Cell culture and clinical samples

Human normal pancreatic ductal epithelial cells (HPDE) and the human pancreatic cancer cell lines ASPC1, SW1990, MIAPACA-2, CFPAC-1, and PANC-1 were obtained from the Chinese Academy of Sciences Committee Type Culture Collection cell bank and incubated in high glucose Dulbecco’s Modified Eagle’s Medium (DMEM, Thermo Fisher, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) at 37 °C in 5% CO2. A total of 90 patients who underwent operation at Shidong Hospital signed informed consent forms between January 2018 and December 2022. Whole blood samples were collected from patients and separation of plasma exosomes was done, according to the manufacturer’s protocol (4484450, Thermo Fisher, USA). The clinical characteristics of the 90 patients included in the study are shown in Table 1. The pathological criteria for PNR diagnosis were adapted from Liebig et al. [20]. A total of 60 paired and 30 unpaired PDAC specimens from patients were subjected to tissue microarray analysis (TMA). The experimental protocol was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Shidong Hospital.

Table 1 Baseline data for patients with pancreatic cancerPlasmid and oligonucleotide transfection

For lenti-Gli1 construction, HEK293T was used for virus packaging after the co-transfection of pWPXL-Gli1 with packaging plasmid psPAX2 and envelope plasmid pMD2.G (Addgene, city, state, USA) by Lipofectamine 3000 (Invitrogen, USA). The efficiency of transfection was examined by qPCR and Western blot analysis. The plasmid pcDNA3.1-CMV-circ-0001261 was constructed by GenePharma (Shanghai, China). Small interfering RNAs (siRNAs) against negative control, circ-0001261, and VGF, as well as siRNAs against miR-negative control and miR-451a, were designed by RiboBio (Guangzhou, China). Lipofectamine 3000 (Invitrogen, USA) was used to transfect cells, according to the manufacturer’s instructions. The sequences are depicted in Supplement 1.

Western blotting

Cells were mixed in a RIPA lysis buffer (Millipore, USA) and protease inhibitor cocktail (Roche, Germany) for 60 min before heating for 10 min at 100℃. The Bicinchoninic Acid (BCA) protein assay (Beyotime, China) was used to test protein concentration. A total of 50 μg protein was probed overnight at 4℃ with the indicated primary antibodies. The antibodies used in the study were as follows: Gli1 (ab217326, Abcam, UK), β-actin (66009–1-Ig, Proteintech, China), Ki-67 (ab16667, Abcam, UK), E-cadherin (60335–1-Ig, Proteintech, China), Caspase-3 (ab32351, Abcam, UK), VGF (26781–1-AP, Proteintech, China), CD63 (67605–1-Ig, Proteintech, China), TSG101 (67381–1-Ig, Proteintech, China), Calnexin (A4846, Abclonal, China), and fluorescence-conjugated secondary antibody (DyLight 800, Odyssey, USA). The Odyssey two-color infrared laser imaging system (LI-COR Biosciences, USA) was used to scan the membrane and assess staining.

Polymerase Chain Reaction (PCR)

Total RNA was extracted by Trizol reagent (Invitrogen, USA), in accordance with the manufacturer’s instructions. Total RNA (1,000 ng) was used to generate cDNA with SuperScript II reverse transcriptase and Oligo (dT) (Invitrogen, USA). The qRT-PCR experiments were conducted with a real-time PCR kit (Takara, China) and analyzed according to the equation 2−ΔΔCt [ΔCt = Ct-Ct (GAPDH)]. A fold change greater than 1 was defined as high, and a fold change ≤ 1 was defined as low. Primer sequences are shown in Supplement 1.

Isolation of DRG cells and cell image acquisition

We placed the newborn mice in iced PBS solution under sterile conditions. After decapitation, we inserted a micro-shear from the opening of the cervical spinal canal, removed the exposed spinal cord, carefully extracted the DRG with a fine microscope, and peeled off its capsule. The separated tissue blocks were stored in iced DMEM culture medium, and then collected into a culture dish containing 1 mg/mL collagenase solution for digestion at 37℃ overnight. After adding 10 drops of FBS to stop digestion, we gently blew and centrifuged at 1,000 rpm for 5 min to remove the supernatant. Then, the single cell suspension was inoculated in a 6-well plate for 4 to 6 h. The serum-free culture medium (neurobasal + B27 + glutamine + 20 ng/mL NGF) was used for culture. On the third day of inoculation, the mitotic inhibitor ARA-C was added to inhibit the proliferation of non-neural cells. Representative higher magnification images were taken with an inverted Nikon TE 200, and the total neurite length per neuron was automatically tested using the neurite outgrowth application in MetaMorph (Molecular Devices). Briefly, we captured five representative DRG images from different fields of each culture group and assembled them into a general view in MetaMorph software. Neurite length was automatically recognized, calculated, and exported through the MetaMorph's neural synapse analysis module for later analysis. Recombinant human NGF protein was purchased from Proteintech (HZ-1222, Wuhan, China).

Cell proliferation, invasion, and chemotaxis

We inoculated the cell suspension in a 96-well plate (approximately 100μL per well). The cells were treated after adhering to the plate. We then added 10μL CCK-8 solution (Dojindo, Japan) to each well, put the culture plate into the incubator for incubation for 2 h. The optical density (OD) at 450 nm wavelength was measured with a microplate reader.

We diluted the Matrigel with serum-free cold cell culture medium DMEM and added it to the upper chamber of 24-well Transwell. We incubated a Transwell insert (Thermo Fisher, USA) at 37℃ for at least 4 to 5 h. The obtained cells were washed with culture medium, and 200μL of cell suspension was added into the upper chamber at a density of 5 × 105 cells/mL. Subsequently, 600μL of cell culture medium was added to the lower chamber before incubation. After 24 h of culture, the cells remaining on the membrane were stained with 0.1% crystal violet for counting and analysis. In the chemotaxis experiment, PDAC and DRG cells were also added into the upper and lower chambers of Transwell at 1:1 ratio, and crystal violet showed the proportion of transmembrane penetrating cells.

Annexin-V/PI staining

DRG cells were incubated in six-well plates and treated as indicated. After treatment at room temperature for 15 min in Annexin-V/PI (BD Biosciences, USA), the cells were tested and analyzed by flow cytometry (BD FACSCanto, USA).

Exosome isolation and identification

Exosomal fraction from 30 mL of cell culture supernatant was extracted by ExoQuick™ Exosome Precipitation Solution, according to the manufacturer’s recommendations (System Biosciences, USA). Briefly, 1/2 volume of ExoQuick Solution was added to cell culture supernatant and samples were refrigerated at 4 °C overnight. The exosome was isolated by centrifugation at 10,000 g for 60 min. The isolated exosome was detected by Grainsize Analyzer (Zetasizer Nano ZS, UK). The measuring range of the zetasizer is from approximately 0.1 nm to 10 µm. The sizes and aggregational states of exosome were further examined using a TEM (JEM-2100, Japan).

For uptake experiments, purified exosomes were labeled using the PKH26 labeling kit (Sigma Aldrich, USA), according to the manufacturer’s instructions. In brief, exosomes were mixed in solvent C (containing PKH26 dye) and incubated for about 5 min. The labeled exosomes were then co-cultured with DRG cells for 24 h. After slides were stained with 4ʹ,6-diamidino-2-phenylindole for 10 min, they were visualized under a confocal laser scanning microscope.

RNA sequencing

The total RNA extracted from DRG and PDAC cells was depleted of linear RNA and ribosomal RNA. RNA was then fragmented and reverse transcribed. After linking to the sequencing adaptor, a library was obtained for the whole genome sequencing. The raw sequencing reads were mapped to the reference human genome through TopHat2 and TopHat-Fusion sequentially. Mapped reads (from TopHat2) and back-spliced reads (from TopHat-Fusion) were used to quantify the enrichment of each circRNAs candidate, indicated in RPM (reads per million mapped reads). The expression level of each circRNA was compared by DEG seq and the images were drawn by Sangerbox [21].

Luciferase reporter assay

Luciferase reporter constructs containing circ-0001261-miR-451a or VGF mRNA 3ʹUTR-miR-451a binding sequences (circ-0001261-WT or VGF-WT) or mutant sequences (circ-0001261-MUT or VGF-MUT) and miR-451a mimic or miR-NC were co-transfected with psiCHECK-2 dual-luciferase plasmid as well as the Renilla luciferase gene. Before luciferase reporter assay, DRG cells (1 × 105 cells/well) were seeded in 24-well plates for 24 h. After 48 h transfection, the cellular luciferase activity was examined by a luciferase reporter assay system (Promega, USA), in accordance with the manufacturer’s protocol.

In vivo study

Nude mice (males, 4 weeks old) were subcutaneously injected with lenti-Gli1 CFPAC-1 or CFPAC-1 cells (1 × 107cells/mouse) to form PDAC. With regard to the si-circRNA treatment, 1 × 108 plaque-forming units/100 mL lentivirus vectors of si-circ-0001261 were transduced into nude mice through tail injection. All animal experimental procedures were approved by the Institutional Animal Use Committee of Shidong Hospital. The tumor volume was measured every three days and the mice were killed three weeks after injection for further analysis of tumor tissues. The number of intra-PDAC nerves was analyzed by calculation of the average number of PGP9.5+ cells from five randomly selected PGP9.5 immunostaining fields per culture group.

Immunohistochemistry and immunofluorescence

For immunohistochemistry experiments, we used negative control (tissue without primary antibody) and positive control (human breast carcinoma) tissues. For immunofluorescence experiments, human peripheral nerve sheath tumor tissue (PNST) from Shidong Hospital was used as a representative Gli1 and PGP9.5 immunostaining positive control. The tumor tissue was embedded in paraffin, and the other part was dehydrated with sucrose and then embedded with optimal cutting temperature compound (OCT), and subsequently sliced for standby. After the paraffin section was dewaxed with xylene and dehydrated with alcohol, 3% H2O2 was added to remove endogenous catalase. After adding serum to block the reaction, we added primary and second antibodies, and finally used diaminobenzidine (DAB) to develop color and obtain images under the microscope. Tissue sections were incubated with rabbit monoclonal anti-Ki-67 (1:150; ab16667; Abcam), rabbit polyclonal anti-VGF (1:100; 26781–1-AP; Proteintech), rabbit polyclonal anti-S100B (1:400;15146–1-AP; Proteintech), mouse monoclonal anti-Gli1 (1:400; 66905–1-Ig; Proteintech), or anti-PGP9.5 (1:100; ab8189; Abcam). For quantification, blinded semi-quantitative scoring was used, and the results were quantified as negative (< 5%) and positive (≥ 5%).

Statistical analysis

All data are expressed as the mean ± standard deviation (SD) and were analyzed by SPSS 22.0 software (IBM, USA). Correlations between clinical categorical parameters and circRNA levels were evaluated by χ2 test. One-way analysis of variance (ANOVA) was used to compare differences among three groups. The student’s t-test was employed to compare grouped differences when the data followed a normal distribution and the Mann–Whitney test was recommended for the comparison of two independent groups. In all cases, a P-value < 0.05 was considered to be statistically significant.