Long non-coding RNA AGAP2-AS1 increases the invasiveness of papillary thyroid cancer

AGAP2-AS1 expression is increased in PTC tissues and correlates with TNM stage and lymph node metastasis in PTC

Using qRT-PCR, we analyzed AGAP2-AS1 expression in 110 paired PTC tissues and adjacent non-cancerous tissues. Compared with non-cancerous tissues, the AGAP2-AS1 expression was significantly increased in PTC tissues (Figure 1A, 1B). The median value of AGAP2-AS1 expression in PTC tissues was 0.6553. Using this median, we divided the 110 patients into two groups: Those with a high relative expression of AGAP2-AS1 (≥ 0.6553; n = 56) and those with a low relative expression of AGAP2-AS1 (< 0.6553; n = 54). The results in Table 1 show that the relative expression of AGAP2-AS1 was associated with the TNM stage (p = 0.006) and lymph node metastasis (p = 0.01). There was no significant correlation between the two groups and other clinicopathological characteristics.

Figure 1. AGAP2-AS1 expression is increased in PTC tissues and cells. (A) The fold change of AGAP2-AS1 expression between PTC tissues and paired non-cancerous tissues. (B) RT-PCR of AGAP2-AS1 levels in 110 pairs of PTC tissues and paired non-cancerous tissues. The Wilcoxon signed-rank test was used to analyze the differences between the two groups; data are presented as the median with a range. **P < 0.01. (C) RT-PRC of AGAP2-AS1 levels in TPC1, K1, BCPAP, and IHH4 PTC cells, and in a normal human thyroid follicular epithelial cell line (Nthy-ori 3-1). Data were analyzed using an independent samples t-test. *P < 0.05, **P < 0.01 vs. the Nthy-ori 3-1 group.

In addition, we analyzed the relative expression of AGAP2-AS1 in PTC cell lines (K1, TPC1, IHH4, and BCPAP) and in normal human thyroid follicular epithelial cells (Nthy-ori 3-1). Compared with Nthy-ori 3-1 cells, the AGAP2-AS1 expression was increased in TPC1, K1, and IHH4 cells (Figure 1C).

Knockdown of AGAP2-AS1 impairs migration and invasion of PTC cells

To investigate the role of AGAP2-AS1 in PTC cells, we first used two short interfering RNA (siRNA) sequences to knockdown the expression of AGAP2-AS1. As shown in Figure 2A, siRNA1 efficiently suppressed the AGAP2-AS1 expression in K1 and TPC1 cells; thus, it was used to investigate the function of AGAP2-AS1 in PTC. Importantly, AGAP2-AS1 suppression decreased invasion and migration of PTC cells analyzed by Trans-well assay (Figure 2B). In addition, AGAP2-AS1 knockdown decreased PTC cell proliferation and migration analyzed by wound-healing assay (Figure 2C, 2D). We also analyzed the levels of proteins involved in migration and invasion using western blotting. While there was no significant difference in the levels of E-cadherin, N-cadherin, MMP9, and vimentin between AGAP2-AS1 knockdown and control cells, AGAP2-AS1 suppression significantly decreased the level of MMP2 (Figure 2E), indicating that AGAP2-AS1 promotes migration and invasion of PTC cells by increasing the expression of MMP2.

Figure 2. Knockdown of AGAP2-AS1 inhibits invasion and migration in PTC cells. (A) RT-PCR of AGAP2-AS1 after si-RNA suppression in PTC cells; **P < 0.01 vs. NC. (B) Transwell assay in PTC cells after transfection with si-AGAP2-AS1 or NC. Data are presented as the mean ± SD, analyzed using independent samples t-test. **P < 0.01 vs. NC. (C, D) Wound healing assay in PTC cells after transfection with si-AGAP2-AS1 or NC. E. Western analysis of N-cadherin, E-cadherin, MMP2, MMP9, and Vimentin after transfection with si-AGAP2-AS1 or NC in PTC cells; **P < 0.01, *P < 0.05 vs. NC.

AGAP2-AS1 suppression decreases MMP2 expression by upregulating miRNA-424-5p in PTC cells

MiRNAs and lncRNAs interact via miRNA recognition sequences, in which the lncRNA acts as a competing endogenous RNA (ceRNA), resulting in sequestration of the miRNA, and inhibition of the miRNA regulatory effect on the target mRNA [20, 21]. The ceRNA mechanism usually occurs in the cytoplasm; this is consistent with the canonical view that miRNAs act in the cytoplasm [22]. To elucidate the mechanism of how AGAP2-AS1 upregulates the MMP2 expression, we first analyzed the cellular localization of AGAP2-AS1 using immunofluorescence. As shown in Figure 3A, AGAP2-AS1 was localized mainly in the cytoplasm in TPC1 and K1 cells. To investigate whether AGAP2-AS1 may function as a ceRNA/miRNA sponge, we searched for miRNAs that could bind to both AGAP2-AS1 and MMP2, using the bioinformatics target prediction tools Targetscan and MirDB. Altogether, 16 miRNAs were identified that might target both AGAP2-AS1 and the 3′ UTR of MMP2. To test whether expression of these 16 miRNAs is regulated by AGAP2-AS1, we analyzed their levels by qRT-PCR after AGAP2-AS1 knockdown in K1 and TPC1 cells. As shown in Figure 3B, high levels of miR-106b-5p, miR-20a-5p, miR-29b-3p, miR-424-5p, miR-4270, and miR-17-5p were observed in cells with knocked-down AGAP2-AS1 compared with control cells. To validate these results, we overexpressed these six miRNAs, and analyzed the effect on MMP2 expression. As shown in Figure 3C, 3D, overexpression of miR-424-5p had the most significant effect on the suppression of MMP2 in K1 and TPC1 cells. Together, these data demonstrate that AGAP2-AS1 is localized mainly in the cytoplasm in PTC cells, and by downregulating the expression of miR-424-5p, it increases the MMP2 expression.

Figure 3. AGAP2-AS1 suppression decreases MMP2 expression by upregulating miRNA-424-5p in PTC cells. (A) FISH analysis demonstrating cytoplasmic localization of AGAP2-AS1 in TPC1 and K1 cells. (B) qRT-PCR of miRNAs in TPC1 and K1 cells after AGAP2-AS1 knockdown. Data are presented as the mean ± S.D, analyzed using an independent samples t-test; **P < 0.01 vs. NC. (C, D) Western blotting of MMP2 in TPC1 and K1 cells incubated with miR-424-5p mimic. Data are presented as the mean ± S.D, analyzed using an independent samples t-test; *P < 0.05, **P < 0.01 vs. NC.

MiR-424-5p binds to AGAP2-AS1 and MMP2 to suppress invasion and migration of PTC cells

Next, we analyzed the miR-424-5p levels in PTC tissues. As shown in Figure 4A, the levels of miR-424-5p were decreased in 40 paired PTC tissues compared with adjacent non-cancerous tissues. Moreover, the levels of AGAP2-AS1 and miR-424-5p negatively correlated in these 40 pairs of PTC tissues (Figure 4B). Overexpression of a miR-424-5p mimic decreased invasion and migration of PTC cells in vitro, as analyzed by Trans-well and wound-healing assays (Figure 4C, 4D).

Figure 4. MiR-424-5p is downregulated in PTC tissues, binds to AGAP2-AS1 and MMP2, and regulates invasion and migration of PTC. (A) Expression of miR-424-5p in 40 pairs of PTC tissues and paired non-cancerous tissues, assessed using qRT-PCR. The Wilcoxon signed-rank test was used to analyze the differences between the two groups; data are presented as the median with a range. **P < 0.01. (B) Pearson’s correlation analysis between AGAP2-AS1 and miR-424-5p expression in PTC tissues (R² = 0.1046, p = 0.0418). (C) Trans-well assay of invasion and migration of PTC cells after transfection with miR-424-5p mimic or NC. Data are presented as the mean ± S.D, analyzed using an independent samples t-test. **P < 0.01 vs. NC. (D) Wound healing assay to assess the migration ability of PTC cells after transfection with miR-424-5p mimic or NC. (E) Predicted miR-424-5p binding sites in AGAP2-AS1 (AGAP2-AS1-Wt) and the designed mutant sequence (AGAP2-AS1-Mt) are indicated. HEK 293T cells were transfected with AGAP2-AS1-Wt, AGAP2-AS1-Mt, and the indicated miRNAs, and luciferase reporter assay was conducted. Data are presented as the mean ± S.D, analyzed using an independent samples t-test; **P < 0.01 vs. AGAP2-AS1-Wt + NC. (F) Predicted miR-424-5p binding sites in the 3′-UTR region of MMP2 (MMP2-3′UTR-Wt) and the designed mutant sequence (MMP2-3′UTR-Mt) are indicated. HEK 293T cells were transfected with MMP2-3′UTR-Wt or ATAD2-3′UTR-Mt and the indicated miRNAs, and luciferase reporter assay was conducted. Data are presented as the mean ± S.D., analyzed using independent samples t-test. **P < 0.01 vs. MMP2-3′UTR-Wt + NC.

To assess whether miR-424-5p binds to both AGAP2-AS1 and the 3′ UTR of MMP2, we used dual-luciferase reporter assays. Simultaneous transfection of AGAP2-AS1 wild-type (AGAP2-AS1 Wt) and the miR-424-5p mimic distinctly decreased the luciferase activity; however, the luciferase activity was not affected when the miR-424-5p mimic and AGAP2-AS1 mutated-type (AGAP2-AS1 Mt) were cotransfected. The miR-424-5p mimic could also reduce the luciferase activity controlled by wild-type MMP2-3′ UTR (MMP2-3′ UTR Wt), but did not show any effect with the mutated-type MMP2-3′ UTR in HEK 293T cells (MMP2-3′ UTR Mt). These results indicate that miR-424-5 binds to both AGAP2-AS1 and MMP2, and suggest that AGAP2-AS1 upregulates MMP2 expression by sponging miR-425-5p (Figure 4E, 4F).

Downregulation of miR-424-5p partly rescues si-AGAP2-AS1-mediated inhibition of invasion and migration in PTC cells

We performed a rescue experiment to validate the interaction between AGAP2-AS1 and miR-424-5p. When si-AGAP2-AS1 and miR-424-5p inhibitor were expressed simultaneously, the suppressed invasion and migration caused by si-AGAP2-AS were not observed in TPC1 and K1 cells (Figure 5A, 5B). Wound assays showed similar results: Downregulation of miR-424-5p could partly rescue the impact of si-AGAP2-AS1 (Figure 5C, 5D). Western blotting verified the rescue effect of miR-424-5p inhibitor, showing that MMP2 expression was distinctly suppressed when cells were co-transfected with the miR-424-5p inhibitor and si-AGAP2-AS1 compared to cells transfected with si-AGAP2-AS1 alone (Figure 5E). Together, our findings demonstrate that the lncRNA AGAP2-AS1 upregulates invasion and migration of PTC cells by enhancing the expression of MMP2 by sponging miR-424-5p.

Figure 5. Downregulation of miR-424-5p partly rescues si-AGAP2-AS1 mediated inhibition of invasion and migration in PTC cells. (A, B) Trans-well assays of PTC cell migration and invasion after transfection with si-AGAP2-AS1, or a co-transfection with miR-424-5p inhibitor and si-AGAP2-AS1 or NC. Data are presented as the mean ± S.D., analyzed using an independent samples t-test; **P < 0.01 vs. the si-AGAP2-AS1 group. (C, D) Wound-healing assay of PTC cell migration and invasion after transfection with si-AGAP2-AS1, or co-transfection with miR-424-5p inhibitor and si-AGAP2-AS1 or NC. (E) Western blotting of MMP2 after transfection with si-AGAP2-AS1 or co-transfection with miR-424-5p inhibitor and si-AGAP2-AS1 or NC. Data are presented as the mean ± S.D., analyzed using an independent samples t-test; **P < 0.01 vs. the si-AGAP2-AS1 group. (F) Schematic illustration of the proposed function of AGAP2-AS1 in PTC. AGAP2-AS1 functions as a competing endogenous RNA that upregulates MMP2 expression and promotes PTC progression by sponging miR-424-5p. (G) AGAP2-AS1 played a role as ceRNA in different cancers.

Patient samples

Paired samples of PTC and adjacent non-cancerous tissues were collected from patients who received surgery at the First Hospital of China Medical University between 2010 and 2017. The sampled tissues were examined pathologically, frozen in liquid nitrogen, and stored at −80 °C. Their clinicopathological characteristics, such as gender, age, extrathyroidal extension, lymph node metastasis, multicentricity, tumor-node-metastasis (TNM) stage, and tumor size, were recorded.

RNA extraction and qRT-PCR

RNAiso (Takara, Dalian, China) was used to isolate total RNA. SYBR Premix Ex Taq II (Takara, Shiga, Japan) was used to perform qRT-PCR on a LightCycler 480 system (Roche Molecular Systems Inc., Branchburg, NJ, USA) according to the manufacturers' instructions. The cycling conditions comprised an initial denaturation at 95 °C for 30 s; followed by 40 cycles of denaturation at 95 °C for 5 s, annealing at 60 °C for 30 s; and dissociation at 95 °C for 60 s, 55 °C for 1 min, and 95 °C for 30 s. The primer sequences used were: AGAP2-AS1 5′- TACCTTGACCTTGCTGCTCTC-3′ (forward) and 5′-TGTCCCTTAATGACCCCATCC-3′ (reverse), hsa-miR-424-5p 5′- GCCAGCAGCAATTCATGT-3′ (forward) and 5′- TATGGTTTTGACGACTGTGTGAT-3′ (reverse), GAPDH 5′-CAGGAGGCATTGCTGATGAT-3′ (forward) and 5′-GAAGGCTGGGGCTCATTT-3′(reverse).

Cell culture

The European Collection of Authenticated Cell Culture (ECACC, UK) provided the Nthy-ori 3-1 and K1 cell lines. Professor Meiping Shen (Department of General Surgery, The First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu) gifted the TPC1 cell line. DSMZ (Braunschweig, Germany) provided the BCPAP cell line. The Health Science Research Resources Bank (Osaka, Japan) provided the IHH4 cells. TPC1 cells were cultured in Dulbecco’s modified eagle’s medium (DMEM) supplemented with 15% fetal bovine serum (FBS). K1 cells were cultured in DMEM:Ham’s F12:MCDB 105 (2:1:1) supplemented with 2 mM glutamine and 10% FBS. BCPAP and Nthy-ori 3-1 cell lines were maintained in RPMI-1640 medium containing 10% FBS. IHH4 cells were cultured in a 1:1 mixture of RPMI-1640 and DMEM with 10% FBS. Short interfering RNA (siRNA) and negative control (NC) were from Gene Pharma (Suzhou, China). The si-AGAP2-AS1 (sense) sequence was: 5′- GGCACAACGACAAAUGUCUTT-3′. The negative control (NC) sequence was 5′-UUC UCC GAA CGU GUC ACG UTT-3′. Lipofectamine 2000 was from Invitrogen.

Cell invasion and migration assays

After 24 h transfection, cell concentration was adjusted to 5 × 10⁴ cells/ml with serum-free medium. The upper chamber of Transwell chamber (Costar; 24-well insert, pore size: 8 μ m) was filled with 200 μl of cell suspension, and the lower chamber was filled with 500 μl of medium with 10% FBS. For the invasion assay, polycarbonate filters coated with 50 μl Matrigel (1:9, BD Bioscience) were placed in a Transwell chamber. Cells were incubated for 12 h for the migration assay, and 36 h for the invasion assay. Cells on the bottom surface of the membrane were fixed with 4% paraformaldehyde and stained with 0.5 % crystal violet. The migratory cells were visualized and counted in five random visual fields per insert under an inverted microscope at 100× magnification (Nikon MicrophotFX, Japan). The experiment was repeated three times.

Wound healing assay

After 24 hours post-transfection, cells were plated in six-well plates at 2 × 10⁵ cells per well. When cells in each well reached 80% confluence, a 200-μl pipette tip was used to scratch the cell surface. PBS was used to wash off free cells, and the wound was photographed; three images were obtained for each wound. The difference in the wound width at 0 and 24 h was used to measure the extent of wound healing. The experiment was repeated three times.

Western blotting

A Total Protein Extraction Kit (KeyGEN, Nanjing, China) was used to extract proteins from PTC cells. Proteins (20–30 μg) were separated using 10% SDS-PAGE and transferred onto polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). The membranes were blocked with 5% skim milk for 2 h, and incubated overnight at 4 °C with primary antibodies against MMP2, E-cadherin, MMP9, N-cadherin (1:2000 dilution; Abcam, Cambridge, MA, USA), and control glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (ZSGB-Bio, Beijing, China; 1:1000 dilution). The membranes were then incubated with secondary antibodies (1:10,000 dilution; Cell Signaling Technology, Danvers, MA, USA), followed by chemiluminescence detection (Thermo Fisher Scientific, Waltham, MA, USA). To quantify the intensity of the protein bands, Image J (NIH, Bethesda, MD, USA) was used.

Fluorescence in situ hybridization analysis (FISH)

A PARIS kit (Life Technologies) was used to separate the nuclear and cytosolic fractions, and RNA FISH probes were designed and synthesized. Briefly, cells were fixed with 4% formaldehyde for 15 min and washed with PBS. After pepsin treatment, the fixed cells were dehydrated through an ethanol series. The cells were air-dried and then incubated with the FISH probes (40 nM) in hybridization buffer (Life Technologies). Thereafter, the slides were washed, dehydrated, and mounted using Prolong Gold Antifade Reagent. Nuclei were stained using 2-(4-amidinophenyl)-1H-indole-6-carboxamidine (DAPI), and cells were visualized and photographed under a fluorescence microscope (DMI4000B, Leica, Wetzlar, Germany). The probe sequences are as follows:

1. 5′-CGTCG ACACC CAGGA GATGC GACCA CAGCT CGCCG GGCAC-3′

2. 5′-TGCCG GAGCG CACAG CGAAA GGGTG TGCGC CAACC GGCAC-3′

3. 5′-TTTGA GAATG GAACT GGAAV GACGA GAGTG TCGTA ACGTC -3′

Negative control is blank control.

Luciferase reporter assay

The fragment from AGAP2-AS1 containing the putative binding sites for miR-424-5p was amplified by PCR, cloned in the firefly luciferase expression vector pMIR-REPORT (Obio Technology, China), and named as AGAP2-AS1-wild-type (AGAP2-AS1-Wt). To mutate the putative binding sites for miR-424-5p in AGAP2-AS1, the sequence of putative binding site was replaced as indicated and was named as AGAP2-AS1-mutated-type (AGAP2-AS1-Mt). HEK 293T cells were seeded into 96-well plates the day before transfection, and transfected with the pMIR-REPORT-AGAP2-AS1-lncRNA-Wt, or pMIR-REPORT-AGAP2-AS1-lncRNA-Mt, together with the Renilla luciferase-expressing vector pRL-TK (Promega, Madison, WI, USA) and miR-424-5p mimic or NC using Lipofectamine 2000 (Invitrogen). Similarly, wild-type MMP2-3′ UTR (MMP2-3′ UTR-Wt) and mutated-type MMP2-3′ UTR (MMP2-3′ UTR-Mt) containing the putative binding site of miR-424-5p were established and cloned into the Firefly luciferase expression vector pMIR-REPORT (Obio Technology, China). HEK 293T cells were seeded into 96-well plates the day before transfection, and transfected with either the pMIR-REPORT-MMP2-3′ UTR-Wt or the pMIR-REPORT-MMP2-3′ UTR-Mt reporter vector, together with the Renilla luciferase-expression vector pRL-TK (Promega) and miR-424-5p mimic or NC using Lipofectamine 2000 (Invitrogen). After 48 h, the cells were harvested, and firefly and Renilla luciferase activities were measured using the dual-luciferase reporter assay system (Promega).

Statistical analysis

GraphPad Prism 6.0 (GraphPad Software, La Jolla, CA, USA) and SPSS 21.0 software (IBM Corp. Armonk, NY, USA) were used to perform the statistical analyses. The relationship between AGAP2-AS1 expression and the clinicopathological characteristics was examined using the chi-squared test. The results were considered statistically significant at a P-value < 0.05.

Ethics statement

The Ethics Committee of the First Affiliated Hospital of China Medical University, Shenyang, China approved this study. All study participants provided written informed consent.