Two human cholangiocarcinoma cell lines, KRAS WT SSP-25 and KRAS mutant HuCCT1, were obtained from Riken Bioresource Research Center and were authenticated by a next-generation sequencing (NGS) analysis. Based on the NGS analysis, the results indicated that the intrahepatic cholangiocarcinoma, SSP-25 cell is ETK-1:TP53; Simple; p.Arg175His (c.524G>A) correlated to results shown on Cellosaurus website (https://www.cellosaurus.org/). Cells were cultured in RPMI-1640 medium (cat. no. 31800022; Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (cat. no. SH30396.03; Hyclone; Cytiva) in a humidified incubator with 5% CO₂ at 37°C.

After cell seeding, the cells were placed in a 0.25% hormone-depleted, serum-supplemented medium for 2 days to induce serum starvation. Following this, the cells were supplemented with fresh medium containing 5% hormone-depleted serum, along with EGF (cat. no. E9644; Sigma-Aldrich; Merck KGaA), heteronemin (purity >98%, cat. no. 258814; Sigma-Aldrich; Merck KGaA), or LY294002 (cat. no. S1105; Selleck Chemicals) at varying concentrations and treatment times according to the experimental design. Cells that did not receive LY294002 or heteronemin were treated with dimethyl sulfoxide (cat. no. D2650; Sigma-Aldrich; Merck KGaA), while cells that did not receive EGF were treated with phosphate-buffered saline.

SSP-25 and HuCCT1 cells were seeded at a density of 2×10⁵ cells/well in six-well plates. After seeding, the medium was replaced with 0.25% hormone-depleted serum-supplemented medium for 48 h. The hormone-depleted serum was prepared as previously described (30), the hormones were removed by AG^® 1–8X resin (Bio-Rad Laboratories, Inc.). Cells were supplemented with fresh medium containing 5% hormone-depleted serum with different concentrations of heteronemin (0.6, 1.25, 2.5, 5 and 10 µM) for 24 h. Total RNA was extracted and genomic DNA was removed with an Illustra RNAspin Mini RNA Isolation Kit (Cytiva). Deoxyribonuclease I-treated total RNA (1 µg) was reverse-transcribed into cDNA using a RevertAid H Minus First Strand Complementary (c)DNA Synthesis Kit (Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. cDNAs were used as the template for the qPCR analysis. The thermocycling conditions for qPCR were 95°C for 3 min followed by 40 cycles of 95°C for 5 sec and 60°C for 10 sec. qPCRs were conducted using a QuantiNovaTM SYBR^® Green PCR Kit (Qiagen) on a CFX Connect^™ Real-Time PCR Detection System (Bio-Rad Laboratories, Inc.). Primer sequences were as follows: Homo sapiens cyclin D1 (CCND1) forward, 5′-CAAGGCCTGAACCTGAGGAG-3′ and reverse, 5′-GATCACTCTGGAGAGGAAGCG-3′ (accession no. NM_053056); H. sapiens proliferating cell nuclear antigen (PCNA) forward, 5′-TCTGAGGGCTTCGACACCTA-3′ and reverse, 5′-TCATTGCCGGCGCATTTTAG-3′ (accession no. BC062439.1); H. sapiens cytosol MYC proto-oncogene, bHLH transcription factor (c-Myc) forward, 5′-TTCGGGTAGTGGAAAACCAG-3′ and reverse, 5′-CAGCAGCTCGAATTTCTTCC-3′ (accession no. NM_002467.4); H. sapiens EGFR forward, 5′-AATTTACAGGAAATCCTGCATGGC-3′ and reverse, 5′-GATGCTCTCCACGTTGCACA-3′ (accession no. NM_005228); and H. sapiens 18S ribosomal RNA (18S rRNA), forward, 5′-GTAACCCGTTGAACCCCATT-3′ and reverse, 5′-CCATCCAATCGGTAGTAGCG-3′ (accession no. NR_003286.2). Calculations of relative gene expression (normalized to the 18S rRNA reference gene) were performed according to the 2^−ΔΔCq method (31). The fidelity of the PCR was determined with a melting temperature analysis.

To examine signal transduction pathways involved in EGF-induced PD-L1 expression and effects of heteronemin on EGF-induced PD-L1 expression in SSP-25 and HuCCT1 cells, a western blot analysis was performed to detect protein expression levels of PD-L1 and activation of PI3K, STAT1 and STAT3 in total cell lysates of cells that had been treated with the vehicle for 24 h. Protein concentration was determined by bicinchoninic acid method and protein samples were resolved by 10% sodium dodecyl-sulfate polyacrylamide gel electrophoresis. A 20-µg quantity of protein was loaded into each well with sample buffer, and samples were resolved by electrophoresis at 100 V for 2 h. The resolved proteins were transferred from the polyacrylamide gel to Millipore Immobilon-PSQ Transfer polyvinylidene difluoride membranes (MilliporeSigma) with the Mini Trans-Blot^® Cell (Bio-Rad Laboratories, Inc.). Membranes were blocked with a solution of 3% bovine serum albumin (cat. no. A7030; MilliporeSigma) in Tris-buffered saline (TBS) with 0.1% Tween-20 (TBST) at room temperature for 1 h and incubated with primary antibodies to PD-L1 (1:1,000; cat. no. GTX104763; GeneTex International Corporation), PI3K (1:1,000; cat. no. 610046; BD Biosciences), phosphorylated (p)-PI3K p85 (Tyr458)/p55 (Tyr199) (1:3,000; cat. no. 4228; Cell Signaling Technology, Inc.), STAT1 (1:1,000; cat. no. 66545-1-1g; Proteintech Group, Inc.), p-STAT1 (Tyr701) (1:1,000; cat. no. 9167, Cell Signaling Technology, Inc.), AKT (1:1,000; cat. no. 60203; Proteintech Group, Inc.), p-AKT (Ser473) (1:1,000; cat. no. 9271, Cell Signaling Technology, Inc.), STAT3 (1:1,000; cat. no. 610190; BD Biosciences), p-STAT3 (Tyr705) (1:1,000; cat. no. 9145, Cell Signaling Technology, Inc.) and GAPDH (1:20,000; cat. no. 60004-1, GeneTex International Corporation) at 4°C overnight. The antibody-probed membrane was washed with TBST containing 5% fat-free milk (5% TBST/milk) three times for 10 min and then probed with goat anti-mouse immunoglobulin G (IgG) (cat. no. GTX213111-05) or goat anti-rabbit IgG (cat. no. GTX213110-04; both from GeneTex International Corporation) horseradish peroxidase (HRP)-conjugated secondary antibodies, which were prepared in 5% TBST/milk at a 1:20,000 dilution at room temperature for 1 h. After the membrane was washed three times for 10 min with TBS, chemiluminescent detection was performed using the Immobilon Western Chemiluminescent HRP Substrate (MilliporeSigma). Bands were imaged with the BioSpectrum Imaging System (UVP) and quantified using densitometry by ImageJ 1.47 software (National Institutes of Health) according to the software instructions.

SSP-25 and HuCCT1 cells were seeded in 96-well plates at a density of 3,000 cells/well. After 24 h for cell attachment, cells were starved with 0.25% hormone-depleted serum-supplemented medium for 48 h. Then, serum-starved cells were treated with various concentrations of EGF (10, 20, 40 and 100 ng/ml) and heteronemin (0.3, 0.6, 1.25, 2.5, 5, 10 µM) or 10 µM LY294002 in 5% hormone-depleted serum-supplemented medium for 72 h. Medium and reagents were refreshed daily. Cell viability was assayed with a Cell Counting Kit-8 (cat. no. 96992; Sigma-Aldrich; Merck KGaA), according to the manufacturer's protocol. Briefly, cells were incubated with 100 µl/well CCK-8 working solution at 37°C for 1 h, and the absorbance was measured at 450 nm.

In the present study, the statistical significance of all data was analyzed using a one-way ANOVA followed by Tukey's post hoc test using the SigmaPlot 14.5 software (Systat Software, Inc.). All data are presented as the mean ± standard deviation (SD). P<0.05 was considered to indicate a statistically significant difference.

EGF stimulates cancer growth via binding to its receptor, EGFR. Although it was reported that EGF induced growth of the KMBC and Witt cholangiocarcinoma cell lines (32), it is unclear how cholangiocarcinoma cells with different KRAS statuses respond to EGF stimulation. In the present study, the proliferative effect of activated EGFR was first examined on two cholangiocarcinoma cell lines, KRAS WT SSP-25 and KRAS mutant HuCCT1, by stimulating them with various EGF concentrations.

SSP-25 and HuCCT1 cells were stimulated with different EGF concentrations for 24, 48 and 72 h to examine the effect of EGF on their proliferation. EGF induced cell proliferation in SSP-25 cells starting at 20 ng/ml during the 24 to 72 h treatment period (Figs. S1A and 1A), whereas HuCCT1 cell proliferation was significantly stimulated only at 100 ng/ml during the same period (Figs. S1B and 1B). Although stimulation with EGF for 24 h significantly induced cell proliferation, the noticeable fold changes compared with the control occurred only at 72 h. Given that EGF-stimulated cell proliferation initiates between 24 and 72 h, and knowing that growth factor stimulation leads to rapid changes in mRNA and protein expression-often within 24 h, which is sufficient to capture immediate signaling events-24 h were selected to assess mRNA and protein expression and 72 h to evaluate cell viability.

Signaling pathways activated by EGF in these two cholangiocarcinoma cell lines were next examined. In SSP-25 cells, EGF induced the phosphorylation of PI3K at 40 ng/ml and the phosphorylation of STAT1 in the range of 20–40 ng/ml without altering STAT1 expression (Fig. 2A), reflecting different mechanisms that regulate the expression of STAT1 and the phosphorylation of STAT1. Unexpectedly, the phosphorylation of STAT1 was decreased at 100 ng/ml EGF treatment. This reduction may be due to an increase in negative regulators of the STAT pathway and the induction of phosphatases such as Src homology region 2 domain-containing phosphatase-2 (33), leading to rapid dephosphorylation of STAT1 at Y701. Besides, EGF inhibited the activation of another STAT member, STAT3, by suppressing its phosphorylation without affecting its protein level (Fig. 2B). Unlike its effect in SSP-25 cells, EGF stimulation of HuCCT1 cells increased the phosphorylation of PI3K and STAT3 in concentration-dependent manners (Fig. 2C). EGF did not obviously alter STAT1 expression but significantly suppressed STAT1 phosphorylation (Fig. 2C). Additionally, EGF even at the lowest concentration stimulated PD-L1 expression, a proliferation-related protein, in both cell lines (Fig. 2B and C). Thus, EGF stimulation activates PI3K/STAT1 signaling to induce the proliferation of SSP-25 cells and PI3K/STAT3 signaling to induce the proliferation of HuCCT1 cells. To optimize the effect of heteronemin in both cell lines under EGF stimulation, 100 ng/ml EGF was chosen for the following experiments.

Heteronemin was shown to have antiproliferative effects in various cancer cell lines, including cholangiocarcinoma (23–28). However, signal transduction pathways underlying how heteronemin causes this antiproliferative effect in cholangiocarcinoma cells remain unknown. The proliferation inhibitory effects of heteronemin were examined on two different types of cholangiocarcinoma cells, SSP-25 and HuCCT1, by treating them with various concentrations of heteronemin for 72 h. Heteronemin caused significant, concentration-dependent cytotoxic effects in both cell lines, starting at 0.3 µM (Fig. 3).

According to the functions of the proliferation-related genes PCNA, CCND1 and c-Myc, these three common proliferation genes were selected to assess the proliferation inhibitory effect of heteronemin on cholangiocarcinoma cells (34). PCNA is crucial for DNA replication, CCND1 regulates the cell cycle and is linked to cancer progression, and c-Myc acts as a master regulator of metabolism and proliferation, driving malignant transformation through various oncogenic pathways. To explore the suppression of proliferation effects by heteronemin in SSP-25 and HuCCT1 cells, particularly related to the differential expression of PCNA, CCND1 and c-Myc, SSP-25 and HuCCT1 cells were treated with the indicated concentrations of heteronemin for 24 h. In SSP-25 cells, heteronemin downregulated PCNA and CCND1 gene expression in concentration-dependent manners (Fig. 4A). The expression of these two genes in HuCCT1 cells was suppressed by ~90% under heteronemin treatment in the range of 0.6–5 µM (Fig. 4B). However, in SSP-25 and HuCCT1 cells, different concentrations of heteronemin showed contrasting effects on c-Myc expression. Lower concentrations (0.6 and 1.25 µM) of heteronemin exhibited an increased effect. By contrast, higher concentrations (5 and 10 µM) inhibited expression of this gene (Fig. 4). In addition, the effect of heteronemin was also examined on EGFR expression in both cell lines. Only 5 and 10 µM heteronemin inhibited EGFR expression in SSP-25 cells (Fig. 4A); however, it significantly suppressed EGFR expression in HuCCT1 cells even at the lowest concentration (0.6 µM) (Fig. 4B). The different concentrations of heteronemin show distinct effects on the gene expression of CCND1, c-Myc and EGFR, which might be due to different binding affinities. Low concentrations of heteronemin effectively block PCNA expression through a specific target molecule. However, the target molecules regulating CCND1, c-Myc and EGFR have low affinity for heteronemin, which may explain the opposite effects of low and high concentrations on their mRNA expression. Thus, these results indicated that heteronemin reduces expression of the PCNA, CCND1, c-Myc and EGFR proliferation-related genes in both cholangiocarcinoma cell lines. Additionally, KRAS mutant HuCCT1 cells exhibited greater sensitivity to heteronemin treatment.

Activation of the EGFR signal transduction pathway induces cellular motility (35) and promotes cancer cell proliferation (1,11). Recently, it was revealed by the authors that heteronemin inhibited the EGFR signal transduction pathway to downregulate the proliferation of colorectal cancer cells (25). To further investigate whether heteronemin suppresses signal transduction pathways activated by EGF in cholangiocarcinoma cells, SSP-25 and HuCCT1 cells were stimulated with 100 ng/ml EGF in the presence and absence of indicated concentrations of heteronemin for 24 h. In SSP-25 cells, PI3K and its phosphorylated forms increased in heteronemin-treated cells in concentration-dependent manners up to 1.25 µM (Fig. 5). After treatment of this cell line with 1.25 and 2.5 µM heteronemin, STAT1 phosphorylation significantly increased compared with untreated cells. In contrast to PI3K phosphorylation, the phosphorylation pattern of STAT3 decreased in heteronemin-treated cells in a concentration-dependent manner. Heteronemin inhibited the EGF-induced phosphorylation of PI3K in SSP-25 cells (Fig. 5). The phosphorylation of both STAT1 and STAT3 increased in concentration-dependent manners under heteronemin treatment in EGF-stimulated cells, with the exception of the effect of 2.5 µM heteronemin on STAT3.

A parallel experiment was conducted in HuCCT-1 cells, and it was found that heteronemin suppressed EGF-induced signal transduction pathways. Heteronemin at up to 1.25 µM induced the phosphorylation of PI3K, STAT1 and STAT3 in HuCCT-1 cells (Fig. 6). On the other hand, 100 ng/ml EGF increased concentrations of PI3K, STAT1 and STAT3, and their phosphorylated forms, compared with unstimulated cells. Heteronemin inhibited the EGF-induced phosphorylation of PI3K, STAT1 and STAT3 in concentration-dependent manners (Fig. 6). In addition, PI3K, STAT1 and STAT3 levels decreased under treatment with 2.5 µM heteronemin in HuCCT1 cells. To confirm that PI3K plays an important role in regulating signaling pathways induced by the EGF in SSP-25 and HuCCT1 cells, 10 µM LY294002, an inhibitor of PI3K, was applied to EGF-stimulated cells. As shown in Fig. S2, inhibition of PI3K activity decreased levels of PD-L1 and p-STAT1 and p-STAT3 in both EGF-stimulated cell lines. This pattern was similar to the effect of heteronemin on EGF-stimulated SSP-25 and HuCCT1 cells. These results indicated that heteronemin inhibits the proliferation of SSP-25 and HuCCT1 cells by suppressing the PI3K-mediated signaling pathways induced by EGF.

As demonstrated in Figs. 1 and 2, PI3K may play a crucial role in promoting EGF-triggered proliferation in cholangiocarcinoma cells. Moreover, EGF led to a concentration-dependent elevation in PD-L1 expression, as depicted in Fig. 1. Notably, heteronemin was observed to suppress EGF-induced PI3K activation (Figs. 5 and 6). To determine whether heteronemin inhibits the proliferation of cholangiocarcinoma cells via the PI3K signaling transduction pathway activated by the EGF, PD-L1 expression in SSP-25 and HuCCT1 cells was examined by stimulating them with 100 ng/ml of EGF in the presence of different concentrations of heteronemin for 24 h. Heteronemin treatment increased PD-L1 expression in SSP-25 cells and decreased PD-L1 expression in HuCCT1 cells in concentration-dependent manners (Fig. 7). EGF induced PD-L1 expression in both cholangiocarcinoma cell lines, and this effect was concentration-dependently inhibited by heteronemin (Fig. 7).

The role of PI3K in the heteronemin-induced inhibitory effect was next investigated on EGF-induced proliferation in cholangiocarcinoma cells. The two cholangiocarcinoma cell lines were treated with EGF and heteronemin, either alone or in the presence of LY294002 for 72 h. In both cell lines, LY294002 not only suppressed cell proliferation on its own but also inhibited EGF-induced cancer cell proliferation (Fig. 8). Heteronemin similarly inhibited cancer cell proliferation in both control and EGF-stimulated cells. Although the combined treatment of heteronemin and LY294002 did not show an additive effect on cell viability, their inhibitory effect was the same as that of treatment with heteronemin, LY294002 or combined treatment in SSP-25 cells but not in HuCCT1 cells. It is possible that the KRAS mutation in HuCCT1 cells constitutively activates other signaling pathways, such as ERK1/2, to offset the inhibitory effect of heteronemin. This situation is similar to previous studies, where a combination treatment (heteronemin plus another inhibitor) does not show an additive effect on cell viability but does suppress target protein activation (25,36). Taken together with the inhibitory effect of heteronemin on PI3K phosphorylation (Figs. 5,6), these results revealed that the inhibitory effect of heteronemin on EGF-induced cell proliferation depends on suppression of PI3K activity.