Cytotoxic and senescence-inducing effects of BLU554 in pancreatic ductal adenocarcinoma: an in vitro study.

Background
Pancreatic cancer is a leading cause of cancer-related death worldwide. Recent data from the World Cancer Research Fund indicate that the global incidence of pancreatic cancer increased from 196,000 in 1990 to 510,000 in 2022 [1, 2]. China, the United States, and Japan reported the highest number of pancreatic cancer cases in 2022, and the incidence of the disease continues to rise annually [3]. In Japan, the 5-year survival rate for pancreatic cancer was approximately 8.9% in men and 8.1% in women between 2009 and 2011 (https://ganjoho.jp/reg_stat/statistics/stat/cancer/10_pancreas.html), and this metric has shown little improvement over the years. Although liposomal irinotecan with oxaliplatin, 5-fluorouracil, and leucovorin (NALIRIFOX); folinic acid, fluorouracil, irinotecan, and oxaliplatin (FOLFIRINOX); and gemcitabine plus nab-paclitaxel are used as first-line chemotherapies for metastatic pancreatic cancer, the associated median progression-free survival durations are 7.4, 7.3, and 5.7 months, respectively [4]. Therefore, new therapeutic strategies are urgently required owing to the limited efficacy of current treatments.
The involvement of fibroblast growth factor receptors (FGFRs) in carcinogenesis has been reported in various organs [5, 6]. Among members of the FGFR family, FGFR4 is less well characterized, although emerging evidence has started clarifying its role in the development of cancer. Silencing FGFR4 expression in gastric cancer cells significantly reduces cell growth and increases apoptosis [7]. The FGFR4-specific inhibitor BLU9931 suppresses the growth of gastric tumor xenografts in nude rats [8–10]. In pancreatic cancer, FGFR4 overexpression has been reported in high-grade pancreatic intraepithelial neoplasia and pancreatic ductal adenocarcinoma (PDAC) [11]. Therefore, FGFR4 was investigated as a molecular target for pancreatic cancer therapy.
In this study, we selected BLU554, a highly specific inhibitor of FGFR4 [12]. Our previous study demonstrated that BLU9931 suppressed proliferation and invasion of pancreatic cancer cells in pancreatic cancer cell lines PK-1 and T3M-4 [13]. Compared with BLU9931, BLU554 exhibits stronger FGFR4 selectivity [14], greater anticancer activity in gastric cancer cell line MKN45 [12], oral bioavailability, and clinical advancement to phase Ib/II clinical trials [15], suggesting superior translational potential. Because BLU9931 was found to induce cellular senescence, we also assessed whether BLU554 exerts similar effects. Given that senescent cells play dual roles in tumor progression [16], their selective elimination following treatment may further enhance therapeutic efficacy.
Senescent cells can be selectively eliminated by senolytic agents such as quercetin and dasatinib [17]. Quercetin, a natural polyphenol, and dasatinib, a tyrosine kinase inhibitor, have been shown to efficiently induce apoptosis in senescent cells without affecting quiescent, proliferating, or differentiated fibroblasts [18]. Therefore, these senolytic agents were employed to eliminate drug-induced senescent cancer cells. In this study, we investigated whether BLU554 induces cellular senescence in PDAC cells and whether subsequent senolysis enhances therapeutic efficacy. This two-step strategy may represent a novel therapeutic approach for pancreatic cancer by combining senescence induction with targeted senolysis.

Methods

Cell culture
The human PDAC cell line PK-1 was obtained from the Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer, Tohoku University (Sendai, Japan). The human PDAC cell line T3M-4 was provided by the RIKEN BioResource Research Center (Tsukuba, Japan). Cells were cultured in RPMI1640 supplemented with 10% fetal bovine serum. Both cell lines were maintained in a 5% CO2 humidified incubator at 37 °C. Genomic DNA was extracted from the PDAC cells using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany), per the manufacturer’s instructions. Short tandem repeats were analyzed using the GenePrint 10 System (Promega, Madison, WI, USA) at BEX Co., Ltd (Tokyo, Japan), according to the manufacturer’s instructions. The PDAC cell lines were confirmed to be correctly genotyped and contamination-free (Figure S1). Mycoplasma contamination was tested using Mycostrip (InvivoGen, Hong Kong, China), and no positive signal was detected in any cultured cells.

WST-8 assay
The viability of BLU554-treated cells was quantified using the Cell Count Reagent SF (Nacalai Tesque, Kyoto, Japan). PDAC cells were treated with 5–50 µM BLU554 for 4 days. After treatment, 10 µL of WST-8 reagent was added to each well and incubated for 3 h. Thereafter, absorbance was measured at 450 nm using a microplate reader. The experiments were performed in two biological replicates, each with three technical replicates (n = 6).

Western blot
PK-1 and T3M-4 cells were lysed in radioimmunoprecipitation assay buffer supplemented with the complete mini protease inhibitor cocktail (Mannheim, Roche, Germany). Equal amounts of proteins (25 or 50 μg) were loaded on 12% TGX gel (Hercules, CA, BioRad), separated by using sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane. The membranes were blocked with Intercept Tris-Buffered Saline (TBS) Blocking Buffer (LICORbio, Lincoln, NE) and incubated with anti-p21 Waf1/Cip1 (12D1) rabbit monoclonal antibody (1:1000, #2947, Cell Signaling) or anti-Lamin B1 (D9V6H) rabbit monoclonal antibody (1:1000, #13435, Cell Signaling) overnight at 4 °C. Goat anti ß-actin antibody (ab8229, Abcam) was used as the loading control antibody. Membranes were washed with TBS (0.05% Tween-20) and incubated with respective secondary antibodies (1:8,000) for 1 h at room temperature. Bands were visualized using the ChemiDoc MP imaging system (BioRad).

Immunocytochemical staining
PK-1 and T3M-4 cells were seeded onto coverslips (13 mm diameter; Matsunami Glass Ind., Ltd., Osaka, Japan) placed in 24-well culture plates. After treatment with BLU554 (Selleck Chemicals, Houston, TX, USA) for 4 days, the cells were fixed in 10% neutral buffered formalin for 30 min and permeabilized with phosphate-buffered saline (PBS) containing 0.2% Triton X-100 for 10 min. After two washes with PBS, the samples were blocked with Protein Block (Agilent, Santa Clara, CA, USA) for 1 h at 24 °C. For γH2AX detection, the cells were incubated with rabbit anti-γH2AX antibody (NOVUS biologicals, Centennial, CO, USA) at 1:2,000 dilution and viewed under the Simple Stain MAX PO (MULTI) detection system (Nichirei Biosciences Inc., Tokyo, Japan). Signal detection was performed using DAB substrate (Nichirei Biosciences Inc). Images were acquired using a Mantra multi-spectral microscope (Perkin-Elmer, Shelton, CT, USA). The images were then analyzed using inForm software ver. 2.4 (Perkin-Elmer) to count γH2AX-positive cells in 20 randomly selected fields at 10 × magnification.

Quantitative PCR (qPCR)
Total RNA was extracted from PK-1 or T3M-4 cells treated with BLU554 for 4 days. To measure mRNA expression of senescence markers, qPCR was conducted after reverse transcription using SuperScript IV VILO Master Mix (ThermoFisher, Waltham, MA, USA), PowerTrack SYBR Green Master Mix (ThermoFisher), and primers listed in Table 1. The conditions for qPCR were as follows: 95 °C for 2 min, 40 cycles of 95 °C for 5 s and 60 °C for 30 s, 95 °C for 15 s, 60 °C for 1 min, and 95 °C for 15 s. Each experiment was independently performed twice, and all qPCR reactions were run in triplicate using the ABI QuantStudio 3 Real-Time PCR System (ThermoFisher). Relative quantification of each mRNA was performed using the comparative Ct method.

Transmission electron microscopic (TEM) analysis
To observe PK-1 and T3M-4 cells under transmission electron microscopy (TEM), PDAC cells were cultured on coverslips in 24-well plates. The cells were fixed with 2.5% glutaraldehyde for 1 h at 4 °C, followed by post-fixation for 30 min with 1% OsO4. After dehydration using graded ethanol, Epon-filled capsules were placed on the coverslips. Three days later, the coverslips were heated at 100 °C on a hot plate, after which the cells on the coverslips adhered to the Epon. Subsequent procedures and TEM (H-7500; Hitachi High-Technologies, Tokyo, Japan) observations were performed according to a previously described protocol [19, 20]. Experiments were conducted in two independent replicates, and 30 fields were observed via TEM. Lysosome quantification was performed using TEM images acquired at 4,000 × to 7000 × magnification. At least six images per condition were analyzed in each experiment, and the procedure was repeated three times. The total number of lysosomes was normalized to the observed area to account for differences in image size.

Senescence associated-β-galactosidase (SA-β-Gal) assay
Senescent cells were stained using the Cellular Senescence Detection Kit-SPiDER-βGal (Dojindo, Kumamoto, Japan), according to the manufacturer’s protocol. PK-1 (0.5 × 104 cells/well) and T3M-4 (1.5 × 104 cells/well) seeded on 24-well plates were treated with BLU554 for 4 days. Cells fixed in 10% neutral buffered formalin were scanned using a Mantra microscope (ParkinElmer, Waltham, MA, USA), and the images were analyzed using inForm software (ParkinElmer) for automated cell segmentation and phenotyping. For each sample, at least 10 fields were randomly imaged.

Cell viability assay
The number of viable cells in culture was quantified using the CellTiter-Glo 2.0 Assay (Promega, Madison, WI, USA). PDAC cells were treated with 50 µM BLU554 for 4 days and then seeded onto 96-well plates (3 × 103 cells/well). The cells were subsequently treated with dasatinib (Selleck) or quercetin (Cayman Chemical, Ann Arbor, MI, USA) for an additional 4 days. ATP levels, representing metabolically active cells, were quantified. Experiments were performed in two biological replicates, each with five technical replicates (n = 5).

Statistical analysis
Statistical analysis was performed using GraphPad Prism 10.4.2 (GraphPad Software, Boston, MA, USA). All data are presented as the mean ± standard deviation (SD). Statistical significance was assessed using two-tailed t tests, one-way analysis of variance (ANOVA), or two-way ANOVA, as appropriate. Statistical significance was set at P < 0.05.

Results

BLU554 suppresses growth of PDAC cell lines with high FGFR4 expression
In our previous analysis of PDAC tissues (Figure S2) and eight PDAC cell lines, PK-1 and T3M-4 were identified as the highest FGFR4-expressing lines, with concurrent high expression of the ligand FGF19 [13]. To evaluate the tumor-suppressive effects of BLU554, PK-1 and T3M-4 cells were treated with different concentrations of BLU554 (5, 15, 30, and 50 μM), and cell numbers were measured on days 0, 1, 2, 3, and 4. BLU554 inhibited the proliferation of both PK-1 and T3M-4 PDAC cells (Fig. 1A and B). On day 4, treatment with 50 μM BLU554 reduced the number of PK-1 cells to approximately 50% of the control, while treatment with 30 μM BLU554 similarly decreased the number of T3M-4 cells to approximately half of the control. Treatment with 50 μM BLU554 resulted in only minimal proliferation of T3M-4 cells. The WST-8 assay also showed a decrease in cell viability, consistent with the reduction in cell number on day 4 (Fig. 1C). The IC50 values were 22.13 μM for PK-1 and 10.58 μM for T3M-4, showing a roughly similar trend (Figure S3).

Phosphorylation of H2AX is induced in BLU554-treated PDAC cells
γH2AX, the phosphorylated form of the H2AX histone, is recognized as an early senescence marker [21]. To determine whether BLU554 induced cellular senescence in PDAC cells, PK-1 and T3M-4 cells were stained with an anti-γH2AX antibody. Approximately 6.7% of PK-1 control cells were positive for γH2AX. Treatment with BLU554 increased the proportion of γH2AX-positive cells to 13% and 17% at concentrations of 5 and 50 μM, respectively (Fig. 2A and B). Similarly, BLU554 treatment significantly increased the proportion of γH2AX-positive T3M-4 cells: 20.9% at 5 μM and 27.5% at 50 μM relative 1.8% in the control (Fig. 2A and B).

BLU554 induced key markers associated with cellular senescence in PK-1 and T3M-4 cells
Next, we examined the expression of other well-established markers of cellular senescence, CDKN1A and LMNB1, in PK-1 and T3M-4 cells treated with BLU554. Quantitative PCR analysis revealed that the expression of CDKN1A mRNA, whose gene product p21 controls cell-cycle arrest in senescent cells, increased in both PK-1 and T3M-4 cells following BLU554 treatment. The relative expression levels were 1.64 and 3.51, respectively, indicating the induction of cellular senescence (Fig. 3A). In contrast, the expression of LMNB1 mRNA, which encodes Lamin B1 and whose loss is associated with nuclear morphology alterations during cellular senescence, decreased in both cell lines (Fig. 3A). The relative expression levels in PK-1 and T3M-4 cells were 0.41 and 0.07, respectively, compared with the control (set to 1.0).
To further confirm the induction of cellular senescence through the upregulation of CDKN1A (p21) and downregulation of LMNB1 (Lamin B1) in BLU554-treated PDAC cells, Western blot analysis was performed. In PK-1 and T3M-4 cells, p21 expression was increased, while Lamin B1 expression was decreased following treatment with 50 μM BLU554 (Fig. 3B). These results were consistent with the mRNA expression data.

Morphological cellular senescence induced by BLU554
To further assess BLU554-induced cellular senescence in PDAC cells, the morphological features of cellular senescence were investigated using TEM analysis. Treatment of PK-1 and T3M-4 cells with BLU554 resulted in an increase in the number of lysosomes with abnormal structures and irregular contents (Fig. 4A). Compared with control cells and cells treated with 5 μM BLU554, most lysosomes remained intact or exhibited only moderate alterations (Fig. 4A, arrows). These abnormalities were more pronounced at 50 μM, and lysosomes with structures similar to myelosomes were also observed (arrowheads). Treatment with 50 μM BLU554 significantly increased the number of lysosomes in both PK-1 and T3M-4 cells (Fig. 4B).

Detection of senescence-associated beta-galactosidase (SA-β-Gal) activity
Senescence-associated beta-galactosidase (SA-β-Gal), another marker of cellular senescence localized in lysosomes, was detected in the nuclei of PDAC cells using SPiDER-β-Gal (Fig. 5A, red). SA-β-Gal activity was barely detectable in the control cells, whereas a clear signal was observed in PK-1 and T3M-4 cells treated with 50 μM BLU554 for 4 days. The cell nuclei were counterstained with Hoechst 33341 (Blue). A significant increase in the number of cells showing high SA-β-Gal activity was observed in PK-1 cells (p < 0.01), whereas T3M-4 cells displayed a higher mean, but the difference was not statistically significant (Fig. 5B).

BLU554 induces senescence-associated secretory phenotype factors in PDAC cells
To investigate the induction of cellular senescence by BLU554 in PDAC cell lines, the gene expression levels of senescence-associated secretory phenotype (SASP) factors, including granulocyte–macrophage colony-stimulating factor (GM-CSF), IL-1α, IL-1β, IL-6, and TNF-α were analyzed (Fig. 6). GM-CSF mRNA expression was markedly upregulated in both PK-1 and T3M-4 cells. The change in IL-1α mRNA expression was relatively small, with significant differences observed only in PK-1 cells treated with 5 or 50 μM BLU554. IL-1β mRNA levels showed a significant 5.7-fold increase in PK-1 cells treated with 50 μM BLU554. In contrast, IL-6 mRNA expression was downregulated by BLU554. In PK-1 cells, suppression of IL-6 expression was observed only at 50 μM, whereas in T3M-4 cells, both 5 and 50 μM BLU554 significantly reduced IL-6 mRNA levels. No significant changes in TNF-α mRNA expression were observed in either PDAC cell line.

BLU554-induced senescence in PDAC cells sensitizes cells to senolytic drugs
To further enhance the anti-cancer effect of BLU554, senolytic drugs were administered to the surviving cancer cells. Treatment with quercetin or dasatinib alone did not reduce the viability of PK-1 or T3M-4 cells, except at 12 and 25 nM dasatinib in PK-1 cells (Fig. 7A). Therefore, 6 or 12 μM quercetin and 6 or 12 nM dasatinib (concentrations with no or relatively weak effects on PDAC cell viability) were applied to BLU554-induced senescent PDAC cells. Quercetin treatment reduced the viability of T3M-4 cells pretreated with BLU554 at 6 μM (Fig. 7 B). A reduction was also observed at 12 μM quercetin administration, but the difference lacked statistical significance. T3M-4 cells pre-treated with BLU554 were also sensitized to dasatinib at 6 nM and 12 nM (Fig. 7C), which reduced cell viability from 82 to 72%. In contrast, the suppressive effects of senolytic drugs (quercetin and dasatinib) were not observed in PK-1 cells. In addition, when BLU554 was combined with either quercetin or dasatinib, these simultaneous treatments had only minimal effects on cell viability, suggesting that senolytic agents are effective only after the induction of senescence (Figure S4).

Discussion
The FGFR4 inhibitor BLU554 was applied to PK-1 and T3M-4 PDAC cells, which express high levels of FGFR4 [13], to suppress tumor cell growth. Although BLU554 inhibited growth of these cell lines, its anti-cancer activity was limited, as cell numbers for PK-1 or T3M-4 cells did not decrease (Fig. 1A). One possible explanation for this moderate response is that BLU554 treatment induced cellular senescence. Senescent cells are known to be resistant to cell death. Notably, p21 reportedly maintains the viability of senescent normal human IMR-90 fibroblasts [22]. In the present study, BLU554 treatment increased p21 expression (Fig. 3A and B), indicating the induction of senescent cells resistant to cell death. This may partly explain the limited growth inhibition observed in PK-1 and T3M-4 cells. As these persistent senescent cells could reduce the overall efficacy of BLU554, their elimination represents a promising strategy to overcome treatment resistance in PDAC.
Although various markers of cellular senescence have been identified, their expression can vary depending on the cell type and senescence-inducing stimulus [21]. Moreover, senescence in cancer cells has been less frequently studied [23]. Therefore, several markers of cellular senescence were analyzed in this study. γH2AX [21], p21 (CDKN1A) [24], Lamin B1 (LMNB1) [25], lysosome number [26], and SA-β-Gal activity [27, 28] are widely used indicators of senescence. Overall, the significant alterations detected in the majority of senescence markers examined support the notion that BLU554 induces cellular senescence in both PK-1 and T3M-4 pancreatic cancer cells.
The number of lysosomes increased in BLU554-treated PK-1 and T3M-4 cells (Fig. 4B), consistent with observations in other cell types [29]. Morphologically, the dense, normal lysosomes in PDAC cells exhibited structural disruption following BLU554 treatment. Under electron microscopy, lysosomal structures in BLU554-treated PK-1 cells underwent dramatic alterations, exhibiting highly aberrant morphologies with multiple concentric layers of fine striations reminiscent of myelinosomes [30, 31] (Fig. 4A). Lysosomal structural alterations in T3M-4 cells appeared less pronounced than those in PK-1 cells. These characteristic structures have not been observed in other senescent cells, such as long-passaged normal human fibroblasts, and their direct relationship with cellular senescence remains unknown. However, increased lysosome activity is a well-recognized hallmark of cellular senescence, and myelinosome formation is thought to result from lysosomal overload-induced dysfunction [30]. Thus, although myelinosome-like structures are not considered classical markers of cellular senescence, their appearance may reflect lysosomal dysregulation commonly associated with senescent cells, particularly under drug-induced stress conditions.
Senescent cells are known to secrete various inflammatory cytokines collectively referred to as SASP factors. In this study, BLU554-induced senescence in PDAC cells resulted in differential regulation of SASP-related genes (Fig. 6), suggesting a heterogeneous SASP profile. Notably, GM-CSF expression was upregulated, which may be associated with reduced metastatic potential. Higher GM-CSF levels have been reported in patients with early-stage disease: patients with stage 1 disease exhibited higher levels than those with stage 4 disease [32]. Conversely, elevated IL-1α and IL-1β expression, particularly in PK-1 cells, may contribute to immunosuppressive tumor microenvironments by inducing CCL22 and promoting Treg recruitment [33, 34]. Although IL-6 and TNF-α are frequently elevated in PDAC and linked to poor prognosis [35], suppressed or unchanged expression of these markers in our model suggests that they may not be major contributors following BLU554 treatment. These findings indicate that senescence induced by BLU554 leads to a mixed SASP response, incorporating both tumor-suppressive and tumor-promoting signals, which could influence therapeutic outcomes. Because the concentration of BLU554 used in our PDAC models (50 μM) was nearly identical to that used in hepatocellular carcinoma trials [36], these SASP responses may also be relevant in patients. Accordingly, particular attention should be given to the induction of pro-inflammatory cytokines when using anticancer agents such as BLU554.
Senolytic drugs can exhibit cell type-dependent efficacy [37]. In the present study, BLU554-induced senescent T3M-4 cells showed reduced viability upon senolytic treatment, while PK-1 cells did not respond similarly (Fig. 7). This difference may be attributed to the absence of p16 in PK-1 cells, which lack CDKN2A gene expression. p16 is a key mediator of stable cell cycle arrest during senescence. In support of this, previous studies in mouse embryonic fibroblasts have shown that H-rasV12 fails to induce full senescence in p16-deficient cells, allowing continued proliferation [38]. Although PK-1 cells expressed some senescence markers, the lack of p16 expression may have impaired full senescence induction. This could explain their reduced sensitivity to dasatinib, which selectively eliminates senescent cells by targeting pro-survival pathways.
When BLU554 induced senescence in PDAC cells, treatment with dasatinib decreased cell viability, but quercetin had no effect. In contrast, when senescence was induced by BLU9931 in PDAC cells, a reduction in cell viability was observed only with quercetin, as previously reported [13]. Thus, even though both agents are FGFR4 inhibitors, they produced different responses to senolytic drugs. This discrepancy may be attributable to the broader off-target reactivity of BLU9931 compared with the clinical candidate BLU554. Stoichiometric analyses and the X-ray crystal structure of the FGFR4–BLU554 complex have demonstrated that BLU554 is a more FGFR4-specific inhibitor than BLU9931 [14]. If the difference in senolytic response observed here is also driven by such target selectivity, BLU9931 may additionally inhibit FGFR3. FGFR3 has been reported to promote neuronal apoptosis by inhibiting Bcl-2 and Bcl-xL through PI3K/Akt signaling [39]. Quercetin is known to inhibit the PI3K/Akt pathway, which would be consistent with this mechanism. In contrast, FGFR4, the primary target of BLU554, is known to inhibit Src family kinases. Dasatinib similarly inhibits Src family kinases, thereby promoting apoptotic signaling [40]. Collectively, these findings suggest that the differential sensitivity of BLU554- and BLU9931-induced senescent cells to senolytic drugs is likely due to differences in their specificity for FGFR family members. Future studies should explore agents that can induce senescence or promote senolysis more efficiently, including not only FGFR4 inhibitors but also clinically used agents for pancreatic cancer. Moreover, future animal studies as well as investigations into the differences in metastatic potential of PDAC are required to confirm the therapeutic relevance of senescence induction and senolytic approaches.

Conclusions
The FGFR4-specific inhibitor BLU554 exhibited cytotoxic effects against the PDAC cell lines PK-1 and T3M-4. Additionally, BLU554 induced senescence in a subset of the treated cancer cells. The subsequent elimination of these senescent cells, known to secrete SASP factors that affect the tumor microenvironment, using senolytic drugs further reduced cancer cell viability. Anticancer drug-induced cellular senescence may occur in cancer cells in the clinical setting. This combinational approach is particularly promising for the treatment of pancreatic cancer, a malignancy with limited therapeutic options and extremely poor 5-year survival rates. Future studies should focus on elucidating the mechanisms of senescence induction and identifying effective senolytic drugs, with the aim of improving therapeutic outcomes in pancreatic cancer. Thus, the combination of senescence-inducing agents with senolytic f.

Supplementary Information