The anti-DESI-2 primary antibody (#20517-1-AP) was purchased from Proteintech Group, Inc. (Rosemont, IL, USA). Primary antibodies targeting BH3 interacting-domain death agonist (BID; #sc-373939), retinoid X receptor (RXR; #sc-553), cytochrome c (#sc-13156), caspase-3 (#sc-271028), Rho-associated protein kinase (ROCK; #sc-17794) and β-actin (#sc-81178) were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). The anti-Ras homolog gene family member A (RhoA) primary antibody (#ab54835) was purchased from Abcam (Cambridge, MA, USA). Horseradish peroxidase (HRP)-conjugated secondary antibodies, including goat anti-rabbit immunoglobulin (Ig)G (#ZB-2301) and goat anti-mouse IgG (#ZB-2305), were purchased from ZSGB-Bio (Beijing, China). The radioimmunoprecipitation assay (RIPA) lysis buffer was from the Beyotime Institute of Biotechnology (Haimen, China). A Histostain-Plus kit and an Immobilon Western Chemiluminescent HRP Substrate were purchased from Invitrogen (Thermo Fisher Scientific, Inc., Waltham, MA, USA) and EMD Millipore (Billerica, MA, USA), respectively.

PANC-1 human epithelial pancreatic cancer cells (ATCC, Manassas, VA, USA) were cultured in Gibco Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (both from Thermo Fisher Scientific, Inc.) at 37°C in a humidified incubator with an atmosphere containing 5% CO₂. The open reading frames of DESI-2 (GenBank accession: NM_016076) were subcloned into a plasmid pcDNA3.1(+) vector (Invitrogen; Thermo Fisher Scientific, Inc.) between the BamHI and XhoI sites according to the manufacturer's instructions, in order to obtain the recombinant plasmid vector pcDNA3.1-DESI-2. The pcDNA3.1-DESI-2 and empty control vector were transfected into PANC-1 cells using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Following a 24-h transfection, the cells were collected and prepared for subsequent experiments.

The Human 12×135K Gene Expression Array was manufactured by Roche NimbleGen, Inc. (Madison, WI, USA). Briefly, total RNA was isolated from PANC-1 cells using Trizol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and quantified using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies; Thermo Fisher Scientific, Inc.). The integrity of the RNA was assessed by electrophoresis on a 1% denaturing agarose gel. Double-stranded cDNA (ds-cDNA) was then synthesized from 1-µg RNA samples using a SuperScript ds-cDNA Synthesis kit (Invitrogen; Thermo Fisher Scientific, Inc.) and labeled with Cy3 fluorescence (Roche NimbleGen, Inc.). The microarrays were hybridized with Cy3-labeled ds-cDNA (Hybridization System; Roche NimbleGen, Inc.) and imaged using an Axon GenePix 4000B microarray scanner (Molecular Devices LLC, Sunnyvale, CA, USA). The images were imported into NimbleScan software version 2.5 (Roche NimbleGen, Inc.) for grid alignment and expression data analysis. Expression data were normalized through quantile normalization and the Robust Multichip Average algorithm included in the NimbleScan software. The differentially expressed genes (fold change ≥2.00) were identified through fold change filtering by Agilent GeneSpring GX software version 11.5.1 (Agilent Technologies, Inc., Santa Clara, CA, USA). Hierarchical clustering analysis was also performed using the Agilent GeneSpring GX software version 11.5.1. Gene Ontology (GO; www.geneontology.org) and Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.genome.jp/kegg/) analyses were performed to determine the respective roles of the differentially expressed genes.

To extract the proteins, PANC-1 cells were collected and lysed using RIPA lysis buffer on ice prior to being centrifuged at 12,000 × g for 30 min at 4°C. Following the determination of protein concentration using a Pierce bicinchoninic acid assay kit (Thermo Fisher Scientific, Inc.), the protein samples (50 µg/lane, reduced with 50 mM dithiothreitol) were separated by 12% SDS-PAGE under denaturing conditions. Subsequently, the separated proteins were transferred onto polyvinylidene fluoride membranes and blocked in phosphate-buffered saline with 0.05% Tween^® 20 (PBST) containing 5% skimmed milk for 2 h. The membranes were washed three times with PBST and incubated with the primary antibodies against DESI-2, RhoA, ROCK, RXR, BID, cytochrome c, caspase-3 and β-actin (dilution, 1:800) for 2 h at room temperature. The membranes were then washed three times with PBST and probed with the HRP-conjugated secondary antibodies (dilution, 1:10,000) for 1 h at room temperature. Following three washes with PBST, the target blots were visualized using Immobilon Western Chemiluminescent HRP Substrate. In order to detect cytochrome c in the cytoplasm, a Mitochondria Isolation kit (Beyotime Institute of Biotechnology) was used to remove the mitochondria, according to the manufacturer's instructions, and these were subsequently subjected to western blotting. The western blot experiments were each performed in triplicate to assess the relative protein abundance.

Pancreatic ductal adenocarcinoma tissue specimens were obtained from 96 patients with pancreatic cancer (age range, 31–74 years; 54 males and 42 females) who had undergone surgery at the West China Hospital of Sichuan University (Chengdu, China). The institutional ethics committee of Sichuan University approved the present study and informed consent was obtained from all patients enrolled in the study. The formalin-fixed and paraffin-embedded tissue samples were cut into 5-µm-thick sections using a microtome. The sections were dewaxed in xylene and rehydrated in graded alcohol. Following heat-induced epitope retrieval, the sections were probed for DESI-2, RXR and RhoA with the corresponding primary antibodies (dilution, 1:200) at 4°C overnight. After washing for 3×5 min in PBS, the sections were stained using an avidin-biotin-peroxidase complex method and visualized with diaminobenzidine (Beyotime Institute of Biotechnology) according to the manufacturer's instructions. The sections were imaged with a Leica DM2500 microscope and analyzed using Leica Application Suite software version 3.8 (Leica Microsystems, Inc., Buffalo Grove, IL, USA). DESI-2 expression level was scored as a staining intensity score (0, negative; 1, mild; 2, moderate; and 3, strong) multiplied by a score representing the percentage of positively staining cells (0, none; 1, 1–25%; 2, 26–50%; 3, 51–75%; and 4, 76–100%), resulting in an overall scoring scale of 0–12 (12). According to the median expression level of DESI-2, the cancer samples were divided into DESI-2-higher and DESI-2-lower tissue samples.

GO analysis was used to analyze differentially expressed genes with respect to GO categories. The P-value denotes the significance of the GO term enrichment in the differentially expressed gene list (threshold for significance, P<0.05). Biological pathways that were significantly enriched among the differentially expressed genes were determined by pathway analysis based on the KEGG database (P<0.05).

The recombinant vector pc-DNA3.1(+) containing DESI-2 was transiently overexpressed in PANC-1 cells, with an empty vector as the control. The total RNA was isolated from the cells and subjected to quality control prior to mRNA microarray. Agarose gel electrophoresis demonstrated that the isolated RNA samples were able to be separated into three distinct bands of 28s RNA, 18s RNA and 5s RNA (Fig. 1A). The gene expression variations were assessed and visualized using the scatter-plot method to identify differential genes (Fig. 1B). Hierarchical cluster analysis was performed based on their expression levels and presented in the dendrogram. The results revealed distinguishable gene expression profiles between DESI-2 overexpressing cells and the control cells (Fig. 1C).

A total of 45,033 genes were examined by NimbleGen Human Gene Expression Microarrays. The genes were considered to be differentially expressed when the mRNA abundance exceeded the fold change threshold (fold change ≥2.0). The altered expression included a total of 1,766 upregulated and 1,643 downregulated genes. These differential genes were subsequently subjected to pathway analysis using the KEGG database. The upregulated genes were primarily distributed in protein degradation, cell death and differentiation (Fig. 2A; only the top 10 altered pathways are presented). It is noteworthy that a number of genes associated with apoptosis signaling, including Fas, caspase-10, BID and TRADD, exhibited increased expression levels (Fig. 2B). The downregulated genes were principally distributed in metabolism, cell movement and adhesion (Fig. 3A; only the top 10 altered pathways are presented). The critical genes Rho-guanine nucleotide exchange factor (GEF), RhoA and ROCK in the Rho signaling pathway exhibited decreased expression levels (Fig. 3B).

As changes to the mRNA must be evaluated at the protein level, the key signaling pathways that were altered in DESI-2 overexpressing cells were further examined. Western blotting demonstrated that RXR and BID are upregulated in PANC-1 cells that overexpress DESI-2 (Fig. 4). BID has an essential role in the apoptosis pathway and, therefore, its downstream signaling pathways, including the release of cytochrome c and the activation of caspase-3, were further investigated. As hypothesized, cleaved caspase-3 and the emergence of cytochrome c in cytoplasm were detected in the cells overexpressing DESI-2, but not in the control PANC-1 cells (Fig. 4). These results suggest that endogenous apoptosis may be induced by the overexpression of DESI-2 in PANC-1 cells.

With regard to the downregulated mRNA, western blotting demonstrated that RhoA and ROCK exhibited decreased expression levels in DESI-2 overexpressing PANC-1 cells, corresponding with the changes in their mRNA (Fig. 4). These results indicated that the RhoA signaling pathway was downregulated following the overexpression of DESI-2 in PANC-1 cells.

RhoA and RXR exhibit marked changes in their expression profiles in DESI-2 overexpressing cells; thus, the expression features of RhoA and RXR in pancreatic cancer tissues were further investigated. A total of 96 pancreatic ductal adenocarcinoma tissue specimens were collected and subjected to immunohistochemistry to analyze the expression levels of DESI-2, RhoA and RXR. The tissue samples were divided into two groups according to the expression levels of DESI-2, which included DESI-2 abundant and DESI-2 scarce pancreatic tumor tissues (Fig. 5). Immunohistochemistry demonstrated that RXR exhibited similar changes in expression levels to DESI-2, with higher expression levels in DESI-2 abundant pancreatic tumor tissues (Fig. 5). However, RhoA expression exhibited lower expression levels in DESI-2-abundant pancreatic tumor tissues, contrary to the expression profile of DESI-2 (Fig. 5).

These results were established by further analysis. The 96 tissue samples were divided into the DESI-2-higher (52 tissue samples) and DESI-2-lower (44 tissue samples) cancer groups, according to the median expression level of DESI-2 present. The tissue samples with RXR or RhoA expression levels higher or lower than the median expression level are presented in Fig. 6. The proportion of tissues with RXR expression levels exceeding the median is markedly increased in DESI-2-higher cancer (31/52), compared with DESI-2-lower cancer (11/44). However, the proportion of tissues with RhoA expression levels that exceed the median is markedly decreased in DESI-2-higher cancer (15/52), compared with DESI-2-lower cancer (29/44).