Previously, we identified 161 DEPs with a ≥1.5-fold change (FC) in PC via proteomic analyses (7). The effect of DEP expression on the prognosis of patients with PC was evaluated using TCGA, GTEx, and GEO databases (32–34). Due to the limited number of normal samples in TCGA, TCGA and GTEx were combined to form the TCGA_GTEx group. Raw RNA-seq data for the TCGA_GTEx group were acquired from the UCSC database, and included 178 PC samples and 171 pancreas samples (35). To reduce bias associated with the use of a single database, two GEO datasets were selected for further analysis: GSE62165 and GSE15471. Stepwise, all data were log2 (x + 1) transformed and normalized via the “LIMMA” package (36). Differentially expressed genes (DEGs) were obtained using the “LIMMA” package with thresholds of |log2 FC| > 1, with adjusted p < 0.05. Differentially expressed miRNAs (DEMIs) were screened using the “edgR” package with the criteria of |log2FC| > 1 and adjusted p < 0.05 (37). Additionally, the Kaplan–Meier analysis was applied to evaluate OS. Univariate and multivariate Cox proportional hazards regression (CPHR) analyses were used to evaluate the impact on survival. Only common protein expression was consistent with prognosis and a p-value lower than 0.05 was regarded as survival-related proteins. Finally, SDCBP was selected for subsequent analysis.

Gene set enrichment analysis (GSEA) was performed to determine SDCBP function in PC. The potential function of SDCBP was determined through Gene Ontology (GO) term analysis and Kyoto encyclopedia of genes and genomes (KEGG) pathway analysis with a cutoff criterion of p < 0.05 (38, 39). Next, the top five most enriched GO terms and KEGG pathways were identified. In addition, upstream miRNAs of SDCBP were predicted by TargetScan and miRNet databases (40, 41). To improve reliability, miRNAs with low expression in PC were selected for further analysis. Additionally, we evaluated OS through Kaplan–Meier analysis. Finally, correlation analysis was applied to calculate the correlation index of SDCBP with miRNAs with a cutoff criterion of p < 0.05. The qualified miRNAs were selected for further analysis.

To detect the prognostic ability of SDCBP in PC, we collected 61 paraffin-embedded samples for which complete clinical data were available from Shengjing Hospital of China Medical University between 2013 and 2016. All samples were obtained from patients who underwent radical surgery for PC, with diagnosis confirmed by a pathologist. Patients with complete clinicopathological data and follow-up data were included. Additionally, to determine the expression of SDCBP through Western blot (WB) and quantitative polymerase chain reaction (qPCR) assays, a further 12 pairs of fresh PC and normal tissues were obtained between 2019 and 2020. All specimens were collected following surgical resection and immediately frozen at −80°C until use. All patients signed written informed consent before enrollment, and the project was approved by the Ethics Committee of Shengjing Hospital of China Medical University (2020PS617K).

Cell lines were purchased from the National Collection of Authenticated Cell cultures (Shanghai, China) and included Aspc-1, BxPC-3, Capan-2, PANC-1, and SW1990 cell lines, and the pancreatic cell line HPDE. The HPDE cell line was established from ductal epithelial cells of the pancreas. The Aspc-1 cell line was established using metastatic cells from PC ascites carrying CDKN2A, FBXW7, KRAS, MAP2K4, and TP53 mutations. The BxPC-3 cell line was established from primary PC carrying mutations of CDKN2A, MAP2K4, SMAD4, and TP53. The Capan-2 cell line was established from primary PC carrying KRAS mutation. The PANC-1 cell line was established from primary PC carrying a KRAS and TP53 mutation. The SW1990 cell line was established from PC using spleen metastatic cells containing CDKN2A and KRAS mutations (42, 43). Aspc-1, BxPC-3, Capan-2, and SW1990 cells were cultured in RPIM1640 medium containing 10% fetal bovine serum (FBS, Gibco, USA), whereas PANC-1 and HPDE cells were maintained in DMEM with 10% FBS. All cells were cultured at 37°C in a humidified incubator supplemented with 5% CO₂.

SDCBP expression in tissues was determined using a two-step immunohistochemistry kit (E-IR-R217, Elabscience, USA), according to the manufacturer’s instructions. Briefly, tissue slices were dewaxed with xylene and hydrated through graded ethanol. Then, slices were heated to retrieve antigens, covered with 3% hydrogen peroxide to inactivate endogenous enzymes, and blocked with normal goat serum. After this, slices were incubated overnight at 4°C with anti-SDCBP (1:200, Proteintech, USA) and incubated with polyperoxidase anti-Rabbit IgG at 37°C for 30 min. Finally, 3,3-diaminobenzidine reagent was applied during the color development stage to visualize staining, and slices were sealed with neutral balsam. Staining intensity was graded as 0 (negative), 1 (weak), 2 (medium), or 3 (strong). Staining extent was scored according to the percentage of positive cells as follows: 0 (<5%), 1 (5%–25%), 2 (26%–50%), 3 (51%–75%), and 4 (>75%). Three fields were randomly selected and used to calculate the average score. The intensity and extent scores were multiplied as the final staining scores (0–12). A final score of >2 was considered as high expression, while others were considered as low expression.

For WB analysis, samples were lysed in RIPA lysis buffer (Solarbio, China) containing 1% protease inhibitor (Solarbio, China) and 1% phosphatase inhibitor (Solarbio, China), and centrifuged at 14,000 rpm for 30 min at 4°C. Protein concentration was determined using the BCA protein assay Kit (Solarbio, China). Then, 30 μg of total protein was loaded subjected to SDS-PAGE (10%) and transferred to a polyvinylidene difluoride membrane (Millipore, USA). The membranes were blocked with 5% bovine serum albumin (BSA) for 2 h at room temperature, incubated with primary antibodies overnight at 4°C, and incubated with secondary antibodies for 2 h at room temperature. Finally, protein bands were visualized using an ECL detection kit under a luminescent image analyzer. The following antibodies were used: anti-SDCBP (1:1,000, Proteintech, USA), anti p-PI3K (1:1,000, Affinity, USA), anti-PI3K (1:1,000, Affinity, USA), anti p-AKT (1:2,000, Affinity, USA), anti-AKT (1:5,000, Proteintech, USA), anti E-cadherin (1:5,000, Proteintech, USA), anti N-cadherin (1:5,000, Proteintech, USA), anti-GAPDH (1:10,000, Proteintech, USA), and anti-β-actin (1:10,000, Proteintech, USA).

Total RNA was extracted with RNAiso Plus reagent (Takara, Japan) and dissolved in 10 μl of RNase-free water. The concentration and purity of total RNA were measured using a NanoPhotometer 50 (Implen, Germany). For mRNA reverse transcription, 1.0 μg of the total RNA was subjected to the HiScript III RT SuperMix (Vazyme, China). For miRNA reverse transcription, the miRNA 1st Strand cDNA Synthesis Kit (Vazyme, China) was used. The reaction conditions were as follows: 37°C for 15 min, 85°C for 5 s, followed by cooling to 4°C. Then, cDNA was subjected to qPCR on an ABI 7500 RT-PCR system (Thermo Fisher, USA). The amplification system consisted of 2 μl of cDNA, 0.4 μl each of forward and reverse primer, 10 μl of ChamQ Universal SYBR qPCR Master Mix for mRNA or miRNA Universal SYBR qPCR Master Mix for miRNA, and ddH₂O up to a final volume of 20 μl. The amplification conditions involved mRNA denaturation at 95°C for 30 s or miRNA denaturation at 95°C for 5 min, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. Primer specificity was determined by melting curve analysis. GAPDH and β-actin were used as endogenous reference genes for mRNA in tissue and cell PCR, respectively, and U6 was used as an endogenous reference for miRNA. Relative expression levels were quantified using the 2^−ΔΔCt method. The primers were synthesized by Sangon (China), and the sequences are as follows: SDCBP: 5’-CTG CTC CTA TCC CTC ACG ATG-3’ (Forward), 5’-GGC CAC ATT TGC ACG TAT TTC T-3’ (Reverse). GAPDH: 5’-GTC TCC TCT GAC TTC AAC AGC G-3’ (Forward), 5’-ACC ACC CTG TTG CTG TAG CCA A-3’ (Reverse). β-actin: 5’-CAC CAT TGG CAA TGA GCG GTT C-3’ (Forward), 5’-AGG TCT TTG CGG ATG TCC ACG T-3’ (Reverse). hsa-miR-216b: 5’-AAA UCU CUG CAG GCA AAU GUG A-3’ (Forward), 5’-TTT AGA GAC GTC CGT TTA CAC T-3’ (Reverse). RT Primer: 5’-GTC GTA TCC AGT GCA GGG TCC GAG GTA TTC GCA CTG GAT ACG ACT CAC AT-3’. U6: 5’-CGC AAG GAT GAC ACG CAA AT-3’ (Forward), 5’-CGG CAA TTG CAC TGG ATA CG-3’ (Reverse).

SDCBP-shRNA lentiviral vectors (hU6-MCS-CBh-gcGFP-IRES-puromycin) were purchased from GeneChem (Shanghai, China). Capan2 and SW1990 cells were transfected according to the manufacturer’s instructions. The cells were subjected to puromycin selection for 1 week, after which the surviving cells were used for further experiments. For miRNA transfection, the negative control (NC), agomir, and antagomir of hsa-miR-216b (agomiR-216b and antagomiR-216b) were synthesized from GenePharma (Shanghai, China), and transfected into Capan2 and SW1990 cells using Lipofectamine 3000 (Invitrogen, USA). For the luciferase assay, we identified hsa-miR-216b as a direct upstream miRNA in SDCBP and predicted its binding sites using the TargetScan database. The wild-type 3ʹ-untranslated region (UTR) of SDCBP and mutant 3ʹ-UTR SDCBP were constructed by GenePharma and cloned into the pmir-GLO-luciferase reporter plasmid. Capan2 and SW1990 were co-transfected with wild-type and mutant pmir-GLO-SDCBP, and miR-216b mimics or NC using Lipofectamine 3000. After 48 h, cell lysis buffer was collected, and luciferase activity was determined using the Dual-Luciferase Reporter Gene Assay Kit (GenePharma, China). Luciferase activity was normalized to that of Renilla.

The cell counting kit-8 (CCK-8) assay is used widely to evaluate cell proliferation and cytotoxicity (44, 45). Cell proliferation was evaluated by the CCK-8 assay according to instructions (Solarbio, China). Briefly, cells were transfected, plated, and grown in 96-well plates. Then, cell viability was determined at 0, 24, 48, 72, and 96 h using CCK-8. The absorbance was determined at 450 nm using a microplate reader (BioTek, USA). All samples were run in triplicate.

For the colony formation assay, 1,000 cells were seeded in six-well plates and then cultured for 1 week. Next, colonies were fixed with 4% paraformaldehyde for 1 h and stained with 1% crystal violet for 1 min. Only clones containing more than 50 cells were counted and analyzed.

Cell migration and invasion were evaluated through wound-healing and Transwell invasion assays, respectively. To assess wound healing, cells were plated on six-well plates and cultured until they reached 90%–100% confluency. Then, the cells were scraped using a sterile 200-μl pipette tip. Scratched areas were photographed after 0 and 24 h at 100× magnification using an inverted microscope (Nikon DS- Ri2, Japan). For the Transwell invasion assay, 800 μl of medium containing 10% FBS was added to the lower chamber of a Transwell plate as a chemoattractant, and 200 μl of serum-free medium was added to the upper chamber, which was coated with Matrigel (Corning, USA). Then, cells were seeded onto the upper chamber and plates were incubated for 24 h.

Cells traversing the membranes were fixed and stained with a Hematoxylin-Eosin Staining Kit (Solarbio, China), according to the manufacturer’s instructions. Images of cells were captured using a microscope (Nikon E800) at 200× magnification, and five random fields were counted and analyzed.

The protocol for the animal experiment was approved by the Ethics Committee at Shengjing Hospital of China Medical University (2020PS617K). Sixteen female BALB/c nude mice (8 weeks old, 18-20 g) were purchased from HFk Bioscience (Beijing, China), and randomly divided into four groups, each containing four mice. Animals were housed in a specific pathogen-free facility under suitable growth conditions. We evaluated the roles of SDCBP and miR-216b in tumor growth via a subcutaneous experiment in nude mice. To generate xenograft tumors in vivo, 1 × 10⁶ cells of each group were subcutaneously injected in the left axillary of mice. Tumor volume (V) was measured at 7, 14, and 21 days according to the following equation: V = π/6 × L × W², where L and W denote the long and short axis of the tumor, respectively. Finally, all mice were euthanized and dissected. Tumors were weighed and images were obtained for further analyses.

All statistical analyses were performed using GraphPad Prism 8.0 (IBM, USA) and the R Software (version 3.6.1). Data are expressed as the mean ± standard deviation (SD). Significant differences were analyzed using Student’s t-test to compare two groups, and one-way ANOVA to compare multiple groups. Survival curves were prepared using the Kaplan–Meier analysis and analyzed using the log-rank test. Correlations between miRNA and SDCBP were evaluated through Pearson correlation analysis. All experiments were repeated three times. p < 0.05 was considered to be statistically significant (*p < 0.05).

Details of 161 DEPs are shown in Supplementary Table S1 . A total of 332 PC samples and 220 normal samples were used to validate the results. Details of each group are presented in Supplementary Table S2 . Briefly, there were 5,624 DEGs in the TCGA_GTEx group, 3,117 DEGs in the GSE62165 group, and 1,492 DEGs in the GSE15471 group based on |log2FC| >1 and adjusted p < 0.05, with 21 common DEGs among the four groups ( Supplementary Figure S1 and Supplementary Table S1 ). We evaluated the expression and prognostic effect of 21 DEPs through Kaplan–Meier analysis, univariate CPHR, and multivariate CPHR. Eight DEPs were found to be upregulated in PC and were associated with poor survival. After univariate and multivariate CPHR, only three DEPs (CLIC1, KRT7, and SDCBP) were validated as survival-related proteins in PC ( Supplementary Figure S2 ). Due to limited research on SDCBP in PC, this was subsequently selected for further analysis.

GSEA was performed to determine the functional roles of SDCBP in PC. The top five GO terms and KEGG pathways are shown in Supplementary Figure S2 . Briefly, the GO terms indicated that SDCBP was most enriched in cell adhesion, extracellular matrix binding, and regulation of epithelial cell migration ( Supplementary Figure S2 ). KEGG pathway analysis revealed that SDCBP was detected in the PI3K-AKT signaling, Jak-STAT signaling, and ECM-receptor interaction pathways ( Supplementary Figure S2G ). These predictions suggested that SDCBP might be involved in the PI3K-AKT signaling pathway and in EMT. Additionally, upstream miRNAs of SDCBP were predicted through the interaction of TargetScan database, miRNet database, and differential expression miRNAs (DEMIs) with thresholds of |log2 FC| > 1 and adjusted p < 0.05 in TCGA database. Overall, we found 279 target miRNAs of SDCBP in TargetScan, 70 target miRNAs in miRNet, and 37 DEMIs ( Supplementary Figures S3A, B ). Three common target miRNAs were identified, including hsa-miR-139, hsa-miR-155, and hsa-miR-216b. Next, we evaluated the expression and prognostic value of target miRNAs. Only miR-216b was found to have low expression and was related to poor survival in PC ( Supplementary Figures S3C, D ). Finally, we validated the inter-relationships between SDCBP and target miRNAs through correlation analysis. Our results indicated that SDCBP was most strongly correlated with miR-216b expression with a relation index (R) equal to −0.34 and p < 0.05 ( Supplementary Figure S3E ). Therefore, we identified miR-216b as the upstream miRNAs of SDCBP in PC, and this was selected for further analysis.

Based on previous predictions, we determined the expression and prognostic roles of SDCBP via immunohistochemistry. Samples were divided into two groups: SDCBP high expression and SDCBP low expression. The clinical data are shown in Supplementary Table S3 . IHC revealed high expression of SDCBP in 67.2% (41/61) of PC tissues, compared with 29.5% (18/61) of normal pancreatic tissues ( Figure 1G ). In normal pancreas tissue, SDCBP was mainly expressed in the acinar cells, with some expression observed in epithelial ductal cells ( Figures 1A, B ). In PC tissues, SDCBP expression was mainly located in the membrane and cytoplasm ( Figures 1C–F ). Additionally, high expression of SDCBP was associated with poor survival with p = 0.031 ( Figure 1H ). WB and RT-qPCR analyses revealed that SDCBP was more highly expressed in PC tissues compared with normal tissues ( Figures 1I, J ).

To explore the role of SDCBP in PC, we evaluated the expression of SDCBP in one pancreatic cell line HPDE, and five PC cell lines using WB and RT-qPCR analyses. As shown in Figure 2A , high levels of SDCBP expression were observed in Capan-2 and SW1990 cells. Next, we knocked down SDCBP expression using lentivirus and evaluated its roles in PC progression and the underlying mechanism of action. Cell proliferation, migration, and invasion were evaluated after SDCBP knockdown. As shown in Figures 2B, C , the proliferation of Capan-2 and SW1990 cells was markedly inhibited after SDCBP knockdown, which indicated that SDCBP could promote PC proliferation. For the colony formation assay, SDCBP knockdown repressed PC proliferation and colony formation in both Capan-2 and SW1990 cells ( Figures 2D, F ); this repression was statistically significant (p < 0.05; Figures 2E, G ). The results of the Transwell invasion assay suggested that SDCBP inhibition significantly attenuated the invasion ability of the cells ( Figures 2H–K ). Moreover, the migratory ability of PC cells was also inhibited following SDCBP knockdown ( Figures 2L, M ). However, it is unknown why SDCBP knockdown inhibited PC proliferation, migration, and invasion. Based on the results of our studies, we investigated the relationship among SDCBP expression, EMT, and the PI3K/AKT pathway. As shown in Figures 3A–D , the expression of a number of proteins was markedly downregulated following SDCBP knockdown, including that of p-PI3K, p-AKT, and N-cadherin. However, no changes in the expression of PI3K and AKT proteins were observed. Notably, the protein expression of E-cadherin was markedly upregulated, while that of N-cad was inhibited, which represent important features of mesenchymal transformation (EMT). Thus, SDCBP might promote PC progression via EMT mediated by the PI3K/AKT pathway.

MicroRNAs are important regulators of protein expression and play essential roles in the progression of PC. To elucidate the relationship between SDCBP and miRNAs, we predicted and validated the upstream miRNA of SDCBP. As shown in Supplementary Figures S3B–E , miR-216b was predicted as a target miRNA of SDCBP. To validate our prediction, the relative expression of miR-216b was determined in PC tissues. The results suggested that miR-216b was expressed at low levels in PC tissues ( Figure 3E ). Correlation analysis revealed that miR-216b was negatively correlated with SDCBP expression ( Figure 3F ), which was consistent with our prediction ( Supplementary Figure S3E ). In addition, low expression of miR-216b was also found in Capan-2 and SW1990 cells ( Figure 3G ). Next, the binding sites of miR-216b and SDCBP were predicted ( Figure 3H ) and a luciferase plasmid was constructed with wild-type and mutant SDCBP. The luciferase activity of wild-type SDCBP was notably reduced following co-transfection with agomiR-216b in Capan-2 and SW1990 cells. However, the luciferase activity of mutant SDCBP was unchanged ( Figure 3I ). Thus, miR-216b could directly regulate SDCBP expression. Notably, the protein and RNA expression of SDCBP was reduced following transfection with agomiR-216b ( Figures 3J and 5A, C ). Overall, these findings revealed that miR-216b might regulate PC progression by directly targeting SDCBP.

To further explore the roles of miR-216b in PC, we up- and downregulated its expression using agomiR-216b and antagomiR-216b, respectively. As shown in Figures 4A, B , the proliferation of Capan-2 and SW1990 cells was markedly inhibited following transfection with agomiR-216b. However, the proliferation of PC cells was notably increased following transfection with antagomiR-216b. Additionally, the inducing effect of antagomiR-216b on PC proliferation could be rescued through SDCBP knockdown ( Figures 4A, B ). In the colony formation assay, the number of colonies formed was suppressed and increased following transfection with agomiR-216b and antagomiR-216b ( Figures 4C–F ). Colony number in PC cells following transfection with antagomiR-216b was also restored by SDCBP knockdown. Moreover, the migration and invasion of PC cells were inhibited and promoted following transfection with agomiR-216b and antagomiR-216b, respectively, and was recovered following SDCBP knockdown ( Figures 4G–L ). These results indicated that miR-216b regulated PC proliferation, migration, and invasion by directly targeting SDCBP.

The previous results revealed that SDCBP promoted EMT of PC cells via the PI3K/AKT pathway. Considering that SDCBP was identified as a direct target of miR-216b, we validated the roles of miR-216b in EMT and the PI3K/AKT pathway. As shown in Figures 5A–D , some proteins were notably down- and upregulated following transfection with agomiR-216b and antagomiR-216b, respectively, including p-PI3K, p-AKT, and N-cadherin. Furthermore, those changes in the promotion of miR-216b could be rescued from SDCBP knockdown. However, there were no changes in the expression of PI3K and AKT proteins. Notably, expression of E-cadherin also changed following transfection with agomiR-216b and antagomiR-216b. Together, our findings indicated that targeting miR-216b with SDCBP induced EMT in PC through the PI3K/AKT pathway.

To confirm the role of miR-216b in vivo, xenograft tumors were established in BALB/c nude mice. Mice were subcutaneously injected with Capan2 cells and then divided into four groups, as follows: NC, antagomiR-216b, agomiR-216b, and antagomiR-216b + sh-SDCBP. We evaluated changes in tumor volume and weight in each group. As shown in Figures 6A, B , tumor volume was markedly up- and downregulated following transfection with antagomiR-216b and agomiR-216b, respectively. Furthermore, antagomiR-216b significantly reduced tumor volume following SDCBP knockdown. The inhibitory effect of SDCBP on tumor volume could be counteracted by transfection with antagomiR-216b. This response was also consistent with regard to tumor weight ( Figure 6C ). These findings revealed that miR-216b directly targeted SDCBP to regulate PC progression and EMT. The interaction between SDCBP and miR-216b is presented schematically in Figure 6D . SDCBP promotes PC progression by inducing EMT, and was negatively regulated by miR-216b at the post-transcriptional level.