A total of 62 primary pancreatic cancer tissues and 20 adjacent normal tissues located >2 cm away from cancer tissues were obtained from patients (38 males and 24 females; age range, 38–72 years; median age, 58.65 years) who underwent surgical resection at Shanxi Dayi Hospital Affiliated to Shanxi Medical University between January 2016 and June 2018. All tissue specimens were examined by pathologists and diagnosed as pancreatic ductal adenocarcinomas. No antineoplastic treatment was administered to the patients prior to the operation. The present study was reviewed and ethically approved by the Shanxi Dayi Hospital Ethics Committee. Informed written consent was obtained from each patient and his/her family.

Human pancreatic cancer cell lines (SW1990, PANC-1, BxPC-3 and Aspc-1) and human pancreatic ductal epithelial cell line (HPDE6c7) were from the American Type Culture Collection. All cell lines were cultured in RPMI 1640 medium (Gibco; Thermo Fisher Scientific, Inc.) containing 100 U/ml penicillin and 100 U/ml streptomycin, supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) in a humidified atmosphere of 5% CO₂ at 37°C.

Total RNA of pancreatic cancer cells was extracted using TRIzol^® reagent (Takara Biotechnology Co., Ltd.) according to the instructions of the manufacturer. A PrimeScript RT Reagent kit (Takara Biotechnology Co., Ltd.) was used for the reverse transcription of mRNA into cDNA at 37°C for 60 min and 85°C for 1 min. Gene amplification was then performed. qPCR was performed using a SYBR PrimeScript RT-PCR Kit (Takara Biotechnology Co., Ltd.) according to the instructions of the manufacturer. The reaction started at 95°C for 5 min, followed by 35 cycles of 95°C for 45 sec, 59°C for 30 sec, and 72°C for 45 sec. The quantification was performed using the 2^−ΔΔCq method (12), and β-actin was used as the internal reference gene to normalize the results. The primer sequences are presented in Table I.

The PANC-1 cell line (2×10⁵ cells/well) in logarithmic growth phase was inoculated in a 6-well plate for 24 h. HPA silencing vectors were constructed using three short hairpin (sh)RNAs (50 nM; Shanghai GeneChem Co., Ltd; Table II) specifically targeting HPA. The HPA overexpression vector was constructed using a recombinant plasmid encoding HPA (50 nM; cat. no. GV147; Shanghai GeneChem Co., Ltd.), and the empty vector was used as a negative control. The PANC-1 cell line was then transfected to construct a silent (shHPA group) or overexpressed (HPA group) cell line. The PANC-1 cell line was transfected with an empty vector containing the GFP sequence as a control (vector group). FGF2 silencing vectors were constructed using (sh)RNA (5′-ACTACAATACTTACCGGTCAA-3′) (50 nM; Shanghai GeneChem Co., Ltd) specifically targeting FGF2. The PANC-1 cell line was then transfected to construct a silent (shFGF2 group) cell line. The SDCI inhibitor synstatin (100 nM; Beyotime Institute of Biotechnology) was added to the HPA group to form the HPA + synstatin group. The FGF2 receptor (FGF2R) inhibitor AZD4547 and the PI3K/Akt-specific inhibitor LY294002 were purchased from Sigma-Aldrich; Merck KGaA. PANC-1 cells were divided into the following groups: i) FGF2 group [PANC-1 cell line transfected with FGF2 recombinant plasmid (100nM; cat. no. QCP5411; QCHENG BIO Inc.)]; ii) AZD4547 group (3.0 nmol-l AZD4547 added to the FGF2 group); iii) LY294002 group (10 µmol-l LY294002 added to the FGF2 group); and iv) control group [PANC-1 cell line transfected with empty plasmid (cat. no. QCP5400; QCHENG BIO Inc.]. Transfection was performed using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the instructions of the manufacturer. Subsequent experiments were performed 48 h after transfection.

Tumor tissues were fixed with 10% formalin solution for 15 min at room temperature, and then blocked in wash buffer containing 5% normal goat serum (Shanghai GeneChem Co., Ltd.) for 2 h at room temperature. Tumor tissues were incubated with rabbit anti-human HPA primary antibody (1:100; cat. no. PB0405) and rabbit anti-human FGF2 primary antibody (1:100; cat. no. PB0916; both Boster Biological Technology) at 4°C overnight, and then incubated with goat anti-rabbit IgG secondary antibody (1:200; cat. no. BA1003; Boster Biological Technology) for 2 h at room temperature. The tissue was then stained with diaminobenzidine, and the nuclei were stained with hematoxylin for 5 min at room temperature and then observed using an optical microscope (Olympus Corporation) at ×200 magnification. Immunohistochemical staining was performed according to the protocols of the manufacturer (Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd.; OriGene Technologies, Inc.). The following scoring system was used to evaluate the staining results: The staining intensity of HPA and FGF2 was rated 0–3 (0, negative; 1, weak; 2, moderate; or 3, strong), while marker expression was evaluated according to the percentage of the stained area: 0, 0–10%; 1, 11–25%; 2, 26–50%; or 3, 50–100%. The final result was calculated as the sum of the two scores. The rating criteria were as follows: 0–3, low expression; and 4–6, high expression. The results were evaluated by two pathologists and the differences were re-evaluated until a consensus was reached.

Total protein was extracted by lysing the cells with RIPA buffer (Beyotime Institute of Biotechnology) and the protein concentration was measured using a BCA protein assay kit (Beyotime Institute of Biotechnology). A total of 50 µg protein from each sample was separated by SDS-PAGE (10% gel) and the proteins were transferred to a nitrocellulose membrane by electrophoretic transfer. After blocking with 5% skimmed milk for 2 h at room temperature, the membranes were incubated with the following primary antibodies: Rabbit anti-human HPA (1:1,000; cat. no. PB0405), rabbit anti-human FGF2 (1:1,000; cat. no. PB0916), rabbit anti-human AKT (1:5,000; cat. no. A00024-1), rabbit anti-human E-cadherin (1:100; cat. no. BA0475), rabbit anti-human N-cadherin (1:100; cat. no. BM3921), rabbit anti-human vimentin 1:3,000 (cat. no. PB0378; all Boster Biological Technology) and rabbit anti-human palladin (1:100; cat. no. PA5-65160; Thermo Fisher Scientific Inc.) were added according to the instructions of the manufacturer and incubated at 4°C overnight. β-actin (1:500; cat. no. BA2305; Boster Biological Technology) was used as an internal loading control. The goat anti-rabbit IgG secondary antibody (1:5,000; cat. no. BA1056; Boster Biological Technology) labeled with horseradish peroxidase was added to the membrane after washing and incubated at room temperature. The gray value of the protein bands was analyzed by ImageJ software (v1.8.0; National Institutes of Health) and the ratio of the gray value of the target band to that of the internal reference β-actin was representative of the expression level of the protein. All assays were performed in triplicate.

The invasive ability of PANC-1 cells was measured using Transwell chambers (Corning Inc.). PANC-1 cells (2×10⁵ cells/ml) starved for 24 h were inoculated into the upper chamber of a 24-well chamber, and each group was analyzed in 6 replicate wells. A total of 50 µl Matrigel solution was placed on the polycarbonate microporous membrane between the upper and the lower chambers, 100 µl RPMI 1640 medium was added to the upper chamber, and a mixture of 600 µl RPMI 1640 and 20% fetal bovine serum was added to the lower chamber. After incubation for 24 h at room temperature, the cells were fixed with a 4% paraformaldehyde solution for 30 min at room temperature, followed by rinsing with PBS and staining with a 0.5% crystal violet dye solution for 10 min at room temperature. Subsequently, the lower chamber was observed and photographed with a light microscope at ×200 magnification, and cell counting was performed.

PANC-1 cells were seeded in a 6-well plate at an adjusted concentration of 1×10⁶ cells/ml. The experiment was performed when the cells reached 100% confluence. A wound was created in the center of the plate with the tip of a 100-ml pipette. The scratched cells were washed away in serum-free medium and this process was repeated 3 times, and then cultured in regular medium. The wound was observed at ×200 magnification with an inverted microscope at 0 and 48 h, and images were captured.

SPSS software (version 23.0; IBM Corp.) was used for statistical analysis throughout the study. Statistical analysis of HPA and FGF2 expression between pancreatic cancer tissues and paired adjacent normal tissues was evaluated using a paired Student's t-test. One-way analysis of variance was performed on ≥3 sets of data using Tukey's post hoc test. GraphPad Prism software (version 8.0; GraphPad Software, Inc.) was used to generate graphical representations. Each experiment was repeated ≥3 times. The experimental results are presented as mean ± SD values. P<0.05 was considered to indicate a statistically significant difference.

In order to investigate the expression of HPA and FGF2 in pancreatic cancer tissues and cells, immunohistochemical staining and RT-qPCR analysis were performed. According to the results of immunohistochemistry, significant upregulation of HPA and FGF2 was observed in pancreatic cancer tissues compared with adjacent non-tumor tissues (Fig. 1A). Based on the aforementioned scoring criteria, the details of HPA and FGF2 expression are presented in Table III. The results of RT-qPCR analysis revealed that HPA and FGF2 expression levels were significantly higher in pancreatic cancer cell lines than in the HPDE6c7 cell line (Fig. 1B). These results implied that the expression of HPA and FGF2 in pancreatic cancer was markedly upregulated.

To further explore whether elevated HPA and FGF2 expression levels are directly associated with the clinicopathological characteristics of pancreatic cancer, the expression of HPA and FGF2 in different pancreatic cancer tissues was examined via immunohistochemistry. The results revealed that the positive rate of HPA in tissues with lymph node metastasis was significantly higher than that in tissues without lymph node metastasis. Tumor-node-metastasis (TNM) stage III–IV tissues were found to have higher HPA-positive expression than TNM stage I–II tissues. Positive expression of HPA was not associated with the age, sex, tumor location or tumor differentiation status of patients with pancreatic cancer. By contrast, the positive expression rate of FGF2 was observed to increase as the degree of tumor differentiation decreased. Moreover, similarly to HPA, the age, sex or tumor location of the patients with pancreatic cancer were not associated with FGF2 expression (Table IV).

According to the aforementioned experimental results, the expression of HPA and FGF2 in pancreatic cancer cells was significantly increased. For the purpose of determining whether the expression of FGF2 was affected by HPA, the expression of FGF2 in the shHPA, HPA and vector groups was compared. Among the three PANC-1 shHPA cell lines selected, the sh1 group had the highest silencing efficiency, compared with the sh2 and sh3 groups (Fig. 2A and B). The RT-qPCR and western blot analyses demonstrated that the expression of FGF2 was significantly increased in the HPA group compared with the vector group. On the contrary, HPA silencing significantly inhibited the expression of FGF2 in the shHPA group compared with the vector group (Fig. 2C and D). Taken together, these results suggested that HPA is able to increase the expression of FGF2, and that HPA silencing results in the downregulation of FGF2 in pancreatic cancer cells.

By binding to SDC1, HPA removes this proteoglycan from the cell surface and promotes its expression (13). To confirm the study hypothesis that HPA affects the expression of FGF2 in pancreatic cancer cells by binding to SDC1, RT-qPCR was performed. The results revealed that SDC1 expression was increased in pancreatic cancer cell lines compared with the HPDE6c7 cell line. (Fig. 3A). Overexpression of HPA resulted in a significant increase in SDC1 expression in the HPA group compared with the vector group, whereas HPA silencing considerably suppressed the expression of SDC1 (Fig. 3B). Moreover, the expression of FGF2 was markedly decreased in the HPA + synstatin group compared with that in the vector group (Fig. 3C). Taken together, these findings indicated that HPA may promote the expression of FGF2 by upregulating SDC1 in pancreatic cancer cells. Notably, neither overexpression nor silencing of FGF2 inflenced HPA expression (Fig. S3)

A previous study demonstrated that Palladin is highly expressed in pancreatic cancer cells (14). Moreover, the PI3K/Akt pathway has been observed to maintain the stability of Palladin and promote its expression (15). It was hypothesized that FGF2 promotes the expression of Palladin via the PI3K/Akt pathway, thereby promoting the invasion and migration of pancreatic cancer cells. According to the western blot results, the expression of Palladin in pancreatic cancer cells was significantly higher than in the HPDE6c7 cell line (Fig. 4A). This result was further confirmed via RT-qPCR analysis (Fig. 4B).

Successful overexpression of FGF2 in FGF2 group cells was determined via RT-qPCR and western blot analyses (Fig. S1). According to the western blot analysis, the expression of Palladin and Akt proteins in the FGF2 group was significantly increased compared with the other groups (Fig. 4C and D). This result was also confirmed by RT-qPCR (Fig. S2). The invasive and migratory abilities of pancreatic cancer cells were assessed by Transwell chamber and wound healing assays. When FGF2 was overexpressed in the FGF2 group, the invasive and migratory abilities of pancreatic cancer cells were significantly enhanced compared with the control group. However, these abilities of pancreatic cancer cells were markedly reduced in the AZD4547 and LY294002 groups compared with in the FGF2 and control groups. (Fig. 4E-G). These results indicated that FGF2 may regulate the invasion and migration of pancreatic cancer cells via the PI3K/Akt signaling pathway, thus promoting the progression of pancreatic cancer.

EMT is closely associated with cell migration and invasion. To examine the effects of HPA and FGF2 on EMT, western blot analysis was conducted in order to detect the markers of EMT. The expression of E-cadherin was found to be inhibited, whereas the expression of N-cadherin and vimentin was considerably increased after HPA overexpression; these effects were reversed by adding AZD4547 (Fig. 4H). These results demonstrated that HPA may activate the EMT process by upregulating the expression of FGF2.