2025 Volume 48 Issue 11 Pages 1784-1793
The hypovascular nature of pancreatic tumors creates a nutrient-scarce, hypoxic microenvironment, yet pancreatic cancer cells adapt by altering their metabolism to thrive under austere conditions—a phenomenon known as “austerity.” Targeting this adaptation offers a promising strategy for next-generation therapeutics that selectively impair pancreatic cancer cell viability in nutrient-deprived states without toxicity under nutrient-rich conditions. Here, we evaluated the anti-pancreatic cancer properties of grandifloridin D, a synthetic derivative of (+)-grandifloracin. In vitro antiausterity assays demonstrated that grandifloridin D potently and preferentially reduced the viability of MIA PaCa-2 pancreatic cancer cells under nutrient deprivation at a PC50 concentration of 0.14 μM. Live-cell imaging and ethidium bromide/acridine orange (EB/AO) dual staining confirmed that grandifloridin D induces cell death by disrupting membrane integrity. Under nutrient-rich conditions, grandifloridin D exhibited antimetastatic activity, significantly inhibiting MIA PaCa-2 cell migration in real-time assays and suppressing colony formation and spheroid formation. Western blot analysis revealed that grandifloridin D is a potent inhibitor of the protein kinase B (Akt) and mammalian target of rapamycin (mTOR) signaling pathway while also suppressing the autophagy-related proteins microtubule-associated protein 1 light chain 3 (LC3). These results suggest that grandifloridin D is a promising antiausterity agent for pancreatic cancer drug development.
Pancreatic cancer remains one of the most formidable malignancies and is characterized by rapid progression, late-stage diagnosis, and poor prognosis. With a 5-year survival rate of approximately 5.5%, it ranks among the deadliest cancers, claiming approximately 23000 lives annually in Japan and approximately 10000 in the United Kingdom.1,2) In the European Union, pancreatic cancer has recently surpassed breast cancer as the third leading cause of cancer-related mortality.3) The hypovascular nature of pancreatic tumors creates an austere microenvironment, severely limiting the nutrient and oxygen supply. Under such conditions, pancreatic cancer cells activate alternative survival pathways, including protein kinase B (Akt)/mammalian target of rapamycin (mTOR) signaling and autophagy, enabling them to tolerate nutrient deprivation for extended periods, unlike normal cells, which die within 24 h.4,5) This adaptive capacity contributes to their intrinsic resistance to conventional chemotherapies such as gemcitabine, suggesting the urgent need for novel therapeutic approaches.6)
One promising strategy is the development of antiausterity agents that exploit the metabolic vulnerabilities of pancreatic cancer cells by targeting their tolerance to nutrient starvation, a hallmark of the tumor microenvironment.7) Several natural products with antiausterity activity have been identified, including arctigenin, angelmarin, (+)-grandifloracin, ancistrolikokine E3, nicolaioidesin C, plumbagin derivatives, and toyaburgine.8–14) Notably, arctigenin (GBS-01) has progressed to clinical trials, demonstrating favorable responses in patients with gemcitabine-refractory pancreatic cancer and validating an antiausterity approach.15)
Among these, (+)-grandifloracin, a dimeric polyoxygenated cyclohexene isolated from Uvaria species (e.g., Uvaria dac), has emerged as a promising compound.16,17) It showed preferential cytotoxicity against the PANC-1 pancreatic cancer cell line in nutrient-deprived medium (NDM), with a PC50 of 14.5 μM—defined here as the concentration that caused 50% cell death in NDM without toxicity in nutrient-rich medium Dulbecco’s modified Eagle’s medium (DMEM). Furthermore, (+)-grandifloracin induced PANC-1 pancreatic cancer cell death by inhibiting the Akt/mTOR pathway and hyperactivating autophagy, a dual mechanism that disrupts austerity tolerance.10) Grandifloracin is unusual in that both enantiomers are produced in nature by different species in the Uvaria genus. Our earlier work isolated the (+)-enantiomer from Uvaria dac and demonstrated the antiausterity activity of this enantiomer specifically. However, the opposite enantiomer, (−)-grandifloracin had previously been isolated from Uvaria grandiflora.18,19) Given its interesting antiausterity activity, as well as known synthetic methods to access the grandifloracin core,20–24) we previously explored the structure–activity relationship (SAR) of (±)-grandifloracin (1) derivatives through the synthesis and evaluation of racemic analogs (Fig. 1). Modifications to the ester side chains yielded grandifloracin derivatives, such as (±)-grandifloridins A–D (2–5),25,26) with enhanced antiausterity activity against PANC-1 cells compared with the parent compound, suggesting side chain diversification as a viable strategy to optimize potency without compromising bioactivity. Among these derivatives, compound grandifloridin D (5) emerged as the most potent antiausterity agent against PANC-1 cells, with a PC50 value of 4.8 μM. However, its broader biological effects, particularly against other pancreatic cancer cell lines, and their impact on cancer cell survival mechanisms remain to be explored. Therefore, in this study, we investigated the antiausterity activity of grandifloridin D (5) against the MIA PaCa-2 human pancreatic cancer cell line, an aggressive phenotype with increased migratory and invasive potential.27) These traits mirror the metastatic behavior of advanced pancreatic cancer, which is a primary driver of mortality.28) We evaluated the ability of grandifloridin D (5) to suppress key hallmarks of cancer progression, cell migration, colony formation, and spheroid formation, together with its mechanistic effects on the Akt/mTOR and autophagy pathways, which are critical mediators of nutrient starvation tolerance, tumor growth, and metastasis.14,29)
All reagents were purchased from commercial suppliers and used without further purification, unless stated otherwise. DMEM; FUJIFILM Wako (Osaka, Japan), Cat. #041-29775, fetal bovine serum (FBS; Gibco, Cat. #10270-106), an antibiotic-antimycotic solution (Sigma-Aldrich (St. Louis, MO, U.S.A.), Cat. #A5955), and phosphate-buffered saline (PBS; FUJIFILM Wako, Cat. #166-23555) were used for cell culture. NDM was prepared in DMEM without glucose or FBS (FUJIFILM Wako, Cat. #042-32255). Grandifloridin D was synthesized as previously described and dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich, Cat. #D2650) for the stock solutions.25) Chloroquine (CQ; FUJIFILM Wako, Cat. #038-17971) was dissolved in distilled water for the stocking.
Cell LinesThe human pancreatic cancer cell line, MIA PaCa-2 (RBRC-RCB2094), was obtained from the Riken BRC Cell Bank (Tsukuba, Japan). The cells were maintained in DMEM supplemented with 10% FBS and 1% antibiotic-antimycotic solution at 37°C in a humidified 5% CO2 incubator. Cell passages were maintained at −80°C to ensure consistency.
Preferential Cytotoxicity AssayThe preferential cytotoxicity of grandifloridin D (5) against MIA PaCa-2 cells was assessed using an established protocol. Briefly, cells (2 × 104 cells/100 μL/well) were seeded in 96-well plates (Corning (Corning, NY, U.S.A.), Cat. #3596) and incubated for 24 h. After washing with PBS, the medium was replaced with NDM or nutrient-rich DMEM containing serial dilutions of grandifloridin D (5) (0–100 μM, 0.1% DMSO final). Following 24 h of incubation at 37°C and 5% CO2, the medium was replaced with 100 μL DMEM containing 10% Cell Counting Kit-8 solution (CCK-8, Dojindo (Kumamoto, Japan), Cat. #CK04). After 3 h, absorbance (Abs) was measured at 450 nm using a microplate reader (Bio-Rad, Model 680). Cell viability was calculated as follows:
where Abs(blank) is the absorbance of the wells with medium only. Experiments were performed in triplicate, and PC50 values (concentration that reduced viability by 50%) were determined using GraphPad Prism 9.3.
In subsequent experiments, grandifloridin D (5) was employed at concentrations ranging from 0.5 to 5 μM in NDM to examine its antiausterity properties. To assess anti-metastatic and inhibitory effects of spheroid formation under non-cytotoxic conditions in DMEM, higher concentrations (5–20 μM) were utilized in DMEM to ensure assay specificity.
Morphological AnalysisMorphological changes were evaluated as previously described. MIA PaCa-2 cells (2 × 105 cells/dish) were seeded in a 35-mm dish (Thermo Fisher Scientific (Waltham, MA, U.S.A.), Cat. #150460) and incubated for attachment in DMEM, then treated with grandifloridin D (0.5 or 1.0 μM) or without (control) for 24 h at 37°C and 5% CO2. Assays were conducted in NDM to model the nutrient-scarce primary tumor microenvironment, as rapid cell death in NDM precludes long-term functional evaluations such as migration or spheroid formation, which were performed in DMEM to mimic nutrient-rich metastatic sites. Cells were stained with ethidium bromide/acridine orange (EB/AO, Sigma-Aldrich, Cat. #E2889, A6014; 1 : 1, 100 μg/mL each) for 15 min, and imaged using an EVOS FL Auto Imaging System (Thermo Fisher Scientific) with GFP (AO) and RFP (EB) filters. Live-cell imaging was performed in parallel using CytoSMART Lux2 (CytoSMART Technologies) at 15-min intervals for 24 h, with representative snapshots analyzed (Supplementary Movie S1).
Real-Time Cell Migration AssayCell migration was assessed using a 2-well culture-insert assay (Ibidi, Cat. #80209) using established methods. MIA PaCa-2 cells (1 × 106 cells/mL in DMEM) were seeded (70 μL/well) into inserts within 35-mm dishes (Ibidi, Cat. #81176) and incubated for 24 h at 37°C and 5% CO2. The inserts were removed to create a 500-μm gap, the cells were washed with PBS, and the medium was replaced with DMEM alone (control) or DMEM containing grandifloridin D (10 or 20 μM). Time-lapse imaging was conducted using CytoSMART Lux2 at 15-min intervals for 48 h in a 5% CO2 incubator (Supplementary Movie S2). The cell-free area (% of the initial gap) was quantified from bright-field images using ImageJ software (NIH).
Colony Formation AssayColony formation was evaluated using 24-well plates (Corning, Cat. #353047). MIA PaCa-2 cells (2.5 × 103 cells/well) were seeded in 1 mL of DMEM and incubated for 24 h at 37°C and 5% CO2. Cells were treated with grandifloridin D (5) (0, 1.3, 2.5, 5, 10, and 15 μM) in DMEM for 24 h, the medium was replaced with fresh DMEM, and the cells were cultured for 10 d. Colonies were fixed with 4% formaldehyde (FUJIFILM Wako, Cat. #064-00406), stained with 0.5% crystal violet (Sigma-Aldrich, Cat. #C0775) for 15 min, and washed with PBS. The colony area (% of well surface) was quantified using the ImageJ “ColonyArea” plugin. The experiments were performed in four replicates.
Western Blot AnalysisMIA PaCa-2 cells (5 × 105 cells/mL, 2 mL/well) were seeded in 6-well plates (Corning, Cat. #3516) in DMEM and incubated overnight. The medium was replaced with NDM alone, or NDM with grandifloridin D (1, 2, and 4 μM); DMEM alone, or DMEM with grandifloridin D (4 μM), and cells were incubated for 6 h at 37°C and 5% CO2. For combination treatment assays, in NDM with 6 h of exposure, grandifloridin D (5) was used at 2 or 4 μM, either alone or in combination with the final concentration of CQ at 25 μM. To assess autophagy flux, additional experiments included chloroquine (CQ, 25 μM) as a positive control to induce LC3-II accumulation by blocking autophagosome-lysosome fusion, allowing comparison with grandifloridin D (5) effects. Cells were harvested, washed with ice-cold PBS, and lysed in RIPA buffer (Thermo Fisher Scientific, Cat. #89900) supplemented with a protease inhibitor cocktail (Sigma-Aldrich, Cat. #P8340), 1 mM sodium orthovanadate, 40 mM β-glycerophosphate, and 1 mM phenylmethanesulfonyl fluoride. Lysates were incubated on ice for 30 min, centrifuged at 13000 rpm for 12 min at 4°C, and the supernatants were stored at −20°C. Protein concentration was determined using the BCA assay (Thermo Fisher Scientific, Cat. #23225). Proteins (20 μg/lane) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (10% polyacrylamide gel) and transferred to polyvinylidene difluoride (PVDF) (polyvinylidene fluoride) membranes (Bio-Rad, Cat. #1620177). Membranes were blocked with 5% skim milk in TBST (Tris-buffered saline with 0.1% Tween® 20, Sigma-Aldrich, Cat. #P9416) for 1 h, incubated overnight at 4°C with primary antibodies, washed thrice with TBST, and incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 h at room temperature. The bands were visualized using ECL solution (GE Healthcare, Cat. #RPN2232) and imaged using ImageQuant LAS 4000 (GE Healthcare). Primary antibodies (Cell Signaling Technology) included: Phosphatidylinositol-3 kinase (PI3K) p110α (#4255), PI3K p110β (#3011), p-Akt (S473, #9271), Akt (#9272), p-mTOR (S2448, #2971), mTOR (#2983), LC3 (#2775), and β-tubulin (#2146). Secondary antibodies were anti-rabbit (#P0448) HRP conjugates (Dako, Glostrup, Denmark). The band intensity was quantified using ImageJ software and normalized to β-tubulin.
Spheroid Formation AssayMIA PaCa-2 cells were seeded at a density of 1000 cells per well in a 96-well ultra-low attachment prime surface plate (FUJIFILM Wako, Cat. #MS-9096M). The plates were centrifuged to promote cell aggregation, and the spheroids were allowed to form over 3 d. On day 3, the spheroids were treated with varying concentrations of grandifloridin D (5, 10, and 20 μM) or a control vehicle. The spheroid growth was monitored every 6 h using an OMNI device. On day 6, 50% of the medium was replaced, and the spheroids were allowed to grow until day 10. At the end of the assay, the spheroids were stained with ethidium bromide (EB)/acridine orange (AO) and captured using a fluorescence microscope to assess viability.
Statistical AnalysisData were analyzed using GraphPad Prism 9.3 (GraphPad Software, San Diego, CA, U.S.A.). One-way ANOVA with Dunnett’s multiple comparisons test was used for the cytotoxicity and colony assays. Differences in the area in the spheroid formation assay were analyzed using two-way ANOVA with Tukey’s post hoc test. Western blot data were analyzed using one-way ANOVA or unpaired t-test, with n = 3 (three independent replicates). Significance was set at p < 0.05 (*p < 0.05, **p < 0.01, ***p < 0.001).
Pancreatic cancer cells, including MIA PaCa-2 and PANC-1, demonstrate remarkable tolerance to nutrient deprivation and survive for over 48 h under complete starvation through adaptive metabolic reprogramming, a key feature of austerity. In this study, we investigated grandifloridin D (5)’s ability to block this survival mechanism in MIA PaCa-2 cells in NDM. Grandifloridin D (5) has been found to display potent, concentration-dependent reduction in cell viability in NDM, with a PC50 of 0.14 μM against MIA PaCa-2 cells (Fig. 2). The PC50, defined as the concentration reducing viability by 50% in NDM without toxicity in DMEM, showed grandifloridin D’s selectivity in NDM over DMEM. At 0.5 μM, grandifloridin D (5) induced complete cell death within 24 h, in stark contrast to the untreated controls, which remained viable beyond 48 h.
Cells were treated with grandifloridin D (0–100 μM) in NDM (red) or nutrient-rich DMEM (blue) and assessed by CCK-8 assay after 24 h. Data represent mean ± standard deviation (S.D.) (n = 3; one-way ANOVA with Dunnett’s multiple comparisons test).
This potency significantly exceeds that of (+)-grandifloracin (PC50 = 14.5 μM, PANC-1), the structural inspiration of grandifloridin D (5), and exhibited approximately 34-fold greater selectivity for MIA PaCa-2 (PC50 = 0.14 μM) over PANC-1 (PC50 = 4.8 μM) cells. When compared with other antiausterity agents, such as arctigenin (PC50 ≈ 0.5 μM, PANC-1) and angelmarin (PC50 ≈ 0.2 μM, PANC-1), grandifloridin D’s PC50 of 0.14 μM ranks it among the top effective compounds in this class. These results suggest that grandifloridin D (5) targets nutrient-starved pancreatic cancer cells, offering a potent and selective antiausterity profile.
Live-Cell Imaging Reveals Rapid Morphological Changes and Cell Death Induced by Grandifloridin D (5) in MIA PaCa-2 Cells under Nutrient DeprivationChanges in cell morphology are critical indicators of cellular status and serve as valuable assays in therapeutic research.30) To assess grandifloridin D (5)’s effects on MIA PaCa-2 pancreatic cancer cells under nutrient-deprived conditions, we employed live-cell imaging to monitor dynamic morphological alterations every 15 min over 24 h (Fig. 3A, Supplementary Movie S1), complemented by an ethidium bromide (EB)/acridine orange (AO) dual-staining fluorescence assay (Fig. 3B). Live-cell imaging captures the real-time cytotoxic impact of grandifloridin D (5), while EB/AO staining (where AO marks viable cells green and EB stains compromised cells red) confirms cell death mechanisms. MIA PaCa-2 cells were treated with grandifloridin D (5) at 0.5 and 1.0 μM, alongside an untreated control, and incubated in NDM for 24 h. As shown in Fig. 3A, untreated MIA PaCa-2 cells remained intact with normal morphology over 24 h in NDM. By contrast, cells treated with 0.5 μM grandifloridin D (5) exhibited dramatic rounding within 12 h, progressing to membrane blebbing, and complete cellular content leakage into the medium by 24 h. At 1.0 μM, these alterations occurred even faster, with pronounced morphological changes evident within 3 h, causing in extensive cell death within 24 h (Supplementary Movie S1). EB/AO staining supported these findings, showing untreated cells as uniformly green and viable, while grandifloridin D (5)-treated cells (0.5 and 1.0 μM) displayed red fluorescence, indicating disrupted membrane integrity and necrotic death (Fig. 3B). These results suggest grandifloridin D (5) as a promising antiausterity agent against pancreatic cancer cells.
Time-lapse images of cells in NDM, either untreated or treated with grandifloridin D (0.5 μM or 1.0 μM), were captured at 0, 12, and 24 h. After 24 h, fluorescence and phase-contrast images were acquired following staining with ethidium bromide/acridine orange (EB/AO; red/green) and visualized using an EVOS FL digital microscope.
Metastasis remains the primary cause of mortality in cancer patients and is driven by a multi-step process involving angiogenesis, tumor cell detachment, invasion into the circulatory system, and formation of micrometastases in nutrient-rich distant organs.31) Thus, inhibiting cancer cell migration under normal nutrient conditions is a critical strategy for the development of antimetastatic therapeutics.32)
To investigate the antimetastatic potential of grandifloridin D (5), we utilized real-time live-cell imaging to monitor the dynamic inhibition of MIA PaCa-2 pancreatic cancer cell migration in a nutrient-rich environment, providing visual evidence of its effects over time (Fig. 4, Supplementary Movie S2). MIA PaCa-2 cells were seeded into a 2-well culture insert and incubated for 24 h under standard conditions (37°C, 5% CO2, humidified). After that, the insert was removed to create an approx. 500-μm cell-free gap. The medium was then replaced in treated groups with fresh medium containing grandifloridin D at 10 or 20 μM or left untreated (control). Time-lapse imaging captured cell migration every 15 min for 48 h, yielding 193 frames (Fig. 4A). In the untreated control, MIA PaCa-2 cells rapidly migrated into the gap, reducing the open area from 100% at 0 h to 53% at 24 h and 6% at 48 h (Fig. 4B). Conversely, grandifloridin D (5) treatment markedly inhibited this process in a concentration-dependent manner; at 10 μM, the gap remained 66% open at 24 h and 23% at 48 h; at 20 μM, it remained open at 94% at 24 h and 92% at 48 h (Fig. 4B, Supplementary Movie S2).
Real-time live-cell imaging of MIA PaCa-2 cells in nutrient-rich DMEM, untreated or treated with grandifloridin D (10 or 20 μM), was performed at 15-min intervals over 48 h using a CytoSMART system (Supplementary Movie S2). Representative images depict wound closure at 0, 12, 24, 36, and 48 h, and the open area (% of initial gap) was quantified from bright-field images at the indicated time points.
Live-cell imaging provided critical real-time insights into these dynamics, revealing that untreated cells exhibited directional migration leading to open area closure within 24 h, whereas grandifloridin D (5)-treated cells showed reduced motility, and open areas remained largely open, particularly at a concentration of 20 μM. This visual evidence suggests that grandifloridin D (5) disrupts cancer cell migratory behavior, a key metastatic trait, under nutrient-rich conditions that mimic metastatic target sites. These findings suggest that grandifloridin D (5) is a potent antimetastatic agent.
Grandifloridin D (5) Inhibits MIA PaCa-2 Colony Formation under Nutrient-Rich ConditionsFollowing invasion of the circulatory or lymphovascular system, metastatic cancer cells disseminate to distant organs with more favorable nutrient conditions than the primary tumor site. These cells adapt to foreign microenvironments, forming colonies that develop into secondary tumors.33) In pancreatic cancer, most patients are diagnosed at advanced stages when local therapies, such as resection or radiotherapy, are ineffective, leading to poor survival outcomes.34) Thus, preventing colony formation is a critical antimetastatic strategy. Here, we evaluated the ability of grandifloridin D (5) to inhibit MIA PaCa-2 pancreatic cancer cell clonogenicity under nutrient-rich conditions.
MIA PaCa-2 cells were seeded in 24-well plates, treated with grandifloridin D (5) at various concentrations (0–15 μM) in DMEM for 24 h, and then cultured in fresh DMEM for an additional 10 d to allow colony formation. Colony area was quantified relative to the surface area of the wells (Fig. 5B). In the untreated control, MIA PaCa-2 cells formed large colonies that occupied 94% of the well area. By contrast, grandifloridin D (5) suppressed colony formation in a concentration-dependent manner; at 5 μM, colony area decreased to approximately 80%, while at 15 μM, it was reduced to approximately 3%, with virtually no colonies observed (Fig. 5A). This near-complete inhibition at higher concentrations suggests that grandifloridin D (5) disrupts critical processes such as cell adhesion, proliferation, or survival post-migration.
After a 24-h treatment with 0–15 μM grandifloridin D, colonies were cultured for 10 d and quantified as a percentage of well surface area. Data represent mean ± S.D. (n = 4), with significant reductions observed at 5 μM (p = 0.03), 10 μM (p = < 0.001), and 15 μM (p = < 0.001) compared with control, while 2.5 μM showed no significant effect (p = 0.74); significance determined by one-way ANOVA with Dunnett’s multiple comparisons test (*p < 0.05, *p < 0.01, ***p < 0.001). Panel (B) shows representative images of colonies at the indicated concentrations.
These findings complement the anti-migration effects of grandifloridin D (5), highlighting its anti-metastatic potential in vitro, suggesting grandifloridin D as a promising candidate for preventing pancreatic cancer metastasis under nutrient-rich conditions that mimic secondary tumor sites.
Inhibition of Spheroid Growth by Grandifloridin D (5) in Real-Time AssaysTraditional two-dimensional (2D) monolayer cultures have long been a cornerstone of in vitro cancer research; however, their limitations in recapitulating the complex tumor microenvironment (TME) are well documented. These models fail to mimic the critical features of solid tumors, such as cell–cell interactions, hypoxia, nutrient gradients, and chemotherapeutic resistance, which are hallmarks of pancreatic ductal adenocarcinoma.35) To address these shortcomings, three-dimensional (3D) spheroid cultures have emerged as powerful tools for bridging the gap between 2D cultures and in vivo models. Spheroids replicate the TME better by fostering cell–cell and cell–matrix interactions, creating hypoxic cores, and exhibiting drug penetration barriers similar to those observed in PDAC tumors.36,37) In this study, we used a 3D spheroid formation assay to assess the efficacy of grandifloridin D (5) in inhibiting the growth of MIA PaCa-2 spheroids. Through real-time imaging and quantitative analysis, we evaluated spheroid growth dynamics and cell viability over 10 d at various concentrations of grandifloridin D (5) (Fig. 6, Supplementary Movie S3). Representative images (Fig. 6A) depict the morphology of spheroids on days 0, 3, 6, and 10 in both control and grandifloridin D (5)-treated groups (5, 10, and 20 μM). Control spheroids demonstrated continuous growth, resulting in the formation of compact structures over time. Conversely, spheroids treated with grandifloridin D (5) exhibited dose-dependent inhibition of growth, with higher concentrations (10 and 20 μM) resulting in a significant reduction in spheroid size by day 10. Quantitative analysis of the spheroid area (Fig. 6B) confirmed these observations. Control spheroids reached an average area of approximately 1.3 × 106 μm2 by day 10, indicating uninterrupted growth. Spheroids exposed to 10 and 20 μM grandifloridin D (5) exhibited significantly reduced growth rates, with final areas approximately 40 and 70% smaller, respectively, than those of the controls (*p = 0.03 and ***p = 2.5e-06, respectively). Notably, the 20 μM group displayed signs of spheroid disintegration by day 10, indicating pronounced cytotoxicity. At the end of the experiment, cell viability within the spheroids was assessed by EB/AO staining on day 10 (Fig. 6C). Control spheroids predominantly emitted green fluorescence, which was consistent with a high proportion of viable cells. By contrast, grandifloridin D (5)-treated spheroids showed a dose-dependent increase in red fluorescence, indicating cell death or damage. The 20 μM group exhibited the highest red-to-green fluorescence ratio, with spheroids comprising predominantly nonviable cells, confirming the observed reduction in the spheroid area. Longati et al.38) reported that PDAC spheroids exhibit resistance to gemcitabine, a standard chemotherapeutic agent, due to impaired drug penetration and upregulated survival pathways. By contrast, grandifloridin D (5)’s efficacy against spheroid formation suggests that it may overcome such barriers, potentially by disrupting cell–cell interactions or inducing apoptosis within the spheroid core. The dose-dependent increase in EB staining further supports the cytotoxic effects of grandifloridin D (5), which is consistent with the mechanisms observed for other novel PDAC therapeutics, such as those targeting hypoxic signaling or matrix remodeling.39)
(A) Representative bright-field images of spheroids cultured in DMEM (control) or treated with 5, 10, or 20 μM grandifloridin D on days 0, 3, 6, and 10 (scale bar: 1000 μm). (B) Quantification of spheroid area over 10 d, measured every 6 h, showing significant reductions at 10 μM (p = 0.03) and 20 μM (p < 0.001) compared with the control; data represent the mean ± standard error of the the mean (S.E.M.) (n = 3), with significance assessed by two-way ANOVA followed by Tukey’s post hoc test (*p < 0.05, **p < 0.01, ***p < 0.001). (C) Fluorescent images of spheroids stained with ethidium bromide/acridine orange (EB/AO) on day 10 (scale bar = 500 μm).
The PI3K/Akt/mTOR pathway is a pivotal regulator of cellular processes and is frequently dysregulated in cancers, including pancreatic adenocarcinoma, where it drives proliferation, survival, angiogenesis, epithelial-to-mesenchymal transition (EMT), and chemoresistance. Targeting this pathway is a well-established therapeutic strategy, with FDA-approved inhibitors such as capivasertib, everolimus, and vistusertib demonstrating clinical efficacy. Similarly, antiausterity agents such as arctigenin, (+)-grandifloracin, (+)-panduratin A, and 4′-O-methylgrynullarin have been shown to inhibit PI3K/Akt/mTOR signaling. Given grandifloridin D (5)’s structural inspiration from (+)-grandifloracin, we investigated its effects on this pathway and autophagy, a stress response mechanism, in MIA PaCa-2 pancreatic cancer cells.
Cells were treated with grandifloridin D (5) (0, 1, 2, and 4 μM) in NDM or with 0 and 4 μM in nutrient-rich DMEM for 6 h. Western blot analysis assessed the expression of PI3K (p110α, p110β), Akt, phosphorylated Akt (p-Akt S473), mTOR, phosphorylated mTOR (p-mTOR S2448), and autophagy markers LC3-I and LC3-II using β-tubulin as a loading control (Fig. 7A). In NDM, grandifloridin D (5) reduced p-Akt (S473) and mTOR levels (**p = 0.008 and **p = 0.004, respectively; Fig. 7B) in a concentration-dependent manner, with significant inhibition observed at 2 μM. Total PI3K (p110α, p110β), Akt, and p-mTOR (S2448) levels showed no consistent changes. There are no significant alterations in protein expression under DMEM.
(A) Protein levels of PI3K subunits (p110α, p110β), Akt, phosphorylated Akt (p-Akt S473), mTOR, phosphorylated mTOR (p-mTOR S2448), and autophagy markers LC3-I and LC3-II were measured after 6 h of treatment with grandifloridin D (0, 1, 2, and 4 μM in NDM); 0 and 4 μM in nutrient-rich DMEM, with β-tubulin as loading control. (B) Quantification of β-tubulin-normalized protein levels (n = 3) relative to the control showed no significant change for PI3K p110α (NDM p = 0.40, DMEM p = 0.99) and p110β (NDM p = 0.91, DMEM p = 0.53); LC3 showed significance only at 4 μM in NDM (p = 0.03) but not in DMEM (p = 0.60); Akt levels were unchanged (NDM p = 0.96, DMEM p = 0.06); phosphorylated Akt (p-Akt S473) was significantly reduced at 2 μM (p = 0.008) and 4 μM (p = 0.001) in NDM but not DMEM (p = 0.41); mTOR was significantly decreased at 2 μM (p = 0.004) and 4 μM (p = < 0.001) in NDM but unchanged in DMEM (p = 0.21); phosphorylated mTOR (p-mTOR S2448) levels showed no significant differences (NDM p = 0.46, DMEM p = 0.44). Statistical analyses were performed using one-way ANOVA with Dunnett’s multiple comparisons test for NDM and unpaired t-tests for DMEM samples (*p < 0.05, **p < 0.01, ***p < 0.001).
Autophagy, a dual-natured process in pancreatic cancer, supports survival under nutrient deprivation but can also promote cell death when dysregulated.40–42) The key marker, LC3 (LC3-I, 16 kDa; LC3-II, 14 kDa), reflects autophagosome formation and flux.43) In NDM, grandifloridin D (5) dose-dependently downregulated LC3-I and LC3-II (Fig. 7A), with significant reductions at 4 μM (*p = 0.03; Fig. 7B), indicating autophagy suppression. To further elucidate the role of grandifloridin D (5) in autophagy inhibition, we performed autophagy flux analysis using chloroquine (CQ), a late-stage autophagy inhibitor that blocks autophagosome-lysosome fusion, leading to LC3-II accumulation. MIA PaCa-2 cells were treated with grandifloridin D (5) (0, 2, or 4 μM) alone or in combination with CQ (25 μM) in NDM for 6 h. Western blot analysis (Fig. 8A) showed that CQ monotherapy induced a significant approx. 2-fold increase in LC3-II levels compared with untreated controls (p < 0.01), consistent with blocked autophagic flux. By contrast, grandifloridin D (5) monotherapy reduced LC3-I/II expression, with significant inhibition at 4 μM (p = 0.04). Co-treatment with CQ and grandifloridin D (5) attenuated CQ-induced LC3-II accumulation in a dose-dependent manner, with a significant reduction at 4 μM (p = 0.002; Fig. 8B). These findings indicate that grandifloridin D (5) inhibits autophagy at an early stage, likely by interfering with autophagosome formation, thereby disrupting adaptive survival under nutrient stress.
(A) Expression of autophagy markers LC3-I and LC3-II after 6 h of exposure to grandifloridin D (0, 2, and 4 μM) with or without 25 μM CQ in NDM. (B) Quantification of β-tubulin-normalized protein levels (n = 3) relative to the control with grandifloridin D (5), showing significant inhibition of LC3-II expression at 4 μM (p = 0.04). The co-treated groups with 25 μM CQ and 4 μM grandifloridin D (5) showed a significant difference (p = 0.002). Statistical analyses were performed using one-way ANOVA with Dunnett’s multiple comparison test (*p < 0.05, ##p < 0.01).
These findings highlight the dual action of grandifloridin D (5), which suppresses Akt/mTOR signaling and autophagy in early stages, to preferentially induce MIA PaCa-2 cell death in NDM. This synergy may disrupt the metabolic plasticity that enables pancreatic cancer cells to thrive in hypoxic, nutrient-scarce microenvironments, offering a mechanistic basis for its potency (PC50 = 0.14 μM).
The hypovascular nature of pancreatic cancer creates a nutrient-scarce, hypoxic microenvironment, driving adaptive mechanisms such as austerity that enable tumor cells to survive and metastasize under stress. Targeting these adaptations offers a novel therapeutic strategy distinct from conventional cytotoxic agents by exploiting cancer cell vulnerabilities in nutrient-deprived states. Here, we demonstrated that grandifloridin D (5), a synthetic derivative inspired by (±)-grandifloracin (1), exhibits potent antiausterity and antimetastatic properties against MIA PaCa-2 pancreatic cancer cells, thereby advancing anticancer drug discovery through a dual mechanistic approach.
Our in vitro antiausterity assays revealed grandifloridin D (5)’s remarkable selectivity, with a PC50 of 0.14 μM in NDM, surpassing the potency of known antiausterity agents like arctigenin (PC50 ≈ 0.5 μM). Live-cell imaging revealed rapid morphological alterations, with cells rounding within 12 h at a concentration of 0.5 μM, and complete leakage of cellular contents by 24 h, indicative of necrotic cell death, as confirmed by EB/AO staining; this observation is consistent with previous research indicating that nutrient deprivation increases the sensitivity of pancreatic cancer cells to agents that disrupt cellular membranes. In contrast to broad-spectrum cytotoxic agents, the limited efficacy of grandifloridin D (5) in nutrient-rich DMEM highlights its specificity for nutrient-deprived conditions, which is a significant advantage for minimizing off-target toxicity.
Beyond its role in austerity, grandifloridin D (5) has demonstrated antimetastatic properties in vitro under standard nutrient conditions. Metastasis presents a significant clinical challenge, as it is responsible for over 90% of the deaths associated with pancreatic cancer. Pancreatic cancer cells from their hypovascular microenvironment tend to migrate to organs abundant in nutrients. Real-time imaging showed that grandifloridin D (5) (20 μM) inhibited MIA PaCa-2 migration by maintaining 92% open area after 48 h, in contrast to the near-complete closure (6%) in the controls. This mirrors the migration-suppressing effects of Akt inhibitors, such as MK-2206, in pancreatic models.44) Grandifloridin D (5) reduced colony formation by 3% at 15 μM, similar to the anti-clonogenic activity of everolimus. These findings suggest that grandifloridin D (5) disrupts metastatic progression under nutrient-rich conditions, mimicking secondary sites and complementing its antiausterity role. In addition, grandifloridin D (5) showed concentration- and time-dependent inhibition of spheroid formation in real time over a period of 10 d, a model that closely mimics the TME. Spheroids replicate the TME better by fostering cell–cell and cell–matrix interactions, creating hypoxic cores, and exhibiting drug penetration barriers similar to those observed in PDAC tumors, suggesting that grandifloridin D (5) is a potential candidate for preclinical evaluation.
Mechanistically, the inhibition of Akt/mTOR and autophagy pathways by grandifloridin D (5) underpins its efficacy. Western blotting revealed significant reductions in p-Akt (S473), mTOR, and LC3-I/II levels at 2–4 μM in NDM, consistent with the suppression of Akt signaling by (±)-grandifloracin (1). The selective reduction in total mTOR without changes in p-mTOR (S2448) suggests that grandifloridin D (5) primarily affects protein abundance, potentially via enhanced degradation, rather than direct phosphorylation modulation, aligning with patterns seen in rapamycin analogs. Furthermore, in the co-treatment assay with CQ, grandifloridin D (5) was found to inhibit the autophagy pathway at an early stage by downregulating the expression of LC3-II in a dose-dependent manner. The PI3K/Akt/mTOR pathway is hyperactive in more than 70% of pancreatic cancers and sustains proliferation and survival under stress. Autophagy, meanwhile, supports tumor cell resilience in nutrient-scarce microenvironments, as observed in KRAS-driven models. By downregulating both, grandifloridin D (5) mirrors dual-targeting strategies similar to those of rapamycin analogs, but its antiausterity selectivity offers a unique edge.
In conclusion, grandifloridin D (5) represents a significant advancement in the development of grandifloracin derivatives as potential therapeutic agents against pancreatic cancer. This compound demonstrates strong efficacy in disrupting the austerity-driven survival mechanisms of pancreatic tumor cells. It effectively suppresses Mia PaCa-2 cell migration, inhibits colony and spheroid formation, and downregulates critical survival pathways, including Akt/mTOR signaling and autophagy in its early stages. By targeting the adaptive resilience of pancreatic cancer cells, grandifloridin D holds promise as an anticancer agent for next-generation therapies aimed at exploiting vulnerabilities within the tumor microenvironment.
This research was supported by the Japan Society for the Promotion of Science (JSPS) through KAKENHI Grant Number JP23K02104 awarded to SA. HHN gratefully acknowledges the Otsuka Toshimi Scholarship Foundation for providing financial support.
SA: Conceptualization, supervision, and funding acquisition; HHN, JM: methodology; SA, HHN, JM: formal analysis; HHN, JM, TF, LC, SL: investigation; SA, HHN, SL: writing—review and editing; HHN, SA: writing—original draft preparation. All authors have reviewed and agreed to the published version of the manuscript.
The authors declare no conflict of interest.
This article contains supplementary materials. They are available at:
Supplementary PDF: 1H NMR, 13C NMR, COSY, NOESY, HMBC, HMQC and HRFABMS spectra of synthesized grandifloridin D; real-time image stacks from the MIA PaCa-2 cell migration assay, spheroid formation assay; Western blot images from three independent in vitro replicates, including autophagy flux analysis with chloroquine co-treatment.
Supplementary Movie S1: Real-time video of MIA PaCa-2 cells under nutrient-deprived conditions, demonstrating grandifloridin D’s effects on morphology via live-cell imaging (24 h, 15-min intervals).
Supplementary Movie S2: Real-time video of MIA PaCa-2 cell migration in nutrient-rich DMEM, untreated or treated with grandifloridin D (48 h, 15-min intervals).
Supplementary Movie S3: Real-time video of MIA PaCa-2 spheroid formation in nutrient-rich DMEM, untreated or treated with grandifloridin D (10 d, 6-h intervals).
