR4 in H1993-shAIB1 cells and found that the attenuated migration and invasion cell abilities caused by AIB1 depletion were rescued by CXCR4 overexpression (P < 0.05, Fig. 4e, f).

Enforced expression of AIB1 enhances the metastasis potential of the lung adenocarcinoma cell line mediated by CXCR4 in vivo

To investigate whether AIB1 could affect the metastatic potential of lung adenocarcinoma cells in vivo and whether this effect could be mediated by CXCR4, we performed in vivo metastasis assays using a Balb/c nude mouse model. We did not detect any tumor nodules in the livers of all mice examined; however, metastatic tumor nodules were frequently found in the lungs of the mice (Fig. 5a). The expression levels of AIB1 and CXCR4, as detected by IHC, were simultaneously higher in the tumors of the A549-AIB1 group than in the tumors of the A549-vector group (Fig. 5b). The overexpression of AIB1 increased the number of lung metastases in mice injected with A549-AIB1 cells by approximately three-fold compared with the number in mice injected with A549 cells, whereas the depletion of CXCR4 dramatically decreased the number of AIB1-induced lung metastases (P < 0.05, Fig. 5b).

Expression of CXCR4 in lung adenocarcinoma tissues and its correlation with AIB1 expression and patient survival

The median staining index of CXCR4 in lung adenocarcinoma was 3; thus, the categories of high and low expression were defined as groups with a staining index above or below 3. In 156 of the 183 samples, AIB1 and CXCR4 IHC were detected successfully and simultaneously. The rate of high CXCR4 expression was significantly greater in carcinomas overexpressing AIB1 (58/83 cases, 69.9%) than in those cases with a normal expression of AIB1 (23/73 cases, 31.5%, P < 0.001, Table 3, Fig. 6a, b). Furthermore, high expression of CXCR4 was significantly associated with poorer survival in lung adenocarcinoma patients (P < 0.001, Fig. 6c). In addition, lung adenocarcinoma patients with high expression of both AIB1 and CXCR4 displayed the poorest survival, whereas patients with low expression of AIB1 and CXCR4 had the best survival (P < 0.001, Fig. 6d).

CXCR4 expression	Cases	AIB1 expression [cases (%)]		P
		Normal	Over
				< 0.001
Low	75	50 (66.7)	25 (33.3)
High	81	23 (28.4)	58 (71.6)