40 paired of ICC and para-cancerous tissue specimens were collected at the Zhongnan Hospital of Wuhan University (Hubei Province, China) after obtaining written consent from the patients. HCCC-9810 and RBE cell lines were obtained from the cell bank of the Chinese Academy of Sciences (Shanghai, China), and were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (GIBCO, US) at 37°C under 5% CO₂. Cells in the logarithmic growth phase were selected for the experiments. The study was conducted according to the guidelines of the “Declaration of Helsinki” and approved by the Hospital Committee for the Protection of Human Subjects.

Total RNA was isolated form the cells and tissues using TRIzol reagent (Thermo Fisher Scientific, MA, USA) according to the manufacturer’s instructions. A reverse transcription kit Vazyme, Hubei, China) was used to synthesize cDNA, and RT-PCR was performed using the SYBR Green PCR Kit (Vazyme, Hubei, China).

The tumor tissue samples were fixed in 10% formalin, embedded in paraffin, and cut into 4μm-thick sections. After clearing with xylene and rehydrating with ethanol, the sections were incubated with 0.3% hydrogen peroxide to inactivate the endogenous peroxidases. The tissue sections were then incubated with the primary antibody, washed with PBS, and probed with the HRP-conjugated secondary antibody and streptavidin (Santa Cruz). Diaminobenzidine substrate was used for color development, followed by counterstaining with hematoxylin. The tissue microarray chips were probed with specific antibodies and digitally scanned.

Western blotting was performed as previously described (31). The proteins extracted from the cells were quantified using the BCA protein assay method (Biyuntian, Wuhan, China), and equal amounts of protein per sample were separated via 6–15% SDS-PAGE and transferred to polyvinylidene fluoride membranes (Millipore, USA). After blocking with 5% skimmed milk, the membranes were incubated overnight with the primary antibodies at 4°C, washed thrice with TBST (10 minutes each), and probed with the secondary antibody at room temperature. The membranes were washed thrice with TBST and developed with ECL.

The cells were seeded in a 6-well plate at the density of 5 × 10⁵ cells per well, and cultured in the presence of EGF or 2-DG 48h. The glucose uptake was measured using a glucose uptake cell-based assay kit (Cayman Chemical) according to the manufacturer’s instructions. The cells were analyzed by flow cytometry. The conditioned media were centrifuged at 13000 g for 10 minutes at 4°C, and the lactate concentration in the clarified supernatants was measured using the L-lactate detection kit (Eton Biosciences) according to the manufacturer’s instructions. The absolute lactate level was calculated from the corresponding standard curve and normalized to the number of cells.

The day before the experiment, ICC cells were planted on the Seahorse XF cell culture plate at 1×10⁶ cells/well. Add the test solution prepared by Seahorse XF Base Medium to the plate, then use NaOH to adjust the pH to 7.4. 2 hours later, change the medium and place it in a carbon dioxide-free incubator for one hour. After adding mitochondrial inhibitors oligomycin, FCCP, and antimycin (AA) plus rotenone (AR), the baseline was measured four times in sequence. Then use the XFp extracellular analyzer (Agilent Technologies, USA) to analyze the oxygen consumption rate (OCR). Similarly, after adding Gluoose, Oligomycin, 2-DG, we can calculate the extracellular acidification rate (ECAR).

EZ ChIP™ Chromatin Immunoprecipitation Kit (Millipore, Billerica, MA) was used for ChIP assay. The treated cells were cross-linked in 1% formaldehyde at 37°C for 10 minutes, and sonicated to obtain chromatin fragments of 200-1000 bp. The lysates were incubated overnight with 3μg anti-Snai1 or anti-IgG antibody (Millipore) at 4°C, followed by magnetic beads for 2h. The protein/DNA complex was eluted and cross-linked, and the immunoprecipitated DNA was analyzed by qRT-PCR.

Cells were co-transfected with FBP1-3’UTR-WT or -MUT reporter plasmid and Renilla luciferase vector using Lipofectamine 3000 reagent (Invitrogen). Luciferase activity was measured 48h later using the dual luciferase reporter gene detection system (Promega), and the data was normalized to Renilla luciferase activity.

The animal experiments were approved by the Institutional Animal Ethics Committee of Wuhan University. 10 of four-week-old female BALB/c nude mice were injected subcutaneously with 1×10⁶ HCCC-9810 cells to induce subcutaneous tumors. The mice were sacrificed 4 weeks later for further analysis. The lung metastasis model was established in four-week-old female BALB/c nude mice by intravenous injection of 3×10⁶ HCCC-9810 cells (n=5 for each group), and the lungs were harvested 4 weeks later for IHC (32).

All data are expressed as the mean ± SD of at least three independent experiments. Two groups were compared by Student’s t test (SPSS statistical software package, version 12; SPSS Inc.). Overall survival rates were analyzed by the Kaplan-Meier method and Cox proportional hazards regression model. P <0.05 was considered statistically significant.

Quantitative real-time PCR was employed to determine the expression levels of aPKC-ι and FBP1 in 40 paired ICC and adjacent nontumor tissues ( Figure 1A ). The results were confirmed by WB and IHC ( Figures 1C, D ). We found that aPKC-ι and FBP1 were both localized to the cytoplasm of cancer cells in IHC experiments. In addition, the expression level of aPKC-ι was significantly higher in tumor tissues compared to that in paired para-matched non-tumor specimens, while FBP1 expression was markedly lower. We next investigated whether the expression of aPKC-ι and FBP1 was associated with clinicopathological characteristics and prognosis. The clinical data of all ICC patients, including age, gender, tumor/non-tumor tissues, nodal invasion, tumor staging and tumor differentiation degree, are summarized in Table 1 .

Overexpression of aPKC-ι was related to tumor–node–metastasis (TNM) stage III-IV (χ2 = 10.417, P=0.004), and medium/poor differentiation (χ2 = 10.157, P=0.005) in ICC patients. And, as expected, low expression of FBP1 was related to tumor–node–metastasis (TNM) stage III-IV (χ2 = 10.417, P=0.004), and medium/poor differentiation (χ2 = 6.144, P=0.034). We further investigated whether there was an correlation between aPKC-ι and FBP1 expression in ICC samples. As shown in Figure 1B , aPKC-ι level is significantly negatively associated with FBP1 (R=-0.605, p<0.01). Meanwhile, a Kaplan-Meier analysis indicated that patients with high expression of aPKC-ι displayed a shorter OS compared to those with low expression, while those with low FBP1 expression exhibited a shorter OS ( Figure 1E ). Multivariate Cox regression analyses also showed that aPKC-ι and FBP1 were independent prognostic factors for OS in ICC ( Table 2 ). These above results suggested that aPKC-ι may interact with FBP1 and promote invasion and metastasis of ICC.

Previous studies reported that aPKC-ι may induce EMT-like changes and promote metastasis in cancer cells (33). In this study, we first established two stable human ICC cell lines, HCCC-9810 and RBE, with down-regulated aPKC-ι expression levels in vitro by transfection with human aPKC-ι siRNA ( Figure 2A ). Along with aPKC-ι down-regulation, the EMT-like protein expression profiles in both HCCC-9810 and RBE cell lines were reversed compared with negative controls ( Figure 2B ), including up-regulation of the epithelial markers E-cadherin and down-regulation of mesenchymal marker Vimentin. These indicated that aPKC-ι promoted early invasion and metastasis of ICC cells by accelerating EMT. Consistent with above findings, ICC cells with down-regulated aPKC-ι also reversed EMT-like cellular characteristics, including decreased cell proliferation, migration and invasion as compared to negative controls ( Figures 2C–E ). To further assess the effects of aPKC-ι in ICC, metabolism-related indices were determined by glucose uptake and lactate secretion assay. We found that knocking down aPKC-ι deceased the rate of glucose uptake and lactate production in the ICC cells ( Figures 2F, G ), which was according with its role in the aberrant glucose metabolism of cancer cells. Thus, we speculated that aPKC-ι promoted EMT and aerobic glycolysis of ICC cells, which could provide a competitive environment for ICC cells invasion and migration.

To investigate the causal relationship between aPKC-ι and FBP1, we first suppressed aPKC-ι expression in two ICC cell lines. FBP1 mRNA and protein levels were up-regulated in aPKC-ι-deficient HCCC-9810 and RBE cells ( Figures 2A, B ), which was suggestive of an inhibitory effect of aPKC-ι on FBP1 mRNA and protein. We further determined whether FBP1 was regulated by aPKC-ι. The mRNA and protein levels of FBP1 were down-regulated in aPKC-ι-overexpressing ICC cells by transfecting with aPKC-ι-cDNA ( Figures 3A, B ). Therefore, FBP1 expressions were markedly increased both in mRNA and protein levels by transfecting aPKC-ι-cDNA ICC cells with FBP1-cDNA, while aPKC-ι levels were no significant changes as compared with negative control.

Following the up-regulation of FBP1 expression in aPKC-ι-overexpressing ICC cells, EMT-like changes were almost reversed compared with negative control groups, including the expression levels of EMT markers ( Figure 3B ) and cell proliferation, invasion, and migration ( Figures 3C–E ). In addition, along with the up-regulation of FBP1 expression, FBP1-cDNA transfection significantly decreased the relative glucose uptake rate and lactate production ( Figures 3F, G ). Those suggested that FBP1 could reverse the aPKC-ι induced EMT-like changes and normalize glucose metabolism in ICC cells.

Along with koncking down aPKC-i, four key glycolytic enzymes (HK2, ENO1, PKM, LDHA) were decreased at the protein level ( Figure 4A ), which additionally supported that aPKC-i could enhance the glycolysis level of ICC. The cellular oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were further measured in ICC cells ( Figure 4B ). According to ECAR and OCR curve chart ( Figures 4C, D ), the glycolysis level, glycolysis capacity and glycolysis capacity reserve were significantly increased in aPKC-i-overexpressing ICC cells compared to negative control, while OCR was decreased as compared with negative control. Following increased FBP1 expression in aPKC-i-overexpressing ICC cells, those glycolysis markers were markedly down-regulated, whereas the OCR was obviously up-regulated. The results reinforced that aPKC-i/FBP1 pathway palys a important role in glucose metabolism of ICC, which was consistent with our previous conclusions.

To further evaluate the role of FBP1 in ICC, a xenograft tumor model and a pulmonary metastasis tumor model were established. We subcutaneously injected aPKC-ι-overexpressing HCCC-9810 cells transfected with FBP1-cDNA or vector into nude mice. The volumes of tumors from aPKC-ι-overexpressing cells transfected with FBP1-cDNA were dramatically smaller compared with negative control groups ( Figure 5A ). Furthermore, the in situ expression of FBP1 and E-cadherin detected by IHC were significantly higher in FBP1-cDNA treatment groups, while Vimentin expression was lower ( Figure 5B ). The in situ expression of HK2,ENO1,PKM,LDHA detected by IHC was higher in FBP1-cDNA treatment groups ( Figure 5C ). Moreover, fewer metastatic nodules were determined in the treatment groups than in the untreated groups, suggesting that FBP1 could significantly inhibit lung metastases of ICC cells induced by aPKC-ι ( Figure 5D ). In conclusion, the assay confirmed that FBP1 played an important role in aPKC-ι-mediated ICC progression and metastasis in vivo.

To explore the underlying molecular mechanisms of decreased FBP1 by aPKC-ι, we first performed coimmunoprecipitation (CO-IP) experiments to assess whether aPKC-ι directly interacts with FBP1. Expression of aPKC-ι and FBP1 was not detected in the precipitated protein complex, suggesting that aPKC-ι indirectly regulated Snail through other ways. Snail, a very well-known transcription factor, was reported to be critical for epidermal growth factor–induced EMT of cancer cells (34). We noticed that nine consensus Snail-binding E-boxes domains (CAGGTG) were identified in the FBP1 promoter (17). Moreover, our previous studies had demonstrated that snail was also crucial for aPKC-ι-induced EMT-like changes in cholangiocarcinoma (29). Hence, we hypothesized that Snail may bind to the FBP1 promoter and is responsible for FBP1 repression in aPKC-ι-overexpressing ICC cells.

The expression of Snail was first examined by qRT-PCR in 40 pair-matched tumor specimens and non-tumor tissues. In agreement with existing results, higher Snail expression was in tumor samples ( Figure 6A ). Pearson correlation analysis showed a significant positive correlation between aPKC-ι and Snail expression (R=-0.643, p<0.001), whereas high Snail was related to low FBP1 (R=-0.484, p<0.01) ( Figure 6B ). In addition, ectopic expression of Snail in the HCCC-9810 and RBE cells markedly decreased FBP1 protein levels, suggesting that Snail is required for loss of FBP1 by aPKC-ι-mediated inhibition ( Figure 6E ).

To investigate whether Snail binds the promoter regions of FBP1, we constructed wild-type FBP1 (FBP1-WT) and mutant FBP1 (FBP1-MUT) eukaryotic expression vectors and generated ICC cells that stably expressed these two constructs. In dual luciferase assay, Snail significantly repressed the FBP1 promoter activity, indicated that Snail inhibited FBP1 transcription ( Figure 6C ). Therefore, chromatin immunoprecipitation (ChIP) was performed in the RBE cells by using three sets of FBP1 primers and confirmed a direct physical interaction between FBP1 promoter and Snail ( Figure 6D ). Taken together, these results implied that FBP1 is a direct target of Snail in ICC cells.