Anterior gradient protein 2 promotes survival, migration and invasion of papillary thyroid carcinoma cells

We measured AGR2 expression by immunohistochemistry with an anti-AGR2 monoclonal antibody in 64 samples including 25 normal thyroid (NT) samples and 39 PTC samples. Representative immunohistochemical staining is shown in Figure 1A, and the entire dataset is reported in Table 1. AGR2 was expressed at low levels in normal thyroid glands (Figure 1A, inset 1). In contrast, several PTC samples (PTC classical variant: 23 out of 28 samples; PTC follicular variant: 5 out of 11 samples) were strongly positive for AGR2 expression, and positivity was confined to tumor cells (Table 1 and Figure 1A, inset 2). No staining was observed in the absence of the primary antibody (Figure 1A, inset 3). Normal colon mucosa was used as positive control (Figure 1A, inset 4).

To determine whether up-regulation occurred also at RNA level, we verified AGR2 expression levels in a DNA array dataset of 15 normal human thyroid (NT) and 43 PTC samples. As shown in Figure 1B, PTC samples showed significantly increased AGR2 levels compared to normal thyroid tissues (P < 0.0001). These findings were further confirmed by quantitative RT-PCR analysis of an independent set of PTC samples (Figure 1C).

We evaluated the expression level of AGR2 in the papillary thyroid cancer cell line TPC-1 in comparison to non-transformed thyroid cells Nthy-ori 3–1. As shown in Figure 2A, AGR2 is over-expressed in TPC-1 cells. We evaluated the effects of AGR2 ablation in TPC-1 cells by RNA interference. TPC-1 cells were transfected with AGR2 siRNA or with scrambled siRNA, counted and lysed at different time points. Forty-eight hours after transfection TPC-1 cells transfected with scrambled siRNA numbered 192 × 10³, whereas those transfected with AGR2 siRNA numbered 146 × 10³ (P = 0.0492). Seventy–two hours after transfection, TPC-1 cells transfected with scrambled siRNA numbered 413 × 10³, whereas those transfected with AGR2 siRNA numbered 279 × 10³ (P = 0.0045) (Figure 2B). Accordingly, the fraction of trypan blue excluding (viable) cells of TPC-1 transfected with AGR2 siRNA, was reduced with respect to scrambled control, at 48 and 72 hours post transfection (P < 0.0001) (Figure 2C).

We evaluated migration (by a wound-closure assay) and invasion (by a Matrigel invasion assay) ability of AGR2 siRNA-transfected cells. A scraped wound was introduced on a confluent monolayer of TPC-1 cells transfected with AGR2 siRNA or scrambled siRNA. Figure 2D and E show that TPC-1 transfected with the scrambled control efficiently migrated into the wound; in contrast, cells transfected with AGR2 siRNA had a reduced migrating ability both at 12 and 24 hours (P < 0.0001). Moreover, cells transfected with AGR2 siRNA showed a reduced ability to invade Matrigel compared to control cells (P < 0.0001) (Figure 2F and G). The efficiency of AGR2 knockdown at the different time points examined is shown in Figure 2H.

We transfected the non-transformed thyroid cells, Nthy-ori 3–1, with the empty vector (pcDNA), AGR2 wild type, AGR2-carrying the cysteine 81 to serine mutation AGR2 (C → S), or AGR2 carrying the glutamic acid 60 to alanine mutation AGR2 (E → A) (Figure 3A). Wild type AGR2, after treatment with disuccinimidyl suberate (DSS), is dimeric (Additional file1: Figure S1). The C81S mutation targets the thioredoxin domain of AGR2 and impairs its catalytic activity and dimer formation, while the E60A mutation impairs AGR2 dimer formation[29‐31]. Nthy-ori 3–1 AGR2, AGR2 (C → S), AGR2 (E → A) and pcDNA cells showed comparable growth rates (Figure 3B). A scraped wound was introduced on the confluent monolayer and cell migration into the wound was monitored after 12 hours. As shown in Figure 3C and D, migration rate was larger in AGR2-transfected cells compared to control vector (pcDNA), AGR2 (C → S) and AGR2 (E → A) cells. Then we seeded AGR2, AGR2 (C → S) and vector control (pcDNA) cells into the top chamber of transwells and evaluated their ability to invade Matrigel. AGR2 over-expression increased the invasive ability of Nthy-ori 3–1 cells in comparison to vector control (P < 0.0001) and AGR2 (C → S) cells (P < 0.0001) (Figure 3E and F).

To confirm these findings, we over-expressed AGR2 in TPC-1 cells. Cells were transfected with AGR2, AGR2 (C → S), AGR2 (E → A) or with the empty vector (pcDNA). Mass populations were selected in G418 (Additional file1: Figure S2A). Growth rates of TPC-1 AGR2, AGR2 (C → S), AGR2 (E → A) and control (pcDNA) cell lines were similar (Additional file1: Figure S2B). We studied cell migration using the wound closure assay. As shown in Additional file1: Figure S2C, migration rate was larger in TPC-1 AGR2 cells compared to AGR2 (C → S), AGR2 (E → A) cells and to vector control (pcDNA) cells. We next seeded TPC-1 AGR2, AGR2 (C → S), and pcDNA control cells into the top chamber of transwells and evaluated their ability to invade Matrigel. AGR2 over-expression increased TPC-1 invasiveness in comparison to AGR2 (C → S) cells and control vector (Additional file1: Figure S2D).

AGR2 has been identified as a protein disulfide isomerase, playing a role in the endoplasmic reticulum (ER) stress response[22‐33]. Protein disulfide isomerases catalyze the formation (oxidation), breakage (reduction) and rearrangement (isomerization) of disulfide bonds between cysteine residues within proteins during their folding process. To determine whether AGR2 had a disulfide isomerase activity in thyroid cells we performed an insulin-reduction assay. In this assay, insulin reduction catalyzed by protein disulfide isomerases in the presence of dithiothreitol (DTT) results in the aggregation of its β chains that can be spectrophotometrically revealed at 650 nm[34]. As shown in Figure 4A, Nthy-ori 3–1 AGR2 cells showed an increased rate of insulin reduction compared to AGR2 (C → S) or vector control cells.

To confirm these findings, we measured the levels of oxidized (GSSG) and of reduced (GSH) glutathione in Nthy-ori 3–1 AGR2, AGR2 (C → S) and vector control cells. As shown in Figure 4B and C, levels of GSSG and GSH were increased in Nthy-ori 3–1 AGR2 cells compared to AGR2 (C → S) and to vector control cells. Finally, as shown in Figure 4D Nthy-ori 3–1 AGR2 cells displayed a decreased level of H₂O₂ compared to vector control and to AGR2 (C → S) cells (P < 0.001).

It was shown that AGR2 has a role in the endoplasmic reticulum (ER) stress response[27‐30]. Thus, we studied the effects of ectopic AGR2 expression in the ER stress pathway upon treatment with Bortezomib. Bortezomib is a potent and selective inhibitor of the proteasome that causes an accumulation of unfolded proteins in the ER[35‐37]. Nthy-ori 3–1 cells transfected with AGR2, AGR2 (C → S), or vector (pcDNA) were treated for 24 hours with 100 nM Bortezomib. After treatment, cells were counted. As shown in Figure 5A, AGR2 protected Nthy-ori 3–1 cells from reduced viability caused by Bortezomib in comparison to AGR2 (C → S) and to control (pcDNA) cells.

Then, we determined by Western blot the expression level of GADD153 (also named Chop), a well-characterized biomarker of ER stress[37], in Nthy-ori 3–1 AGR2, AGR2 (C → S) or empty vector (pcDNA) cells treated with Bortezomib (100 nM) for 24 hours. As shown in Figure 5B, Nthy-ori 3–1 AGR2 cells presented a blunted increase of GADD153 upon Bortezomib treatment compared to pcDNA and to AGR2 (C → S) cells. These data suggest that ectopic AGR2 expression protects non-transformed thyroid cells from ER stress induced by Bortezomib.

Tumor (n = 39) and normal (n = 25) thyroid tissue samples for immunohistochemical analysis were retrieved from the Pathology Department of the Istituto Pascale, Naples, Italy. In addition, RNA from frozen tissues of 10 PTC cases and 10 normal matched thyroid controls was obtained from the Chernobyl Tissue Bank (http://​www.​chernobyltissueb​ank.​com)[43]. A set of 43 PTC and 15 normal thyroid tissues, also obtained from Chernobyl Tissue Bank, was used in the DNA array study (HG-U133 Plus 2.0, Handkiewicz Junak et al., manuscript submitted for publication) carried out in MSC Memorial Cancer Center and Institute of Oncology, Gliwice, Poland. In all cases, the respective Institutional review boards approved the study. The samples were classified according to the diagnostic criteria required for the identification of PTC[44].

Sections of paraffin-embedded samples were stained with hematoxylin and eosin for histological examination[44]. For immunohistochemistry, 2 μm paraffin sections were deparaffinized and placed in a solution of 0.3% hydrogen peroxide in absolute methanol for 30 minutes and then washed in phosphate buffered saline (PBS) before immunoperoxidase staining. The slides were then incubated overnight at 4°C in a humidified chamber with mouse monoclonal antibody against AGR2 (PO1, Abnova, Taipei City, Taiwan) diluted 1:100 in PBS. The slides were subsequently incubated with biotinylated goat anti-mouse IgG for 20 minutes (Vectostain ABC kits, Vector Laboratories) and then with premixed reagent ABC (Vector) for 20 minutes. The immunostaining was performed by incubating the slides in diaminobenzidine (DAB, DAKO) solution containing 0.06 mM DAB and 2 mM hydrogen peroxide in 0.05% PBS, pH 7.6, for 5 minutes, and after chromogen development, the slides were washed, dehydrated with alcohol and xylene and mounted with coverslips using a permanent mounting medium (Permount). Micrographs were taken on Kodak Ektachrome film with a Zeiss system. Negative controls were performed in each case by incubating tissue slides with secondary antibody only. The expression of AGR2 was categorized as positive (staining of ≥10% of the tumor cells) or negative (staining of < 10% of the tumor cells). The staining intensity was further graded in: score 1 (10-50% of positive cells), score 2 (51-75% of positive cells) and score 3 (76-100% of positive cells). Score values were independently assigned by two blinded investigators (G.C. and M.M.) and a consensus was reached on all scores used for computation.

Human papillary thyroid cancer cell line TPC-1 was obtained from M. Nagao (Carcinogenesis Division, National Cancer Center Research Institute, Tokyo, Japan) and identified based on the presence of the RET/PTC1 rearrangement. TPC-1 cell line was grown Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen) supplemented with 10% fetal bovine serum (FBS), L-glutamine and penicillin/streptomycin (Invitrogen). Nthy-ori 3–1 cells are non-transformed human thyrocytes immortalized by the Large T of SV40 and were obtained from the European Tissue Culture collection (Sigma Aldrich, St. Louis, MO, USA). Nthy-ori 3–1 cell line was grown in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% fetal bovine serum (FBS), L-glutamine and penicillin/streptomycin (Invitrogen, Carlsbad, California, USA).

Bortezomib was obtained from Millenium (Cambridge, MA, USA). Human insulin, disuccinimidyl suberate (DSS) and dithiothreitol (DTT) were from Sigma Aldrich (St. Louis, MO, USA). Glutathione reduced and oxidized assay kit was obtained from Bioassay Systems (Hayward, CA, USA).

For cell viability determination, cells were collected by trypsinization stained for 10 minutes with 0.4% trypan-blue (Sigma Aldrich) according to manufacturer’s instructions, and counted in triplicate.

Cell invasion was examined using a reconstituted extracellular matrix (Matrigel, BD Biosciences, San Jose, CA). The cell suspension (1 × 10⁵ cells per well) was added to the upper chamber of transwell cell culture chambers on a prehydrated polycarbonate membrane filter of 8 μm pore size (Costar, Cambridge, MA) coated with 35 μg of Matrigel (BD Biosciences). The lower chamber was filled with 10% medium. After incubation at 37°C, non-migrating cells on the upper side of the filter were wiped-off. Invading cells were mounted on glass slides using mounting medium and stained with Hoechst (Sigma Aldrich). Cell migration was quantified by counting the number of stained nuclei in three individual fields in each transwell membrane, by fluorescence microscopy, in duplicate.

A wound was introduced on the confluent monolayer cells using a micropipette tip. Photographs were taken at 40 X magnification using phase-contrast microscopy immediately after wound incision and at selected time-points. Wound closure was measured by calculating pixel densities in the wound area by Cell^a software (Olympus Biosystem Gmb, Hamburg, Germany) and expressed as percentage of wound closure of triplicate areas ± SD.

A pool of 4 small inhibitor duplex RNAs (ON-TARGETplus siRNA SMARTpool) targeting human AGR2 (#L-003626-00), and a non-targeting pool (ON-TARGETplus Non-Targeting Pool) control (#D 001810-10-20) were purchased from Dharmacon RNAi Technologies (Dharmacon Inc., Chicago, IL, USA). Cells were grown under standard conditions. TPC-1 cells were plated in 60-mm dishes in complete medium without antibiotics and electroporated using MicroPorator (MP-100, Digital Bio, Euroclone, Milan, Italy) according to the manufacturer’s instructions. Cells were harvested at different time points after transfection, counted and analyzed for protein expression.

Immunoblotting was carried out according to standard procedures. Anti-AGR2 (PO1) monoclonal antibody was from Abnova (Taipei City, Taiwan); anti-GADD153 (B-3, sc-7351) and anti-α-tubulin monoclonal antibody was from Sigma-Aldrich. Secondary anti-mouse and anti-rabbit antibodies coupled to horseradish peroxidase were from Santa Cruz Biotechnology.

cDNA containing the complete coding sequence of human AGR2 was produced by reverse transcription of RNA from HT29 colon adenocarcinoma cells (American Type Culture Collection, Rockville, MD, USA) using the following primers:

AGR2 Cl For: 5′-ATCCTCGAGCGATCATGGAGAAAATTCCA-3′;

AGR2 Cl Rev: 5′-ATCGGATCCCAATTCAGTCTTCAGCAA-3′ (annealing temperature: 60°C).

AGR2 cDNA was cloned into pcDNA3.1 (named pcDNA) (Invitrogen) by using BamHI and XhoI restriction enzymes and verified by sequencing.

Point mutations in AGR2 that changed the cysteine at position 81 to serine (C → S) and glutamic acid at position 60 to alanine (E → A) were introduced in the pcDNA-AGR2 plasmid by the Stratagene QuikChange site-directed mutagenesis kit (Agilent Technologies, Inc. Life Sciences and Chemical Analysis Group, Santa Clara, CA, USA).

The following oligonucleotides were used:

AGR2 C81S For: 5′- CATCACTTGGATGAGTCCCCACACAGTCAAGC-3′;

AGR2 C81S Rev: 5′-GCTTGACTGTGTGGGGACTCATCCAAGTGATG-3′ (annealing temperature: 55°C).

AGR2 E60A For: 5′- CTGGACTCAGACATATGAAGCAGCTCTATATAAATCCAAGAC-3′;

AGR2 E60A Rev: 5′-GTCTTGGATTTATATAGAGCTGCTTCTTCATATGTGTCCAG-3′ (annealing temperature: 55°C).

The Nthy-ori 3–1 and the TPC-1 cells were transfected with pcDNA-AGR2, AGR2 (C → S), AGR2 (E → A) or the empty vector (pcDNA) using the Lipofectamine Reagent (Invitrogen) according to the instructions of the manufacturer. Two days later, G418 (Invitrogen) was added at concentration of 0.3 mg/ml (Nthy-ori 3–1) or of 1.2 mg/ml (TPC-1). Several clones and mass populations of transfectants were isolated, expanded and screened for AGR2 expression by Western blotting.

Nthy-ori 3–1 AGR2 transfected cells were treated with 1 mM disuccinimidyl suberate (DSS). Cells were harvested and washed 3 times with ice-cold PBS. The cross-linked cells were incubated for 60 minutes with mild shaking at room temperature, and then quenched by quenching buffer (1 M Tris, pH 7.5) to a final concentration of 10 mM for 15 min at room temperature to stop the chemical reaction. The chemically cross-linked cells were subjected to Western blotting analysis.

To measure the oxidized glutathione (GSSG) and reduced glutathione (GSH), the EnzyChrom™ GSH and GSSG Assay Kit (EGTT-100) (Bioassay Systems, Hayward, CA, USA) was used according to manufacturer’s instructions. Intracellular hydrogen peroxide levels were determined using the cell-permeable probe 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCFDA) that upon cleavage of the acetate groups by intracellular esterases and oxidation is converted to 2′,7′-dichlorofluorescein (DCF) (Invitrogen, Carlsbad, CA). DCF is a highly fluorescent compound that can be detected by fluorescence spectroscopy using excitation and emission wavelength of 485 nm and 538 nm, respectively. Cells were incubated with H₂DCFDA (100 μM) in complete media for 1 hour at 37°C. After that, cells were washed twice in PBS and fluorescence intensity was quantified by microplate reader (Perkin-Elmer Envision 2103 multilabel reader).

The disulfide isomerase activity was measured by the reduction of disulfide bonds in human insulin in the presence of DTT, giving rise to the aggregation of its β chain, a process that can be followed by turbidimetry[32]. The assay mixture was prepared in a cuvette by addition of 60 μl of insulin (10 mg/ml) to cell lysates in a buffer containing 100 mM sodium EDTA pH7 and 100 mM sodium phosphate pH 7 to give a final volume of 500 μl. The reaction started with the addition of 2 mM DTT and was followed kinetically by the increase of absorbance at 650 nm, due to insulin β chain precipitation. The absorbance at 650 nm was plotted.

Total RNA was isolated with the RNeasy Kit (Qiagen, Crawley, West Sussex, UK). The quality of the RNAs was verified by the 2100 Bioanalyzer (Agilent Technologies, Waldbronn, Germany); only samples with RNA integrity number (RIN) value > 7 were used for further analysis. One μg of RNA from each sample was reverse-transcribed with the QuantiTect® Reverse Transcription (Qiagen).

Quantitative RT-PCR was applied to study AGR2 expression in PTC samples (n = 10) and normal thyroid control (n = 10). RNA from each sample was reverse-transcribed with the QuantiTect® Reverse Transcription (Qiagen). For quantitative RT-PCR, we used the Human ProbeLibray™ system (Roche). AGR2 and RNA polymerase 2 primers sequences were:

AGR2 For: 5′-CTG GCC AGA GAT ACC ACA GTC-3′;

AGR2 Rev: 5′-AGT TGG TCA CCC CAA CCT C-3′;

RNA pol-For: 5′-GCGATGAGAACAAGATGCAA-3′;

RNA pol-Rev: 5′-CGCAGGAAGACATCATCATC-3′.

PCR reactions were performed in triplicate and fold changes were calculated with the formula: 2^{-(sample 1 ΔCt - sample 2 ΔCt)}, where ΔCt is the difference between the amplification fluorescent thresholds of the mRNA of interest and the mRNA of RNA polymerase 2 used as an internal reference. Sample 1 represented each single PTC sample, and sample 2 was the average of all (n = 10) normal thyroid samples.

Data are presented as mean ± standard deviations. Two-tailed unpaired Student’s t test was used for all statistical analysis. Mann–Whitney U test was used only for DNA array data analysis. P < 0.05 was considered statistically significant. Statistical analyses were carried out using the Graph Pad InStat (version 3.1a, La Jolla, CA, USA).