Human PDAC cell lines (Capan-2, PANC-1, SW1990, COLO357 and MIAPaCa-2) and normal human pancreatic duct epithelial (HPDE) cell line were obtained from Shanghai Cell Bank (Shanghai, China). PDAC cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS). HPDE cells were maintained in keratinocyte serum-free medium (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with bovine pituitary extract and epidermal growth factor. All cells were cultured at 37°C in humidified incubator with 5% CO₂.

The present study was approved by the Medical Ethics Committee of LanLing County Hospital. All patients included in this research were required to provide written informed consent. Fifteen paired PDAC and adjacent normal tissues were obtained from patients who underwent surgical resection in LanLing County Hospital. All tissue samples were snap-frozen in liquid nitrogen immediately after surgery and stored at −80°C until RNA extraction.

shRNA plasmids for human HMGN5 were designed against the HMGN5 gene and constructed in Phblv-u6-puro vectors (Shanghai GeneChem Co., Ltd., Shanghai, China). The shRNA sequences used in the present study were as follows: shHMGN5 5′-ATGAGAAAGGAGAAGATGC-3′ and shcontrol: 5′-GAAGAATATCGAAGGAAGA-3′. To generate stable HMGN5 knockdown cells, PANC-1 cells were grown in 6-well plates until they reached 60% confluency. The medium was replaced with 1 ml of fresh culture medium supplemented with 100 µl viral supernatant (1×10⁸ UT/ml) and 8 µg/ml Polybrene for 24 h. The PANC-1 cells were further cultured in medium containing puromycin at 3 µg/ml. Individual puromycin-resistant colonies were isolated during drug screening. The open reading frame of β-catenin was inserted into pcDNA3.1 vector to generate pcDNA3.1/β-catenin overexpression vector. Transient transfection was performed with a Lipofectamine 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's instructions. Knockdown or overexpression efficiency was examined via western blot assay.

PANC-1 cells transfected with shHMGN5 or shcontrol and β-catenin plasmid were seeded at 200 cells/well in 6-well plates. Seven days later, the colonies were stained with crystal violet, photographed, and then scored.

Cell viability was determined via Cell Counting Kit-8 (CCK-8) (Beyotime Institute of Biotechnology, Haimen, China). Briefly, cells (3,000/well) transfected with shHMGN5 or shcontrol and β-catenin plasmid in 200 µl medium were plated into 96-well plates. Zero, 24, 48 and 72 h later, 10 µl of CCK-8 solution was added into each well. After 2 h, the optical density values of each well were measured at 450 nm using a microplate reader (Thermo Fisher Scientific, Inc., Waltham, MA, USA). CCK-8 assay were repeated 3 times.

Cell metastasis was determined using Boyden chamber assay. For the invasion assay, the upper sides of the filters were coated with 50 µl Matrigel (BD Biosciences, Bedford, MA, USA). Cells (5×104) with 200 µl of serum-free medium were seeded in the upper chamber. The lower chamber was filled with medium supplemented with 5% FBS. Following incubation at 37°C with 5% CO₂ for 8 h (migration) or 12 h (invasion), cells on the lower filter were fixed with methanol, stained with crystal violet, and then counted under a light microscope.

Equal amounts of the protein from lysates of PDAC cells were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to polyvinylidene fluoride (PVDF) membranes. The immunoreactive bands were firstly incubated with the primary antibodies, including HMGN5 (1:500), wnt1 (1:500), total or phosphorylated β-catenin (1:500), (all from Cell Signaling Technology, Inc., Danvers, MA, USA) E-cadherin (1:1,000), N-cadherin (1:1,000), Vimentin (1:1,000) (all from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), β-actin (1:2,000; Beyotime Institute of Biotechnology) overnight at 4°C and horseradish peroxidase-conjugated goat anti-rabbit antibody (Santa Cruz Biotechnology, Inc.) at room temperature for 2 h. The intensity of protein bands was detected by Image-Pro Plus 6.0 software. β-actin served as the loading control.

Total RNA from cells was extracted using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and then reverse-transcribed through a reverse transcription kit (Takara Biotechnology Co., Ltd., Dalian, China) following the manufacturer's protocol. RT-qPCR was conducted with All-in-One^™ miRNA RT-qPCR Detection Kit (AOMD-Q020; GeneCopoeia, Inc., Rockville, MD, USA) on CFX96^™ Real-Time PCR Detection System supplied with analytical software (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The primers used for amplification were: HMGN5, forward 5′-CATGGACATGAGCACAATCA-3′ and reverse 5′-CCCTTTTCTGTGGCATCTTC-3′; E-cadherin, forward 5′-GTACTTGTAATGACACATCTC-3′ and reverse 5′-TGCCAGTTTCTGCATCTTGC-3′; N-cadherin, forward 5′-ATCAAAGACCCATCCACC-3′ and reverse 5′-CCTCCTCACCACCACTA-3′; vimentin, forward 5′-CTTCCGCGCCTACGCCA-3′ and reverse 5′-GCCCAGGCGACCTACTCC-3′; β-actin, forward 5′-GATCATTGCTCCTCCTGAGC-3′ and reverse, 5′-ACTCCTGCTTGCTGATCCAC-3′.

The protocols for animal experiment study were approved by the Animal Care and Welfare Committee of LanLing County Hospital and conducted in strict accordance with the guidelines of the National Animal Welfare Law of China. Four-week-old male BALB/c nude mice (n=12) were purchased from the Laboratory Animal Center of JiLin University (JiLin, China) and maintained in a SPF environment. Stable PANC-1 shHMGN5 and PANC-1 shcontrol cells (1×10⁶) were subcutaneously injected into the right flanks of athymic nude mice. Tumor volume (mm³) was calculated every 7 days for 35 days using the formula: V = 0.5 × length × width². After 7 weeks, the mice were euthanized. The tumors were isolated, weighed, photographed.

Paraffin-embedded tissues were cut into 4 µm-thick consecutive sections and were then dewaxed in xylene and rehydrated in graded ethanol solutions. Antigen retrieval was performed following the standard procedure. Sections were cooled and immersed in a 0.3% hydrogen peroxide solution for 15 min to block endogenous peroxidase activity, and then rinsed in phosphate-buffered saline (PBS) for 5 min. Non-specific labeling was blocked by incubation with 5% bovine serum albumin at room temperature for 30 min. Sections were then incubated with primary rabbit anti-human antibody against Ki67 (1:200; Santa Cruz Biotechnology, Inc.) at 4°C overnight, rinsed with PBS with Tween-20 (PBST), incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG secondary antibody (Santa Cruz Biotechnology, Inc.). The immunohistochemistry results were scored by the percentage of positive detection of the staining.

All experiments were repeated 3 times. Unless otherwise indicated, experimental values are expressed as mean ± standard error mean. Differences between 2 groups were assessed using Student's t-test (two-tailed). Data of >2 groups were analyzed using one way analysis of variance with Tukey's post hoc test. Statistical analyses were performed using SPSS 13.0 software (SPSS, Inc., Chicago, IL, USA). P<0.05 was considered to indicate statistical significance.

To investigate the potential role of HMGN5 in PDAC, we first examined the expression level of HMGN5 in the PDAC cell lines Capan-2, PANC-1, SW1990, COLO357 and MIAPaCa-2, as well as in the normal pancreatic epithelial cell line HPDE by qPCR and western blot assay. The results showed that the mRNA and protein expression levels of HMGN5 were significantly higher in PDAC cells compared to HPDE cells (Fig. 1A and B). The highest expression levels of HMGN5 were detected in PANC-1 cells. Furthermore, the expression of HMGN5 was detected in PDAC tissues and matched normal tissues from 15 patients using qPCR. The results showed that the expression level of HMGN5 was higher in PDAC samples than in adjacent normal tissues (Fig. 1C). The increased levels of HMGN5 in PDAC indicate that HMGN5 may serve as oncogene in PDAC.

To examine the role of HMGN5 in human PDAC cell growth, PANC-1 cells were transfected with shcontrol or shHMGN5. Western blot results showed that the expression level of HMGN5 was significantly downregulated in cells transfected with shHMGN5 (Fig. 2A). CCK-8 assay showed that HMGN5 silencing in PANC-1 cells significantly inhibited cell viability (Fig. 2B). Colony formation assays illustrated that HMGN5 silencing in PANC-1 cells significantly inhibited cell proliferation (Fig. 2C and D).

Cell metastasis is critical events in the tumor progression. Therefore, the effect of HMGN5 on the migration and invasion of PDAC cell was explored in vitro. PANC-1 cell migration and invasion were investigated through Boyden chamber assay. Results showed that HMGN5 silencing significantly decreased the cell migration and invasion (Fig. 3A-D).

To show the HMGN5 biofunction in promotion of PDAC cell tumorigenicity in vivo, PANC-1 cells stable expression shcontrol or shHMGN5 was respectively injected into nude mice. The tumor volume was measured every 7 days. At 35 days after cell implantation, the tumors were removed and photographed (Fig. 4A). Tumor volume and the final weight of excised tumors demonstrated that HMGN5 knockdown had significantly retarded tumor growth at termination (Fig. 4B and C). Histologic analysis of tumor proliferation showed that HMGN5 silencing had significantly fewer Ki-67 positive cells compared to shcontrol tumors (Fig. 4D and E). These data suggested that HMGN5 may play a critical role in PDAC proliferation and tumor maintenance in vivo.

EMT is a vital biological process involved in cell differentiation in normal embryonic development and is a potential mechanism for tumour cell metastasis. To address whether HMGN5 has effects on EMT, The expression of the epithelial marker E-cadherin and mesenchymal markers N-cadherin and vimentin was determined in PANC-1 cell transfected with shcontrol and shHMGN5 by western blotting and qPCR. The results showed that the expression of E-cadherin was significantly up-regulated compared with the controls, whereas N-cadherin and vimentin was significantly down-regulated in PANC-1 cell (Fig. 5B-E).

To confirm the above data in vivo, we detected the expression of E-cadherin, N-cadherin and vimentin in the tumour transfected with shcontrol and shHMGN5 by western blotting. The results showed that the expression of E-cadherin was significantly up-regulated in shHMGN5 group, whereas N-cadherin and vimentin was significantly down-regulated in shHMGN5 group (5F).

Wnt/β-catenin pathway is one of the major signaling pathways involved in EMT and plays an important role in metastasis. Western blot results showed that HMGN5 positively regulated Wnt1 expression in protein (Fig. 5A). Furthermore, we showed the activation of GSK3β and β-catenin, the canonical pathway of Wnt signaling pathway. The western blot results showed that the amount of p-GSK3β (Fig. 5A) and active β-catenin (dephosphorylated β-catenin, Fig. 5A) was dramatically reduced by HMGN5 silencing.

We then investigated whether rescuing HMGN5 silencing-mediated-Wnt/β-catenin signaling pathway suppression with a β-catenin overexpression plasmid attenuates the progression inhibition effect of HMGN5 silencing on PDAC cells. The expression of β-catenin was detected by western blot assay (Fig. 5A). The CCK-8 (Fig. 2B) and colony formation assay (Fig. 2C and D) results showed that restoration of β-catenin expression reversed the inhibition of cell viability and proliferation by HMGN5 silencing in PANC-1 cell. Furthermore, the Transwell assay showed that β-catenin overexpression significantly reversed the inhibition of cell migration and invasion by HMGN5 knockdown in PANC-1 cell (Fig. 3A-D). Taken together, these results demonstrated that HMGN5 promoted the PDAC progression via activation of Wnt/β-catenin signaling pathway.

We further explored the effect of regulatory relationship of HMGN5 and Wnt/β-catenin signaling pathway on EMT of PANC-1 cells. qPCR and western blot assay results showed that the overexpression of β-catenin counteracted the effect of HMGN5 on EMT as demonstrated by changed expressions of molecular markers associated with EMT (Fig. 5B-E).